Because the sidgen plugin is a postop plugin, it is not
always triggered before the result of an ADD is returned
and the objectclasses of the user may / may not contain
ipantuserattrs.
Fix the expected object classes.
Related: https://pagure.io/freeipa/issue/9062
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
IPA DNS zone and forwardzone commands allow to use nameservers with ports
as "SERVER_IP port PORT_NUMBER". bind is supporting this syntax, but the
Resolver in dnspython that is used to verify the list of forwarders
(nameservers) is only allowing to have IP addresses in this list. With
dnspython version 2.20 there is a new validator in dns.resolver.BaseResolver
that ensures this.
Refs:
- https://bind9.readthedocs.io/en/v9_18_4/reference.html#zone-statement-grammar
- https://github.com/rthalley/dnspython/blob/master/dns/resolver.py#L1094
ipapython/dnsutil.DNSResolver derives from dns.resolver.Resolver. The setter
for nameservers has been overloaded in the DNSResolver class to split out
the port numbers into the nameserver_ports dict { SERVER_IP: PORT_NUMBER }.
After the setter for nameservers succeeded, nameserver_ports is set.
nameserver_ports is used in the resolve() method of dns.resolver.Resolver.
Additional tests have been added to verify that nameservers and also
nameserver_ports are properly set and also valid.
Fixes: https://pagure.io/freeipa/issue/9158
Signed-off-by: Thomas Woerner <twoerner@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There are currently three sets of CA schema changes applied
in ipa-server-upgrade:
* addition of ACME schema
* addition of certificate profile schema
* addition of lightweight CA schema
None of these require a restart of the CA to be supported.
There is an issue in schema parsing such that it doesn't handle
X-ORIGIN properly. A difference is detected and a change applied
but no change is recorded in LDAP so every time upgrade is
run it thinks a CA restart is needed. The CA is not quick to
restart so avoiding one is best, particularly when the update is
run as part of an rpm transaction where a user with an itchy finger
may think things have hung and break out of it.
https://github.com/389ds/389-ds-base/issues/5366 was
filed to track this.
Related: https://pagure.io/freeipa/issue/9204
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
New Tests include
TestRSNPKIConfig
TestRSNVault
The new tests are just extending existing classes to be run
with random serial numbers enabled
The tests also include a new method to check params set in CS.cfg for both CA and
KRA, and another test to check Random Serial Number version while
running `ipa ca-find`
Added nightly definitions
Related Ticket: https://pagure.io/freeipa/issue/2016
Signed-off-by: Sumedh Sidhaye <ssidhaye@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
There was no support for setting the grace login limit on the WebUI. The
only way to so was only via CLI:
`ipa pwpolicy-mod --gracelimit=2 global_policy`
Thus, the grace login limit must be updated from the policy section and
this will reflect also on the user settings (under the 'Password Policy'
section)
Fixes: https://pagure.io/freeipa/issue/9211
Signed-off-by: Carla Martinez <carlmart@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This testcase checks that when ipa-client-install command
is run with --subid option, /etc/nsswitch.conf file is updated
with the below entry
subid: nss
Related: https://pagure.io/freeipa/issue/9159
Since the newly added testsuite required client
system, hence modified the below yaml files to change the topology
from *master_1repl to *master_1repl_1client in the below files
gating.yaml
nightly_latest.yaml
nightly_rawhide.yaml
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The test TestInstallDNSSECFirst is failing because of one of its
dependencies (the most likely suspect is the update of openssl-pkcs11).
Disable the test from azure gating until the issue is solved.
Related: https://pagure.io/freeipa/issue/9216
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Carla Martinez <carlmart@redhat.com>
Originally the code treated 0 as disabled. This was
changed during the review process to -1 but one remnant
was missed effetively allowing gracelimit 0 to also mean
disabled.
Add explicit tests for testing with gracelimit = 0 and
gracelimit = -1.
Also remove some extranous "str(self.master.domain.basedn)"
lines from some of the tests.
Fixes: https://pagure.io/freeipa/issue/9206
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
With some recent changes the Azure Agent has decreased performance.
For example, `InstallDNSSECFirst_1_to_5` (5 subjobs) job took ~33min
and now it takes ~40min. In the same time there are jobs having only
1 or 2 subjobs and they should be used more.
Fixes: https://pagure.io/freeipa/issue/9207
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
An ACI with rights of read, write, search and/or compare without
attributes to apply the rights to is effectively a no-op. Allow
the ACI to be created but include a warning. Ignore the add
and delete rights. While they make no sense in the context of
the other rights we should still warn that they are a no-op
with no attributes.
Use the existing make_aci() object method to create the
message and update the add/mod callers to capture and add the
message to the result if one is provided.
When updating an existing ACI the effective attributes will
not be included so fall back to the attributes in the resulting
permission.
Prior to checking for rights and attributes convert any deprecated
names for older clients into the newer values needed by make_aci
This is exercised by existing xmlrpc permission tests that
create such permissions without attributes.
https://pagure.io/freeipa/issue/9188
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The vagrant box must be freeipa/ci-ipa-4-10-f36 on this branch
instead of freeipa/ci-master-f36.
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
Pagination size must be required, the current validators are triggered after
form is submitted, thus the only way for check if data is not empty is by making
the field required.
Fixes: https://pagure.io/freeipa/issue/9192
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The user's pwd expiration was retrieved but inadvertently was never
compared to current time. So any LDAP bind, including from the
IPA API, counted against the grace period. There is no need to go
through the graceperiod code for non-expired passwords.
https://pagure.io/freeipa/issue/1539
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Currently the cert_dir attribute is only present if IPA installation was
done. If IPA was not installed the attribute does not exist.
In order that the uninstall code finds the attribute a class attribute
is added.
Pagure Issue: https://pagure.io/freeipa/issue/9179
Signed-off-by: Sumedh Sidhaye <ssidhaye@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In the memory check failure message, add a hint to the administrator
that they can use the --skip-mem-check flag to skip the check.
Related: https://pagure.io/freeipa/issue/8404
Signed-off-by: Fraser Tweedale <frase@frase.id.au>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Some of the test data are not expected to cause a rewrite in
the upgrade code. Those that do will set the rewrite flag.
In that case there is a new server.xml to be read. This is
handled with mock_open(). The contents can be retrieved via
mocked_file().write.call_args but the repr() of it is:
call(b'<Server port="1234" shutdown="SHUTDOWN">\n ...')
In at least Python 3.10 one can use write.call_args.args to get
just the raw data. This does not work with Python 3.6.8 and
returns the string 'args' instead results in a TypeError.
TypeError: a bytes-like object is required, not 'str'
Instead drop the args and use the data directly.
For the case of x = mocked_file().write.call_args:
x[0] is a tuple with the first element being the data
x[0][0] is the raw data
So use x[0][0] to get at the data instead of x.args[0]
Fixes: https://pagure.io/freeipa/issue/9190
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Named is able to find plugins if they are installed in the plugindir,
so drop the hardcoded path from named.conf.
Signed-off-by: Timo Aaltonen <tjaalton@debian.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The library moved there some years ago, and the compat symlink might go
away at some point. Better prepare for it.
Signed-off-by: Timo Aaltonen <tjaalton@debian.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When ipa user-del --preserve is called, the command output
prints a summary with:
Deleted user: user1
although the user was preserved.
Replace the summary with
Preserved user: user1
to reflect what was actually done.
Fixes: https://pagure.io/freeipa/issue/9187
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
New box contains the fixed mock configuration needed when
`enable_testing_repo` is True.
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Increase the default timeout for expect function when testing
interactive mode to mitigate an issue when the tests are failing
on the slow systems.
Fixes: https://pagure.io/freeipa/issue/9183
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
Change xfail for test_replica_promotion.py/TestHiddenReplicaPromotion/test_ipahealthcheck_hidden_replica
to respect platform and pki version as the related issue is fixed.
Implement tasks/get_platform_version which returns a platform version
number(s) of a provided host in a form of a tuple.
Related: https://pagure.io/freeipa/issue/8582
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
One some distributions, namely Suse, the SSSD_PUBCONF_KRB5_INCLUDE_D_DIR
does not exist by default. Ipa-client-install will fail to initialize
the kerberos ticket and error when this directory does not exist.
This patch simply creates the directory if it does not exist before
adding the include statement into /etc/krb5.conf
Fixes: https://pagure.io/freeipa/issue/9174
Signed-off-by: Matthew Davis github@virtual.drop.net
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
For ID overrides 'ipaOriginalUid' value should be the human-readable
version of the ID override anchor. Since we would have it already set in
the ID override entry, prefer using it instead of looking up the
override anchor.
This should speed up significantly operations which list all ID
overrides in the view, like Web UI views.
Fixes: https://pagure.io/freeipa/issue/9178
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The LDAP cache log is rather chatty and a bit overwhelming when
looking for error messages. Disable it by default but allow it
to be enabled when a new config option, ldap_cache_debug, is
enabled.
Fixes: https://pagure.io/freeipa/issue/9180
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
PKI 10.10 unconditionally added an upgrade script for the AJP
connector which replaced the AJP secret regardless of tomcat
version. It replaced requiredSecret with secret. IPA expects
the attribute by version so this could make the secrets out of
date and/or have connectors with both secrets and different
values.
PKI commit e70373ab131aba810f318c1d917896392b49ff4b has since
been reverted but there may be servers with both secrets still.
On next IPA upgrade clean them up.
Also allow re-writing ipa-pki-proxy.conf in case the secret
changes to ensure they remain in sync.
Fixes: https://pagure.io/freeipa/issue/9176
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The ACI "Read DNS entries from a zone" is defined when
ipa-server-upgrade is run but not for new installations.
In order to have consistent ACI (same set for new install
and for install + upgrade), define this ACI in
install/share/dns.ldif instead of "Allow read access".
Fixes: https://pagure.io/freeipa/issue/9173
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Now that IPA API properly maps LDAP boolean attributes to the
python bool type, they are displayed as True/False instead
of TRUE/FALSE in the ipa *-show outputs.
Update the expected output for DNS Active Zone.
Related: https://pagure.io/freeipa/issue/9171
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
nsds5replicaUpdateInProgress is defined in LDAP schema as a boolean.
Now that IPA API is able to properly map booleans to the python
bool type, this attribute is not a string any more and
comparisons can be done directly based on its real type.
The code in ipa-replica-install was reading nsds5replicaUpdateInProgress
and calling value.tolower() == 'true' but should now use
value == True instead.
Related: https://pagure.io/freeipa/issue/9171
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This commit updates the subordinate design document to reflect the current state
and remove "outdated" message.
Signed-off-by: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This testscase checks that when permission of
ipaserver-upgrade.log
file is changed healtcheck tool reports the correct warning message.
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Pre ipa-healthcheck 0.11, failures detected by DogtagCertsConnectivityCheck
were reported as:
"msg": "Request for certificate failed, <error>"
but the output is now the following:
"msg": "Request for certificate failed: {error}"
"error": <error>
Update the expected output to be compatible with both versions.
Fixes: https://pagure.io/freeipa/issue/9175
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In IPA framework we don't properly convert to Python bool type and just
return a string (TRUE or FALSE). This can be seen with many boolean
attributes, like
Bool('idnsallowdynupdate?',
cli_name='dynamic_update',
label=_('Dynamic update'),
doc=_('Allow dynamic updates.'),
attribute=True,
default=False,
autofill=True
),
in 'ipa dnszone-show':
> > > api.Command.dnszone_show('ipa.test')['result']['idnsallowdynupdate']
['TRUE']
This is because we don't have the reverse (from LDAP to Python) mapping
for the LDAP boolean OID 1.3.6.1.4.1.1466.115.121.1.7.
When Web UI asks for the entry, it gets back JSON output that contains
this 'TRUE' value:
"idnsallowdynupdate": [
"TRUE"
],
Add proper mapping from LDAP to Python bool type. With this, a simple
'checkbox' type can be used in Web UI instead of a complex radio-box
setup.
Note that when IPA API is asked to return raw values, 'TRUE' and 'FALSE'
still returned. These are the actual LDAP boolean attribute values. Care
needs to be done in tests:
- if output is from a command with --raw option, 'TRUE' or 'FALSE'
should be expected
- if output if from a normal (non-raw) command, True or False would be
returned
Fixes: https://pagure.io/freeipa/issue/9171
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Extend existing tests and enabled random serial numbers
during installation. A subset of tests was identified that
exercise the code sufficiently to ensure proper operation.
Update the xml-rpc test to allow the new RSN version number
attribute.
Update some certificate tests where a specifc serial number
is expected.
Fixes: https://pagure.io/freeipa/issue/2016
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Transmitting a big integer like a random serial number over
either xmlrpc or JSON is problematic because they only support
32-bit integers at best. A random serial number can be as big
as 128 bits (theoretically 160 but dogtag limits it).
Treat as a string instead. Internally the value can be treated
as an Integer to conversions to/from hex as needed but for
transmission purposes handle it as a string.
Fixes: https://pagure.io/freeipa/issue/2016
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Remove a duplicated policy change which sets the gracelimit
to 3.
We don't typically run tests individually but as a whole. If
we ever need to call this one test directly we can ignore
failures.
Fixes: https://pagure.io/freeipa/issue/9167
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The graceperiod plugin was only being enabled on new
installations. Enable also on upgrade.
Loading a new plugin requires a restart. Do so if a
new one is configured.
Fixes: https://pagure.io/freeipa/issue/1539
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
This had been hardcoded to debug=True but it spams the logs
with a lot of unnecessary information.
Allow it to be enabled for troubleshooting purposes but keep it
disabled by default.
Enabling debug would involve created /etc/ipa/dns.conf:
[global]
debug = True
I didn't add a more generic mechanism because for now we only need
the value of debug and it introduces a lot of type conversion
headaches. ipalib handles this automatically but to duplicate this
would be corner-case city.
Fixes: https://pagure.io/freeipa/issue/9128
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Include functions to manage IdM service using ipactl, in particular
starting, stopping and restarting the service.
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Test for RFE: Improve error message with more detail for
ipa-replica-install command. If the replication agreement already
exists, check if the error message contains
a particular command needed to delete it.
Related: https://pagure.io/freeipa/issue/9162
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
These pre-existed in ipaserver/install/replication.py.
The constants were only originally used in ldapupdate.py
but have subsequently been switched to the replication.py
versions so they are not used anywhere in the code.
https://pagure.io/freeipa/issue/1539
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Add support for bind grace limiting per
https://datatracker.ietf.org/doc/html/draft-behera-ldap-password-policy-06
389-ds provides for alternative naming than the draft, using those
instead: passwordGraceUserTime for pwdGraceUserTime and
passwordGraceLimit for pwdGraceLoginLimit.
passwordGraceLimit is a policy variable that an administrator
sets to determine the maximum number of LDAP binds allowed when
a password is marked as expired. This is suported for both the
global and per-group password policies.
passwordGraceUserTime is a count per-user of the number of binds.
When the passwordGraceUserTime exceeds the passwordGraceLimit then
all subsequent binds will be denied and an administrator will need
to reset the user password.
If passwordGraceLimit is less than 0 then grace limiting is disabled
and unlimited binds are allowed.
Grace login limitations only apply to entries with the objectclass
posixAccount or simplesecurityobject in order to limit this to
IPA users and system accounts.
Some basic support for the LDAP ppolicy control is enabled such that
if the ppolicy control is in the bind request then the number of
remaining grace binds will be returned with the request.
The passwordGraceUserTime attribute is reset to 0 upon a password
reset.
user-status has been extended to display the number of grace binds
which is stored centrally and not per-server.
Note that passwordGraceUserTime is an operational attribute.
https://pagure.io/freeipa/issue/1539
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
For a well-known anonymous principal an Anonymous PKINIT method is used
which ignores the password set in the principal entry. For these
principals any defined user auth type is irrelevant, their use is
defined in RFC 6112. This gets confusing when a default user auth type
requires a particular authentication method.
When AS request for Anonymous PKINIT is used, a TGT would contain no
authentication indicator. It means we cannot apply any specific
indicator policy and must skip the checks.
Fixes: https://pagure.io/freeipa/issue/9165
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Vagrant templates for `pki-`, `389ds-` and `testing-` are no longer
needed after feature added by https://github.com/freeipa/freeipa-pr-ci/pull/463.
This updates the test definitions to use PR-CI's custom arguments with
standard `ci-master-f36` vagrant box.
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
In bind 9.17.2+ all dnssec utilities were moved to /usr/bin with
commit 4419606c9d2a52536a6dd0882ac0c7068ac27f30.
Since we only use those utilities in the specialized tool, do a fixup of
the paths in the tool.
Fixes: https://pagure.io/freeipa/issue/9157
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Add the --subid option to client, server and replica installers.
This option allows to configure authselect with the sssd
profile + with-subid feature, in order to have SSSD setup as
a datasource for subid in /etc/nsswitch.conf.
The default behavior remains unchanged: without the option,
/etc/nsswitch.conf keeps the line subid: files
Fixes: https://pagure.io/freeipa/issue/9159
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This will allow custom hardened password policy to be applied.
Without this then the policy will be skipped because the UA
is not enabled.
The KDC and client will prefer SPAKE any time it is available.
For IPA this should mean we should choose hardened setting over a
default one any time SPAKE is used.
Related: https://pagure.io/freeipa/issue/9121
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
So that it prints out a valid command:
- replace "ipa-replica-manage del" by "ipa server-del" (only domain-level1 is now supported)
- the commands needs to be run on a working server, not on the host where ipa-replica-install failed
Fixes: https://pagure.io/freeipa/issue/9162
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Moving 'latest' to Fedora 36 and 'previous' to Fedora 35.
Update Rawhide template.
The following templates are updated:
- gating
- nightly_latest
- nightlñy_latest_selinux
- nightly_previous
- nightly_rawhide
- temp_commit
These are not updated as they will receive a new update after
"custom COPR" feature in PR-CI is merged:
- nightly_latest_389ds
- nightly_latest_pki
- nightly_latest_testing
Signed-off-by: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
If the client was installed with authconfig, with
automount configured to use ldap (--no-sssd), and later
updated to a version using authselect, the uninstaller
tries to disable the authselect feature with-custom-automount
but fails because there is no authselect profile in use.
(Upgrade of a client does not transform authconfig settings
into authselect settings because we don't have any client
upgrader, as opposed to the ipa-server-upgrade for the
servers).
To avoid uninstallation failure, ignore the error and log a
warning.
The second part of the commit leverages the "complete" state
stored in the statestore, in order to fix issues when
a client installation fails and the installation is reverted
by the ipa-client-install tool itself.
The fix checks if the statestore shows an incomplete
installation. If the install was incomplete and failed before
any attempt to configure authselect, then unconfigure doesn't
need to do anything. In the other cases, unconfigure needs
to revert to the pre-ipa state.
Fixes: https://pagure.io/freeipa/issue/9147
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In RHEL there is a split of packages between Base OS and AppStream
repositories. While both repositories are accessible and enabled by
default, there are different requirements towards binary packages in
both. Namely, Base OS packages cannot have runtime dependencies to
AppStream packages and they should have a stricter lifecycle promises in
terms of API and ABI stability.
SSSD 2.7.0 adds sssd-idp package which provides actual implementation of
OAuth 2.0 integration. Since SSSD is provided as part of Base OS, if
sssd-idp is placed there, then all its dependencies would have to be in
Base OS. Unfortunately, libjose is already part of AppStream.
SSSD team currently pulls sssd-idp as a dependency of sssd-ipa so
FreeIPA didn't need to change anything. However, Base OS requirements
will force SSSD team to drop sssd-idp dependency from sssd-ipa. This
means FreeIPA will have to explicitly depend on sssd-idp.
Fixes:https://pagure.io/freeipa/issue/9155
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
When IdP configuration is provided, take it into account:
- idp-specific Kerberos ticket policy would be applied
- Presence of IdP link in a Kerberos principal entry would cause KDB to
enable `idp` pre-authentication method on KDC side.
The latter requires additional pre-authentication method supplied with
SSSD 2.7.0.
Fixes: https://pagure.io/freeipa/issue/8804
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Signed-off-by: Pavel Březina <pbrezina@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Sumit Bose <sbose@redhat.com>
SSSD OIDC helper is used for negotiating with OAUTH2 or OIDC end points
of external identity providers (IdPs).
ipa-otpd daemon now is capable to take either Issuer URL or individual
endpoints and call SSSD OIDC helper accordingly.
Communication with SSSD OIDC helper can be debugged with the use of a
debug variable set in /etc/ipa/default.conf. Man page for
default.conf(5) has been updated to provide this information.
Fixes: https://pagure.io/freeipa/issue/8805
Signed-off-by: Sumit Bose <sbose@redhat.com>
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Sumit Bose <sbose@redhat.com>
RTD default theme removes discs from the section list items which makes
design pages look strange. Add them back via small CSS override.
Also, add 1em on the left side of the disc to provide visual cue that
this is a list item.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Sumit Bose <sbose@redhat.com>
The test test_ipahealthcheck.py::TestIpaHealthCheckWithoutDNS
is installing the server without DNS but calls the installer
with --no-dnssec-validation option.
Remove the --no-dnssec-validation option as it is incompatible
with a non-DNS setup.
Fixes: https://pagure.io/freeipa/issue/9152
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
healthcheck related tests are failing because of the below issue
"client @0x7f8ee47c4d48 : servfail cache hit (CD=0)"
and as a result healthcheck related packages are not downloaded on test
system.
Hence adding the --no-dnssec-validation option to install_master
and install_replica function
https://pagure.io/freeipa/issue/9151
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Allow npm to install and use latest grunt that is compatible with newer
OpenSSL.
This, in turn, requires ATK interfaces to be present for the chromium
installed by puppeteer.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Newer grunt will pull a PhantomJS that is compatible with newer OpenSSL
so the workaround is not needed anymore.
Additionally, OpenSSL 3.0 is more strict and does not tolerate
non-existing default configuration file.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Use fedora-toolbox:36 image as it is prepared to work with systemd and
sudo
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
subid range management does not allow to delete ranges
If subid range was allocated, it cannot be removed because there might
be file objects associated with it on one of IPA clients.
In Web UI a button to remove the range should not be shown.
Remove corresponding test from the Web UI test for subid as the button
to remove the subid range is not present anymore.
Fixes: https://pagure.io/freeipa/issue/9150
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
The test is calling getcert rekey -G DSA in order to rekey
a certificate with a DSA key, but DSA support has been disabled
in the default crypto policy, and certmonger does not support it
any more (see the BZ
https://bugzilla.redhat.com/show_bug.cgi?id=2066439)
Remove the test as it's not relevant anymore. The rekey
operation is tested anyway in other tests:
- test_certmonger_rekey_keysize
- test_rekey_keytype_RSA
- test_rekey_request_id
Fixes: https://pagure.io/freeipa/issue/9140
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
ReadTheDocs somehow caches requirements and insists in using old
version of Sphinx (1.8). We have to force using newer one (4.5)
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Documentation generator can be run inside Python virtual environment.
This allows to isolate from the system-wide changes and add Sphinx
extensions that aren't packaged in a distribution.
The only exception right now is plantuml package. We rely on plantuml to
generate diagrams and since it is written in Java, it cannot be
installed directly into the Python venv through 'pip' tool.
Fixes: https://pagure.io/freeipa/issue/9148
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
m2r project was forked to m2r2 which is actively developed.
m2r2 works with new Sphinx versions.
Update our list of documentation requirements and add support for
plantuml to be able to integrate diagrams.
Fixes: https://pagure.io/freeipa/issue/9148
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
The file epn.conf has been updated when fixing issue 9145
and the test test_epn.py::TestEPN::test_EPN_config_file
is comparing its sha256sum with the checksum of the
shipped file from the package ipa-client-epn.
The expected checksum needs to be updated.
Fixes: https://pagure.io/freeipa/issue/9146
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In /etc/ipa/epn.conf it is possible to customize the
e-mail subject by setting msg_subject=<value> but this
setting is not documented in the man page.
Add the options in epn.conf man page and in the template.
Fixes: https://pagure.io/freeipa/issue/9145
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The tests
TestNonPosixAutoPrivateGroup::test_idoverride_with_auto_private_group
and
TestPosixAutoPrivateGroup::test_gidnumber_not_corresponding_existing_group
are expected to fail until SSSD fixes issues 5988 and 5989.
They currently define an xfail_context with a condition based on
sssd version but that condition is wrong (as of today, no version
of sssd provides the fix).
Remove the wrong condition so that the test is always expected to fail.
Fixes: https://pagure.io/freeipa/issue/9141
Reviewed-By: Anuja More <amore@redhat.com>
In many cases it is impossible to investigate test failures of
environments where a trust to Active Directory is establishe without
Samba logs.
Collect Samba logs by default and make sure Samba is configured with
higher log levels if we are going to configure IPA to setup trust to
Active Directory.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
When NRPC netr_ServerAuthenticate3 call is performed, a trusted AD DC
would use trusted domain account to authenticate to Samba. This means
that Samba would do internally samr_QueryUserInfo2 request with level 16
(UserControlInformation), coming to PDB module via pdb_getsampwsid()
call.
For normal user or workstation accounts we expect to have Kerberos keys
available and may be able to extract NTLM hash data from them. However,
trusted domain account is not a normal Kebreros principal. It stores TDO
credential in a different way. Since we never processed it through the
pdb_getsampwsid() call, it was not possible to retrieve the NTLM hash
for TDO account at all, hence netr_ServerAuthenticate3 call was failing.
NTLM hash is used internally in Samba. An external communication with AD
DC will use an AES-based session key that is derived from the TDO
credential. The credential itself can be treated as a plaintext here.
Fix it by adding a recognition of the trusted domain object account and
retrieve the NTLM hash from the correct attribute of the TDO.
Fixes: https://pagure.io/freeipa/issue/9134
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This is a problem since we added commit b5fbbd1 in 2019. Its logic
allowed to add RC4-HMAC keys for cifs/.. service principal but it didn't
account for the case when cifs/.. principal initiates the request.
Since ipasam only uses GETKEYTAB control, provide this extension only
here and don't allow the same for SETKEYTAB. At the point of check for
the bind DN, we already have verified that the DN is allowed to write to
the krbPrincipalKey attribute so there is no leap of faith to 'any
cifs/... principal' here.
A principal must be member of cn=adtrust
agents,cn=sysaccounts,cn=etc,$SUFFIX to allow perform this operation
Fixes: https://pagure.io/freeipa/issue/9134
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
topologysegment name can be different depending on suffix.
This patch determines and supply the name of topologysgement
as per the suffix
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
There is no guarantee for the topologysegement name, it could be
master-to-replica or replica-to-master. If it is master-to-replica
then --right should be used with the command else --left.
Fixes: https://pagure.io/freeipa/issue/9137
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Continuation of the commit 2eee5931d7:
Use AD-SUPPORT subpolicy when testing trust to Active Directory in FIPS
mode. This is required in FIPS mode due to AD not supporting Kerberos
AES-bases encryption types using FIPS-compliant PBKDF2 and KDF, as
defined in RFC 8009.
Fixes: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
This test checks that when
'pki.server.healthcheck.clones.connectivity_and_data' check is run
'Source 'pki.server.healthcheck.clones.connectivity_and_data' not found'
is not displayed.
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2041995
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The test is calling run_command with a list of arguments:
run_command(['rm', '-f', paths.CERTMONGER_REQUESTS_DIR + '/*'])
but this format does not support shell expansion.
Replace with a str parameter:
run_command('rm -fv' + paths.CERTMONGER_REQUESTS_DIR + '/*')
to make sure all the files in the directory are actually removed.
Fixes: https://pagure.io/freeipa/issue/8506
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
test_ipahealthcheck_expiring is moving the date in the future
in order to check that certmonger properly warns about expiring
certificates, then uninstalls the master.
The uninstallation randomly fails with a DBus error communicating
with certmonger because of a contention between certmonger being
waken up by the call to stop tracking certs and the certmonger
helpers trying to renew the certs.
The test is stopping PKI server, then moves the date in the future.
At this point, certmonger is still running (we are testing that
getcert list properly warns about near expiration). This means that
chances are high that certmonger has enough time to launch the CA helper
for renewal, that takes the lock. But since PKI is down, the helper
remains running for a while and does not release the lock. Then
certmonger is stopped, the tracking files are removed, certmonger is
restarted.
To avoid the contention, manually remove the tracking before
calling uninstall and remove the renewal lock file.
Fixes: https://pagure.io/freeipa/issue/9123
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This makes automount configurable only using sssd and not LDAP.
The reason is that authselect 1.3 no longer supports
user-nsswitch.conf which is where we made direct changes to the
nss configuration on Fedora/RHEL.
The equivalent option was removed from ipa-client-install in
https://pagure.io/freeipa/issue/7671
Fixes: https://pagure.io/freeipa/issue/9084
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Python 3.10 deprecated ssl.PROTOCOL_TLS and ssl.PROTOCOL_SSLv23
constants which were aliases to each other. Use of them now causes a
warning to be displayed:
/usr/lib/python3.10/site-packages/ipalib/util.py:347: DeprecationWarning: ssl.PROTOCOL_TLS is deprecated
ctx = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
Use ssl.PROTOCOL_TLS_CLIENT instead, this constant is available since
Python 3.6.
Fixes: https://pagure.io/freeipa/issue/9129
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When an entry is loaded the incoming values are converted
into python datatypes automatically based on the _SYNTAX_MAPPING
value in ipaldap.
When using delattr to remove a mapped value it will fail because
the datatypes do not match up. For example date types are
datetime.datetime structions and won't match a generalized time
string.
So try to map the value to delete using _SYNTAX_MAPPING before
trying to remove the value. Fall back to trying to remove the
raw value if the mapping fails.
This won't work for some mapping types, DNs for example. Providing
only the RDN value for a DN-type, manager for example, lacks the
context to know how to construct the DN (RDN and contaner).
Fixes: https://pagure.io/freeipa/issue/9004
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
When minlife > maxlife specified on commandline, it says:
"ipa: ERROR: invalid 'maxlife': Maximum password life must be
greater than minimum."
But when minlife == maxlife specfied, It works.
This test check that error message says what exactly it does
related: https://pagure.io/freeipa/issue/9038
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The vault plugin has used TripleDES (des-ede3-cbc) as default wrapping
algorithm since the plugin was introduced. Allow use of AES-128-CBC as
alternative wrapping algorithm for transport of secrets.
Fixes: https://pagure.io/freeipa/issue/6524
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
SPAKE is based on the crypto primitives which are not FIPS compliant
yet. This means that in FIPS mode use of 'hardened' authentication
indicator is not possible. Skip corresponding tests in FIPS mode.
Related: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
paramiko cannot be used in FIPS mode. We have few tests that import
generic methods from test_otp (add_token/del_token) and those tests fail
in FIPS mode due to unconditional 'import paramiko'.
Instead, move 'import paramiko' to the ssh_2f() helper which is not used
in FIPS mode (the whole SSH 2FA test is skipped then).
Related: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
KDC configuration in /var/kerberos/krb5kdc/kdc.conf is generated from
the template in install/share/kdc.conf.template. Master key encryption
type specified there is used to bootstrap the master key in LDAP
database. Once it is done, actual deployment does not rely on the
master_key_type value anymore. The actual master key(s) get loaded from
LDAP database where they stored in a BER-encoded format, preserving all
parameters, including encryption type.
This means we can safely migrate to AES256-SHA2 as the default master
key encryption type for new installations. Replicas will get their
master key encryption type details from the server they were provisioned
from.
MIT Kerberos supports AES256-SHA2 since 1.15 (2015), meaning RHEL 7.4 is
the earliest supported version as it provides krb5 1.15.1. Current
supported RHEL 7 version is RHEL 7.9. Since RHEL 6 already cannot be
used as a replica to IPA 4.5+ due to a domain level 1 upgrade, this
change does not affect old releases.
Migration from the previously deployed master key encryption type is
described by MIT Kerberos upstream in
http://web.mit.edu/kerberos/krb5-latest/doc/admin/advanced/retiring-des.html#the-database-master-key
One would need to use '-x ipa-setup-override-restrictions' to allow
the `kdb5_util` utility to modify the data over IPA KDB driver.
Fixes: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
New calls, getorigbyusername() and getorigbygroupname(), are added to
libsss_nss_idmap. They allow to query the AD specific attributes for a
user or a group directly. Besides a minor performance benefit it helps
to avoid issues if there are users and groups with the same name and the
group is not a user-private group but a real group with members.
Fixes: https://pagure.io/freeipa/issue/9127
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The subCA keys are loaded out-of-band after creation into the
CA so they may have been replicated but not loaded. Give more
time for them to appear in the remote CA.
Use a loop for the checking instead of a raw sleep because most
of the time this is very fast (< 15 seconds) but sometimes it
requires just a bit more. Allow up to 60 seconds.
To avoid output difference, strip the token name out of certutil
output. We don't care about the token a certificate is stored
in, the internal or the FIPS token. We just care that they exist
on both servers and that the keys match.
Apparently in some cases the token name is displayed and not in
others so lets normalize the output to make comparisons more
consistent.
Fixes: https://pagure.io/freeipa/issue/9096
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
In some contexts, filesystem mounts may be owned by unmapped users
(e.g. `emptyDir` mounts in Kubernetes / OpenShift when using user
namespaces). This causes `systemd-tmpfiles(8)` to fail, as a
consequence of systemd's path processing routines which reject this
scenario. Therefore, in Fedora container context, if the
`IPA_TMPFILES_PROG` environment value is set, use the program
specified by its value instead of `/bin/systemd-tmpfiles`.
Signed-off-by: Fraser Tweedale <ftweedal@redhat.com>
Fixes: https://pagure.io/freeipa/issue/9126
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint doesn't handle flow control and thus, doesn't understand
that a key of type `str` is not reachable at this point:
> ipalib/base.py:472: [E1126(invalid-sequence-index),
NameSpace.__getitem__] Sequence index is not an int, slice, or instance
with __index__)
Note: I faced this error on Python3.9 and didn't see it using
Python3.10.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Found by new Pylint:
> ipaclient/install/client.py:1926:
[W1310(format-string-without-interpolation), get_ca_certs] Using
formatting for a string that does not have any interpolated variables)
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
As of Python3 `currentThread`, `thread.getName` are aliases for
`threading.current_thread()` and `threading.Thread.name`
respectively.
In Python3.10:
> bpo-43723: The following threading methods are now deprecated and
should be replaced:
currentThread => threading.current_thread()
activeCount => threading.active_count()
Condition.notifyAll => threading.Condition.notify_all()
Event.isSet => threading.Event.is_set()
Thread.setName => threading.Thread.name
thread.getName => threading.Thread.name
Thread.isDaemon => threading.Thread.daemon
Thread.setDaemon => threading.Thread.daemon
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.11.0 extends consider-using-in check to work for
attribute access.
> To check if a variable is equal to one of many values,combine the
values into a tuple and check if the variable is contained "in" it
instead of checking for equality against each of the values.This
is faster and less verbose.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.9.0 introduced new checker which was a subset of
arguments-differ:
> Used when a method parameter has a different name than in the
implemented interface or in an overridden method.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.12.0 introduced new checker:
> Used when Pylint detects that collection literal comparison is being
used to check for emptiness; Use implicit booleaness insteadof a
collection classes; empty collections are considered as false
Comparison of variable to equality to collection:
> Lexicographical comparison between built-in collections works as follows:
For two collections to compare equal, they must be of the same type,
have the same length, and each pair of corresponding elements must
compare equal (for example, [1,2] == (1,2) is false because the type is
not the same).
Collections that support order comparison are ordered the same as their
first unequal elements (for example, [1,2,x] <= [1,2,y] has the same
value as x <= y). If a corresponding element does not exist, the shorter
collection is ordered first (for example, [1,2] < [1,2,3] is true).
So, `assert value == {}` is not the same as `assert not value`.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.9 introduced new check:
> New checker consider-using-dict-items. Emitted when iterating over
dictionary keys and then indexing the same dictionary with the key
within loop body.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.9 introduced new check:
> Emitted when iterating over the dictionary items (key-item pairs) and
accessing the value by index lookup. The value can be accessed directly
instead.
Note: in Python3 removing from dict during an iteration is not
possible even. For example,
```
cat a.py
d = {"a": 1}
for k, v in d.items():
if v is not None:
del d[k]
python3 a.py
Traceback (most recent call last):
File "/usr/src/RPM/BUILD/freeipa/a.py", line 3, in <module>
for k, v in d.items():
RuntimeError: dictionary changed size during iteration
```
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
There is no actual usage of deprecated classes for Python3.
Pylint complains about such for Python2. Since Python2 is no
longer supported these imports were removed.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
__convert_iter was added in 24b6cb89d, but it was never used.
Found by Pylint:
```
ipalib/frontend.py:696: [W0238(unused-private-member),
Command.__convert_iter] Unused private member
`Command.__convert_iter(self, kw)`)
```
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
> This mangling is done without regard to the syntactic position of the
identifier, as long as it occurs within the definition of a class.
`__set_attr` is called for instance of the class within its
classmethod.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The private member `__finalized` has been added in
7db3aae1b2, later removed in
6b8abb0d78, but `_API__finalized`
(access via mangled attribute name) was not cleaned up and finally
refactored back to `__finalized` in
b1fc875c3a.
Found by Pylint:
```
ipalib/plugable.py:807: [W0238(unused-private-member), API.finalize]
Unused private member `API.__finalized`)
```
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
__remove_lightweight_ca_key_retrieval_custodia has been added in
8700101d9, but it was never used.
Caught by Pylint:
```
ipaserver/install/cainstance.py:1308: [W0238(unused-private-member),
CAInstance.__remove_lightweight_ca_key_retrieval_custodia]
Unused private member
`CAInstance.__remove_lightweight_ca_key_retrieval_custodia(self)`)
```
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
__convert_to_gssapi_replication has been added in a0bfbec19 and
then removed in ce2bb47cc without clean up.
Found by Pylint:
```
ipaserver/install/krbinstance.py:589: [W0238(unused-private-member),
KrbInstance.__convert_to_gssapi_replication] Unused private member
`KrbInstance.__convert_to_gssapi_replication(self)`)
```
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.9.0 new checker:
> Emitted when accessing only the first or last element of str.split().
The first and last element can be accessed by using str.split(sep,
maxsplit=1)[0] or str.rsplit(sep, maxsplit=1)[-1] instead.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.10 introduced new checker:
> It is better to specify an encoding when opening documents. Using the
system default implicitly can create problems on other operating
systems. See https://www.python.org/dev/peps/pep-0597/
According to that PEP:
> open(filename) isn't explicit about which encoding is expected:
- If ASCII is assumed, this isn't a bug, but may result in decreased
performance on Windows, particularly with non-Latin-1 locale
encodings
- If UTF-8 is assumed, this may be a bug or a platform-specific script
- If the locale encoding is assumed, the behavior is as expected (but
could change if future versions of Python modify the default)
IPA requires UTF-8 environments.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.10 introduced new checkers:
> Emitted when using dict() to create an empty dictionary instead of the
literal {}. The literal is faster as it avoids an additional function
call.
> Emitted when using list() to create an empty list instead of the
literal []. The literal is faster as it avoids an additional function
call.
Too many unessential changes.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.11 introduced new checker:
> Used when we detect a string that is being formatted with format() or
% which could potentially be a f-string. The use of f-strings is
preferred. Requires Python 3.6 and ``py-version >= 3.6``.
- f-strings are not mandatory
- format can be more readable
- there are ~5.5K spotted issues
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Pylint 2.10 introduced new checker `redundant-u-string-prefix`:
> Used when we detect a string with a u prefix. These prefixes were
necessary in Python 2 to indicate a string was Unicode, but since Python
3.0 strings are Unicode by default.
There are ~31K emitted warnings right now. They can be fixed on
refactorings without any rush.
Fixes: https://pagure.io/freeipa/issue/9117
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Use AD-SUPPORT subpolicy when testing trust to Active Directory in FIPS
mode. This is required in FIPS mode due to AD not supporting Kerberos
AES-bases encryption types using FIPS-compliant PBKDF2 and KDF, as
defined in RFC 8009.
Fixes: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Julien Rische <jrische@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
RHEL 9 system-wide crypto policies aim at eventual removal of SHA-1 use.
Due to bootstrapping process, force explicitly supported encryption
types in kdc.conf or we may end up with AES128-SHA1 and AES256-SHA2 only
in FIPS mode at bootstrap time which then fails to initialize kadmin
principals requiring use of AES256-SHA2 and AES128-SHA2.
Camellia ciphers must be filtered out in FIPS mode, we do that already
in the kerberos.ldif.
At this point we are not changing the master key encryption type to
AES256-SHA2 because upgrading existing deployments is complicated and
at the time when a replica configuration is deployed, we don't know what
is the encryption type of the master key of the original server as well.
Fixes: https://pagure.io/freeipa/issue/9119
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Julien Rische <jrische@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
The test
test_integration/test_installation.py::TestInstallWithoutNamed
was missing in some nightly definitions.
Add the job definition for:
- nightly_latest_selinux.yaml
- nightly_latest_testing.yaml
- nightly_latest_testing_selinux.yaml
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The example text included a trailing dot which isn't actually
allowed in a system hostname (just DNS). Remove the suggestion
to include it and strip off any trailing dot so that the install
can proceed.
Related: https://pagure.io/freeipa/issue/9111
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The refactor change 9094dfc had a slight error where the
user-input provided value in input wasn't being validated. Only
the command-line or the current FQDN was being verified so
if the FQDN was bad any value input by the user was being skipped.
Fixes: https://pagure.io/freeipa/issue/9111
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
There is no use-case to keep a restored server in a hidden
state. It can be re-marked as hidden once the installation is
recovered from the restore. So mark all restored services as
enabled so they are visible to existing clients during the
remaining recovery.
Fixes: https://pagure.io/freeipa/issue/9095
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The IPA tools will create /var/log/ipaupgrade.log with mode
0600. If for some reason this file doesn't exist during
upgrade then it will be created by the RPM transaction with
mode 0644 (because of umask).
So always set the mode once the snippets are done. This
will ensure that a newly created log will have the expected
mode and also fix any previous incorrectly set mode.
Fixes: https://pagure.io/freeipa/issue/8899
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Turn off string interpolation on the FileStore class to avoid
exceptions when a value to be saved contains a percent sign (%).
The underlying SafeConfigParser that is used interprets percent
signs as placeholders to be interpolated which then causes an
exception as the placeholder isn't properly formatted.
ipa-client-install uses the FileStore class to backup certain
values that it overwrites as part of the installation. If those
pre-existing, backed-up values contained a percent sign,
ipa-client-install would throw an exception and thus prevent
installation.
https://pagure.io/freeipa/issue/9085
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
ipa_cldap_encode_netlogon() allocates memory to store binary data as part of
berval (bv_val) when processing a CLDAP packet request from a worker. The
data is used by ipa_cldap_respond() but bv_val is not freed later on.
This commit is adding the corresponding free() after ipa_cldap_respond()
is completed.
Discovered by LeakSanitizer
Fixes: https://pagure.io/freeipa/issue/9110
Signed-off-by: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Thierry Bordaz <tbordaz@redhat.com>
There is some contention between certmonger starting during the
uninstallation process in order to stop the tracking and activity
going on within certmonger helpers.
As near as I can tell certmonger is not running, then IPA is
stopped in order to uninstall, then certmonger is started to stop
the tracking. certmonger checks cert status on startup but since
IPA isn't running it can't get a host ticket. During this time any
request over DBus may time out, causing a test to fail when we're
just trying to clean up.
https://pagure.io/freeipa/issue/8506
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The quiet option was supposed to suppress unnecessary output
on successful invocations. Instead it was suppressing most
error messages as well.
Remove all the quiet checks when an error would be displayed.
Pass the quiet option to ipa-getkeytab so it honors the
request as well.
Also tidy up some of the debug output.
https://pagure.io/freeipa/issue/9105
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The timer was only being enabled during package installation
if IPA was configured. So effectively only on upgrade.
Add as a separate installation step after the ccache directory
is configured.
Fixes: https://pagure.io/freeipa/issue/9107
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
A change to replace -h and -p options in OpenLDAP command line utilities
causes also an update in the help text in sudorule plugin. This, sadly,
makes existing translations of that text not valid anymore. However, we
have to change the text as OpenLDAP 2.6+ will make the command
referenced in the help text incorrect.
The change in OpenLDAP 2.6+ implements deprecation that was announced by
OpenLDAP project around 20 years ago, so all existing tools support -H
option.
See also: https://bugs.openldap.org/show_bug.cgi?id=8618
Related: https://pagure.io/freeipa/issue/9106
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
pylint 2.9 thinks that __add_principal is a class-level method that is
unused. It is a local function inside one of class methods and is used
directly inside that method.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
OpenLDAP 2.6+ finally deprecated -h and -p options in all its command
line tools. They are not allowed anymore and cause ldap* tools to stop
hard with 'unknown option' error.
Fix this by always using -H url option instead. Deriving default value
for -H url from the configuration file still works, it is only -h and -p
that were deprecated.
See also: https://bugs.openldap.org/show_bug.cgi?id=8618
Fixes: https://pagure.io/freeipa/issue/9106
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
I can only guess to the original purpose of this override. I
believe it was because this is called in the installer prior
to Apache being set up. The expectation was that this would
only be called locally. It predates the RestClient class.
RestClient will attempt to find an available service. In this
case, during a CA installation, the local server is not
considered available because it lacks an entry in
cn=masters. So it will never be returned as an option.
So by overriding the port to 8443 the remote connection will
likely fail because we don't require that the port be open.
So instead, instantiate a RestClient and see what happens.
There are several use-cases:
1. Installing an initial server. The RestClient connection
should fail, so we will fall back to the override port and
use the local server. If Apache happens to be running with
a globally-issued certificate then the RestClient will
succeed. In this case if the connected host and the local
hostname are the same, override in that case as well.
2. Installing as a replica. In this case the local server should
be ignored in all cases and a remote CA will be picked with
no override done.
3. Switching from CA-less to CA-ful. The web server will be
trusted but the RestClient login will fail with a 404. Fall
back to the override port in this case.
The motivation for this is trying to install an EL 8.x replica
against an EL 7.9 server. 8.5+ includes the ACME service and
a new profile is needed which doesn't exist in 7. This was
failing because the RestClient determined that the local server
wasn't running a CA so tried the remote one (7.9) on the override
port 8443. Since this port isn't open: failure.
Chances are that adding the profile is still going to fail
because again, 7.9 lacks ACME capabilities, but it will fail in
a way that allows the installation to continue.
I suspect that all of the overrides can similarly handled, or
handled directly within the RestClient class, but for the sake
of "do no harm" I'm only changing this instance for now.
https://pagure.io/freeipa/issue/9100
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Since 5d51ae5, principal keys were dropped in case user auth indicator
was not including password. Thereafter, the key removal behavior was
removed by 15ff9c8 in the context of the kdcpolicy plugin introduction.
Support for hardened pre-auth methods (FAST and SPAKE) was added in
d057040, and the removal of principal keys was restored afterwards by
f0d12b7, but not taking the new hardened auth indicator into account.
Fixes: https://pagure.io/freeipa/issue/9065
Related to: https://pagure.io/freeipa/issue/8001
Signed-off-by: Julien Rische <jrische@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francisco Trivino <ftrivino@redhat.com>
The access to /kra/admin/kra/getStatus will be needed
in order to fix pki-healthcheck.
Note that this commit is a pre-requisite for the fix
to be done on PKI side. No test added since the full
integration test already exists in test_replica_promotion.py,
in TestHiddenReplicaPromotion::test_ipahealthcheck_hidden_replica
Fixes: https://pagure.io/freeipa/issue/9099
Related: https://pagure.io/freeipa/issue/8582
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The Custodia key export handler is using the default's OpenSSL encryption
scheme for PKCS#12.
This represents an issue when performing a migration from CentOS Stream 8 (C8S)
to CentOS Steam 9 (C9S) where the Custodia client running in the new C9S
replica talks to the Custodia server on C8S source server. The later creates an
encrypted PKCS#12 file that contains the cert and the key using the OpenSSL's
default encryption scheme, which is no longer supported on C9S.
This commit enforces a stronger encryption algorigthm by adding following
arguments to the Custodia server handler:
-keypbe AES-256-CBC -certpbe AES-256-CBC -macalg sha384
The new arguments enforce stronger PBEv2 instead of the insecure PBEv1.
Fixes: https://pagure.io/freeipa/issue/9101
Signed-off-by: Francisco Trivino <ftrivino@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The test TestIpaClientAutomountFileRestore expects a
specific order for the automount sources to query
in /etc/nsswitch.conf.
With authselect update 1.3.0, the databases are sorted in
order of likelihood and the following line in seen:
automount: files sss
instead of
automount: sss files
Since the test doesn't care about the order but rather about
the list of sources, ignore the order.
Fixes: https://pagure.io/freeipa/issue/9067
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
MIT Kerberos 1.20 changes DAL interface around PAC record issuance:
sign_authdata callback is removed and replaced with issue_pac one.
The signatures are different and logic changed as well.
Prepare for KDB version 9 by moving PAC implementation into separate
source files. ipa_kdb_mspac.c is left with most of the common code.
FreeIPA supports sign_authdata callback since KDB version 6, move current
implementation to ipa_kdb_mspac_v6.c.
KDB version 8 actually changed sign_authdata interface and we accounted
to that in ipa_kdb.c with a stub that re-uses v6 version. Keep it as it
is right now.
Finally, add KDB version 9 stub files. Compiling against MIT Kerberos
1.20 does not work yet, thus explicit #error message in ipa_kdb.c. This
will be worked on later.
Related: https://pagure.io/freeipa/issue/9083
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
I got the following message when installing a replica with CA:
2021-11-22T21:15:35Z DEBUG [5/30]: configuring certificate server instance
...
WARNING: The 'pki_ssl_server_token' in [CA] has been deprecated. Use 'pki_sslserver_token' instead.
Installation log: /var/log/pki/pki-ca-spawn.20211122221535.log
Installing CA into /var/lib/pki/pki-tomcat.
With the following change the message no longer appears when installing a replica.
This commit fixes the firt (and simple) part of https://pagure.io/freeipa/issue/9056
Signed-off-by: Jochen Kellner <jochen@jochen.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
FullLoader class for YAML loader was introduced in version 5.1 which
also deprecated default loader. SafeLoader, however, stays consistent
across the versions and brings added security.
This fix is necessary as PyYAML > 5.1 is not available in downstream.
Related: https://pagure.io/freeipa/issue/9009
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Add a test to test_config_plugin, that calls ipa config-mod
with the same value as already present in LDAP.
The call must return EmptyModlist.
Related: https://pagure.io/freeipa/issue/9063
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When ipa config-mod is called with the option --enable-sid,
the code needs to trap EmptyModlist exception (it is expected
that no LDAP attribute is modified by this operation).
The code had a flaw and was checking:
'enable_sid' in options
instead of
options['enable_sid']
"'enable_sid' in options" always returns true as this option
is a Flag with a default value, hence always present even if
not specified on the command line.
Fixes: https://pagure.io/freeipa/issue/9063
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The automember-default-group commands inherit from
the automember commands but should not provide the
--desc parameter.
Remove 'description' from the list of parameters.
Fixes: https://pagure.io/freeipa/issue/9068
Reviewed-By: Thomas Woerner <twoerner@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Following test cases would be checked:
- when called with --principal (it should then prompt for a password)
- when called with --principal / --password
- when called without principal and password but with a kerberos TGT,
kinit admin done before calling ipa-replica-conncheck
- when called without principal and password, and without any kerberos
TGT (it should default to principal=admin and prompt for a password)
related: https://pagure.io/freeipa/issue/9047
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
OpenLDAP 2.6 deprecated separate libldap/libldap_r, there is only one
(reentrant) variant for the library.
Attempt to use _r variant by default. In case it is missing, assume we
are using OpenLDAP 2.6 which has libraries without _r suffix. The
functions are still reentrant so there is not functional difference.
Fixes: https://pagure.io/freeipa/issue/9080
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The test sets 389-ds nsslapd-idletimeout to 60s, then does a
kinit with an otp token (which makes ipa-otpd create a LDAP
connection), then sleeps for 60s. The expectation is that
ns-slapd will detect that the LDAP conn from ipa-otpd is idle
and close the connection.
According to 389ds doc, the idle timeout is enforced when the
connection table is walked. By doing a ldapsearch, the test
"wakes up" ns-slapd and forces the detection of ipa-otpd
idle connection.
Fixes: https://pagure.io/freeipa/issue/9044
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
New versions of MS-KILE and MS-SFU after Windows Server November 2021
security updates add PAC_REQUESTER_SID buffer check behavior:
- PAC_REQUESTER_SID should only be added for TGT requests
- if PAC_REQUESTER_SID is present, KDC must verify that the cname on
the ticket resolves to the account with the same SID as the
PAC_REQUESTER_SID. If it doesn't KDC must respond with
KDC_ERR_TKT_REVOKED
Change requester SID check to skip exact check for non-local
PAC_REQUESTER_SID but harden to ensure it comes from the trusted domains
we know about.
If requester SID is the same as in PAC, we already do cname vs PAC SID
verification.
With these changes FreeIPA works against Windows Server 2019 with
November 2021 security fixes in cross-realm S4U2Self operations.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
MS-KILE 3.3.5.6.4.8 in revision after Windows Server November 2021
security fixes added the following requirement:
- PAC_REQUESTER_SID is only added in TGT case (including referrals and
tickets to RODCs)
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Fixture `expire_certs` moves date back after renewing the certs.
This is causing the ipa-replica to fail. This fix first uninstalls
the server then moves back the date.
Fixes: https://pagure.io/freeipa/issue/9052
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The OpenDNSSec integration code requires:
/usr/sbin/dnssec-keyfromlabel-pkcs11
which is provided by bind-pkcs11-utils, but that package is
only available on RHEL<9.
With this change, freeipa-server-dns depends on bind-dnssec-utils
on all Fedora releases and RHEL==9+, and uses:
/usr/sbin/dnssec-keyfromlabel -E pkcs11
instead of dnssec-keyfromlabel-pkcs11.
Fixes: https://pagure.io/freeipa/issue/9026
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Timo Aaltonen <tjaalton@debian.org>
Reviewed-By: Antonio Torres Moríñigo <atorresm@protonmail.com>
Due to a bug in GCC 9 and GCC 10 optimizing code, all C library
comparison functions should be used with explicit result comparison in
the code to avoid problems described in
http://r6.ca/blog/20200929T023701Z.htmlhttps://gcc.gnu.org/bugzilla/show_bug.cgi?id=95189
The code below is affected:
```
if (strcmp(a, b) || !strcmp(c, d)) ...
```
while the code below is not affected:
```
if (strcmp(a, b) != 0 || strcmp(c, d)) == 0
```
for all C library cmp functions and related:
- strcmp(), strncmp()
- strcasecmp(), strncasecmp()
- stricmp(), strnicmp()
- memcmp()
This PR idea is based on the pull request by 'Nicolas Williams <nico@twosigma.com>'
to Heimdal Kerberos: https://github.com/heimdal/heimdal/pull/855
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When running IPA in locale de_DE.UTF-8 I got an internal error:
jochen@freeipa1:~$ ipa server-del freeipa4.example.org
Removing freeipa4.example.org from replication topology, please wait...
ipa: ERROR: Ein interner Fehler ist aufgetreten
This is not the complete messages. Using en_US.UTF-8 would be ok.
In the httpd error_log:
] ipa: ERROR: non-public: TypeError: unhashable type: 'Gettext'
] Traceback (most recent call last):
] File "/usr/lib/python3.10/site-packags/ipaserver/rpcserver.py", line 407, in wsgi_execute
] result = command(*args, **options)
] File "/usr/lib/python3.10/site-packages/ipalib/frontend.py", line 471, in __call__
] return self.__do_call(*args, **options)
] File "/usr/lib/python3.10/site-packages/ipalib/frontend.py", line 499, in __do_call
] ret = self.run(*args, **options)
] File "/usr/lib/python3.10/site-packages/ipalib/frontend.py", line 821, in run
] return self.execute(*args, **options)
] File "/usr/lib/python3.10/site-packages/ipaserver/plugins/baseldap.py", line 1686, in execute] return self.execute(*args, **options)
] File "/usr/lib/python3.10/site-packages/ipaserver/plugins/baseldap.py", line 1686, in execute
] delete_entry(pkey)
] File "/usr/lib/python3.10/site-packages/ipaserver/plugins/baseldap.py", line 1637, in delete_entry
] dn = callback(self, ldap, dn, *nkeys, **options)
] File "/usr/lib/python3.10/site-packages/ipaserver/plugins/server.py", line 755, in pre_callback
] self._ensure_last_of_role(
] File
"/usr/lib/python3.10/site-packages/ipaserver/plugins/server.py", line
520, in _ensure_last_of_role
] handler(
] File "/usr/lib/python3.10/site-packages/ipaserver/plugins/server.py", line 482, in handler
] raise errors.ServerRemovalError(reason=_(msg))
] File "/usr/lib/python3.10/site-packages/ipalib/errors.py", line 269, in __init__
] messages.process_message_arguments(self, format, message, **kw)
] File "/usr/lib/python3.10/site-packages/ipalib/messages.py", line 55, in process_message_arguments
] kw[key] = unicode(value)
] File "/usr/lib/python3.10/site-packages/ipalib/text.py", line 296, in __str__
] return unicode(self.as_unicode())
] File "/usr/lib/python3.10/site-packages/ipalib/text.py", line 293, in as_unicode
] return t.gettext(self.msg)
] File "/usr/lib64/python3.10/gettext.py", line 498, in gettext
] tmsg = self._catalog.get(message, missing)
] TypeError: unhashable type: 'Gettext'
] ipa: INFO: [jsonserver_session] admin@EXAMPLE.ORG:
server_del/1(['freeipa4.example.org'], version='2.245'): InternalError
Alexander suggested to remove _() in local handler() function in
_ensure_last_of_role():
else:
raise errors.ServerRemovalError(reason=_(msg))
Looks like all the callers give already gettext-enabled message (wrapped
with _() already).
At least for my case I now get a complete error message.
Fixes: https://pagure.io/freeipa/issue/9046
Signed-off-by: Jochen Kellner <jochen@jochen.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The "is this the last KRA" test did a role-find including the
current server. This skewed the result if the server to be
removed has a KRA installed, it would always return "not allowed"
because len(roles) == 1 and the name matched, regardless of
whether other servers also provided a KRA.
https://pagure.io/freeipa/issue/8397
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
ipa pwpolicy-mod --minlife $min --maxlife $max
accepts $max >= $min, yet the error message says:
"Maximum password life must be greater than minimum."
Change the error message so that it conveys the
actual logic.
Fixes: https://pagure.io/freeipa/issue/9038
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
KDB driver extensively uses NDR parsing and marshalling code provided by
Samba libraries. Since these libraries are internal to Samba, they often
change structures without updating SONAME. Typical changes include
adding new structures, so we should require samba-client-libs we were
built against.
There used to be %requires_eq macros in RPM but it was removed from
Fedora some time ago. We need greater than or equal version of it, thus
%ipa_requires_gt is defined in the spec file.
Related: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Previously, ipadb_check_logon_info() was called only for cross-realm
case. Now we call it for both in-realm and cross-realm cases. In case of
the S4U2Proxy, we would be passed a PAC of the original caller which
might be a principal from the trusted realm. We cannot validate that PAC
against our local client DB entry because this is the proxy entry which
is guaranteed to have different SID.
In such case, validate the SID of the domain in PAC against our realm
and any trusted doman but skip an additional check of the DB entry in
the S4U2Proxy case.
Related: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
If the SID was explicitly set for the host or service entry, honor it
when issuing PAC. For normal services and hosts we don't allocate
individual SIDs but for cifs/... principals on domain members we do as
they need to login to Samba domain controller.
Related: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
As a part of CVE-2020-25717 mitigations, Samba now assumes 'CLASSIC
PRIMARY DOMAIN CONTROLLER' server role does not support Kerberos
operations. This is the role that IPA domain controller was using for
its hybrid NT4/AD-like operation.
Instead, 'IPA PRIMARY DOMAIN CONTROLLER' server role was introduced in
Samba. Switch to this role for new installations and during the upgrade
of servers running ADTRUST role.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
As part of CVE-2020-25717 mitigations, Samba expects correct user
account flags in the PAC. This means for services and host principals we
should be using ACB_WSTRUST or ACB_SVRTRUST depending on whether they
run on IPA clients ("workstation" or "domain member") or IPA servers
("domain controller").
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
PAC_ATTRIBUTES_INFO PAC buffer allows both client and KDC to tell
whether a PAC structure was requested by the client or it was provided
by the KDC implicitly. Kerberos service then can continue processing or
deny access in case client explicitly requested to operate without PAC.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
CVE-2020-25721 mitigation: KDC must provide the new PAC_REQUESTER_SID
buffer with ObjectSID value associated with the requester's principal.
The mitigation only works if NDR library supports the PAC_REQUESTER_SID
buffer type. In case we cannot detect it at compile time, a warning will
be displayed at configure stage.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
CVE-2020-25721 mitigation: KDC must provide the new HAS_SAM_NAME_AND_SID
buffer with sAMAccountName and ObjectSID values associated with the
principal.
The mitigation only works if NDR library supports the
PAC_UPN_DNS_INFO_EX buffer type. In case we cannot detect it at compile
time, a warning will be displayed at configure stage.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When working with aliased entries, we need a reliable way to detect
whether two principals reference the same database entry. This is
important in S4U checks.
Ideally, we should be using SIDs for these checks as S4U requires PAC
record presence which cannot be issued without a SID associated with an
entry. This is true for user principals and a number of host/service
principals associated with Samba. Other service principals do not have
SIDs because we do not allocate POSIX IDs to them in FreeIPA. When PAC
is issued for these principals, they get SID of a domain computer or
domain controller depending on their placement (IPA client or IPA
server).
Since 389-ds always returns unique entry DN for the same entry, rely on
this value instead. We could have used ipaUniqueID but for Kerberos
principals created through the KDB (kadmin/kdb5_util) we don't have
ipaUniqueID in the entry.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
If the principal entry in LDAP has SID associated with it, store it to
be able to quickly assess the SID when processing PAC.
Also rename string_to_sid to IPA-specific version as it uses different
prototype than Samba version.
Fixes: https://pagure.io/freeipa/issue/9031
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Robert Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Because the sidgen plugin is a postop plugin, it is not
always triggered before the result of an ADD is returned
and the objectclasses of the user may / may not contain
ipantuserattrs.
Fix the get_user_result method to work in all the cases.
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The API schema is not checked for changes until after a TTL
is expired. A one-hour TTL was hardcoded which makes development
tedious because the only way to force a schema update is to
remember to remove files between invocations.
This adds a new environment variable, schema_ttl, to configure
the TTL returned by the server to schema() calls. This can be
set low to ensure a frequent refresh during development.
If the client is in compat mode, that is if client is working
against a server that doesn't support the schema() command,
then use the client's schema_ttl instead so that the user still
has control.
Re-check validity before writing the cache. This saves us both
a disk write and the possibility of updating the expiration
with a ttl of 0. This can happen if the fingerprint is still
valid (not expired, no language change) the schema check is
skipped so we have no server-provided ttl.
https://pagure.io/freeipa/issue/8492
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Fixture issue_and_expire_cert() kinit after moving the date to
expire certs. This fix is to rely on kinit from fixture.
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Even though specifying 'raw' format is allowed, this breaks Read The
Docs which is not able to complete the full build of documentation and,
therefore, doesn't publish an updated documentation.
Pushed under one-liner rule.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
The integration tests that check backup-reinstall-restore
scenario need to clear sssd cache before checking the uid
of the admin user. For instance:
backup: saves the original admin uid
reinstall: creates a new admin uid, potentially cached by SSSD
restore: restores the original admin uid
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
When a preserved user entry is moved to staged state, the SID
attribute must not be provided to user-stage command (the option
does not exist and the SID will be re-generated anyway).
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The standard installer now configures all the items needed
for SID generation. Add a new test with the following scenario:
- install IPA server
- create an active user
- ensure the user's entry has an attribute ipantsecurityidentifier
- ensure that the kerberos ticket for the user contains PAC data
by using the utility ipa-print-pac
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Add new options to ipa config-mod, allowing to enable
SID generation on upgraded servers:
ipa config-mod --enable-sid --add-sids --netbios-name NAME
The new option uses Dbus to launch an oddjob command,
org.freeipa.server.config-enable-sid
that runs the installation steps related to SID generation.
--add-sids is optional and triggers the sid generation task that
populates SID for existing users / groups.
--netbios-name is optional and allows to specify the NetBIOS Name.
When not provided, the NetBIOS name is generated based on the leading
component of the DNS domain name.
This command can be run multiple times.
Fixes: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ADTRUSTInstance performs only sid configuration when it is
called without --setup-adtrust.
Update man pages for ipa-server-install and ipa-replica-install
with the SID-related options.
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Move the SID-related options into a separate InstallInterface
(--add-sids, --netbios-name, --rid-base and --secondary-rid-base),
make ADTrustInstallInterface inherit from SIDInstallInterface.
Related: https://pagure.io/freeipa/issue/8995
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
New PR-CI template based on compose `Fedora-Rawhide-20211021.n.0`,
future Fedora 36.
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Previously, `test_master_install_ca1` and `test_master_install_ca2`
attempt to create tempdirs on local host and later write some
content into the returned paths on remote host. This fails if
a remote host is a local one.
The existent `create_temp_file` function has been extended to
support `suffix` option of `mktemp`.
Fixes: https://pagure.io/freeipa/issue/9013
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The default platform task enable_sssd_sudo() writes directly
to nsswitch.conf to enable sudo. This isn't necessary to do on
systems with authselect where we already pass in with-sudo as a
profile option.
Override the default function with does a direct write with a no-op.
https://pagure.io/freeipa/issue/8755
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ipa-server-install uninstaller must remove samba *.tdb files
in /var/lib/samba, /var/lib/samba/private and /var/lib/samba/lock.
The current code calls rm on the relative path filename
instead of building an absolute path filename,
resulting in failure to remove the tdb files.
Related: https://pagure.io/freeipa/issue/8687
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ipa-client-samba uninstaller must remove samba *.tdb files
in /var/lib/samba, /var/lib/samba/private and /var/lib/samba/lock.
The current code calls rm on the relative path filename
instead of building an absolute path filename,
resulting in failure to remove the tdb files.
Fixes: https://pagure.io/freeipa/issue/8687
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
As of 21.3 pip:
> Remove the --build-dir option and aliases, one last time. (pypa/pip#10485)
https://pip.pypa.io/en/stable/news/#v21-3
Previous versions warn about deprecation.
The builddir is provided to pip via env variable PIP_BUILD in Tox task.
The purpose of changing of default builddir was noexec mount option for
/tmp in Travis (see 17d571c961). Since Travis is no longer used and
Azure lacks this issue the PIP_BUILD can be safely removed.
Note: pip 21.3 just ignores this env variable, which is more than can be
said for the command line option. It's better to clean it up, since the
behaviour may be changed in future.
This is effectively the revert of 17d571c961.
Fixes: https://pagure.io/freeipa/issue/9011
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
I don't know why this wasn't always multi-value but if one wanted
to set multiple options they needed to call add-option multiple
times. The LDAP attribute is already multi-value.
This shouldn't cause API issues as it understood the attribute as
multi-value just didn't expose it. Client output on the CLI will
look a bit different:
Added option "('one', 'two')" to Sudo Rule "test"
or
Added option "(u'one', u'Two')" to Sudo Rule "test"
instead of with this change:
Added option "one,two" to Sudo Rule "test"
Removing an option works in a similar way.
The value is normalized on the client side in order to ensure that
the option value is always a tuple.
https://pagure.io/freeipa/issue/2278
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The above test is installing 2 Let's Encrypt certificates:
the root ISRG Root X1 and a subca. The subca expired Oct 6 and needs to
be replaced with a valid one, otherwise ipa-cacert-manage install
refuses to install it.
Fixes: https://pagure.io/freeipa/issue/9006
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
Get size of the log file immediately before main test action to avoid
capturing messages written to log during environment preparation.
Fixes https://pagure.io/freeipa/issue/8987
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The test test_undo_refresh_reset_update_cancel is sometimes
failing because a notification obscures the selinuxmap record.
After saving the modification on the record, close any notification
to make sure the test succeeds.
Fixes: https://pagure.io/freeipa/issue/8846
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
The test TestIpaHealthCheckWithExternalCA::test_ipahealthcheck_ipaopensslchainvalidation
needs to be adapted with the new error message returned by
openssl verify when the provided certificate file does not exist.
The message changed with openssl3.
Fixes: https://pagure.io/freeipa/issue/8999
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Previously, all the commands of schema plugin derived from
BaseMetaSearch require metaobject as their argument
(by implementation), but the spec for some of them only optionally
asks for search criteria arg. This patch fixes this inconsistency.
Fixes: https://pagure.io/freeipa/issue/8954
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
usercertificate often has a subclass and both the plain and
subclassed (binary) values are queried. I'm concerned that
they are used more or less interchangably in places so not
caching these entries is the safest path forward for now until
we can dedicate the time to find all usages, determine their
safety and/or perhaps handle this gracefully within the cache
now.
What we see in this bug is that usercertificate;binary holds the
first certificate value but a user-mod is done with
setattr usercertificate=<new_cert>. Since there is no
usercertificate value (remember, it's usercertificate;binary)
a replace is done and 389-ds wipes the existing value as we've
asked it to.
I'm not comfortable with simply treating them the same because
in LDAP they are not.
https://pagure.io/freeipa/issue/8986
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
SSSD provides Kerberos plugin
> to tell the Kerberos libraries what Realm and which KDC to use.
It's useful to see what is happening during kinit in case of any
issues.
Related: https://pagure.io/freeipa/issue/8353
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There are time gaps in which kinit requests may fail due to
offlined SSSD's locator and replication delays.
Since `IPA` provider or SSSD offline the locator plugin for libkrb5
(man 8 sssd_krb5_locator_plugin) can do nothing about this and kinit
fallbacks to the standard libkrb5 algorithm described in `man 5 krb5.conf`.
`krb5.conf` on IPA server doesn't include `kpasswd_server` and kinit
fallbacks to DNS way. DNS (URI or SRV) RRs don't preserve any order
and kinit may contact either master or replica kpasswd servers.
This may result in a password was changed on a replica but was not
replicated to master:
master(kinit)->master(initial)->replica(kpasswd)->master(can't
obtain initial creds with new password)
So, `kpasswd_server` serves as fallback for the offlined locator.
Note: primary_kdc(the former master_kdc) doesn't help here because
it is only used if the initial credentials obtaining fails (see
`krb5_get_init_creds_password` in libkrb5) and not a password change.
Fixes: https://pagure.io/freeipa/issue/8353
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Sometimes tar fails on changed in process files:
```
[2021-09-07 11:03:33] + tar --ignore-failed-read -czf ipaserver_install_logs.tar.gz --warning=no-failed-read /var/log/dirsrv /var/log/httpd2 /var/log/ipa /var/log/ipaclient-install.log /var/log/ipa-custodia.audit.log /var/log/ipaserver-install.log /var/log/krb5kdc.log /var/log/pki /var/log/samba /var/lib/bind/data systemd_journal.log
[2021-09-07 11:03:33] tar: Removing leading `/' from member names
[2021-09-07 11:03:33] tar: Removing leading `/' from hard link targets
[2021-09-07 11:03:33] tar: /var/log/dirsrv/slapd-IPA-TEST/access: file changed as we read it
[2021-09-07 11:03:33] + tests_result=1
```
This is expected failure since processes are not stopped during logs
collection and can flush their logs.
Fixes: https://pagure.io/freeipa/issue/8983
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Add a template for new features in doc/designs/template.md
The template is excluded from automatic doc generation.
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
OpenSSH 8.7p1 changed the message logged on successful
authentication (see commit 9e1882ef6489a7dd16b6d7794af96629cae61a53).
As a result, the method run_ssh_cmd is failing and needs to be
adapted in order to be compatible with old and new openssh versions.
Fixes: https://pagure.io/freeipa/issue/8989
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Using JSON by default within Dogtag appears to cause ipa cert-find to
return JSON, when the request was made with XML. We can request that XML
is returned as before by specifying so in the request header.
Fixes: https://pagure.io/freeipa/issue/8980
Signed-off-by: Chris Kelley <ckelley@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
The logic of test_reset_password_unlock is twisted.
Currently it's doing:
- reset password on replicas[0]
- wait for replication on master
- kinit on master
The call to wait_for_replication should be done on
replicas[0], not on master, according to the method doc:
Note that this waits for updates originating on this host, not those
coming from other hosts.
Fixes: https://pagure.io/freeipa/issue/8975
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
Migrate-ds needs to check if compat tree is enabled before
migrating users and groups. The check is doing a base
search on cn=compat,$SUFFIX and considers the compat tree
enabled when the entry exists.
Due to a bug in slapi-nis, the base search may return NotFound
even though the compat tree is enabled. The workaround is to
perform a base search on cn=users,cn=compat,$SUFFIX instead.
Fixes: https://pagure.io/freeipa/issue/8984
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
A similar change was attempted years ago in commit
9724251292 but it was
never applied because it used the wrong DN and because
nsslapd-timelimit is already present in the entry
the default keyword won't trigger.
Use replace instead to increase the value to 100k from
the default as originally intended.
nsslapd-sizelimit can be changed only with a MOD_REPLACE
otherwise a LDAP_NO_SUCH_ATTRIBUTE error is thrown. IPA
only uses MOD_REPLACE for single-value attributes but
nsslapd-sizelimit is not yet in schema. Add it to
the known set of exceptions for single-value attributes.
https://pagure.io/freeipa/issue/8962
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Previously, the subid-match command would output the full
DN of the owner of the matched range.
With this change, the UID of the owner is displayed, just like
for other subid- commands.
Fixes: https://github.com/freeipa/freeipa/pull/6001
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Rather than stopping the installer entirely, catch and report
errors adding new certificate profiles, and remove the
broken profile entry from LDAP so it may be re-added later.
It was discovered that installing a newer IPA that has the
ACME profile which requires sanToCNDefault will fail when
installing a new server against a very old one that lacks
this class.
Running ipa-server-upgrade post-install will add the profile
and generate the missing ipa-ca SAN record so that ACME
can work.
https://pagure.io/freeipa/issue/8974
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
If a client sends a request to lookup an object from a given trusted
domain by UID or GID and an object with matching ID is only found in a
different domain the extdom should return LDAP_NO_SUCH_OBJECT to
indicate to the client that the requested ID does not exists in the
given domain.
Resolves: https://pagure.io/freeipa/issue/8965
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
MIT KRB5 1.15 introduced KDC service discovery with URI records.
_kerberos and _kpasswd URI records can provide TCP, UDP, and Kerberos
KDC-Proxy references. URI lookups take precedence over SRV lookups,
falling back to SRV lookups if no URI records are found.
Also reduce TTL for system records from one day to one hour. It allows
users to remove or update discovery entries in a timely fashion.
See: https://web.mit.edu/kerberos/krb5-latest/doc/admin/realm_config.html#kdc-discovery
Fixes: https://pagure.io/freeipa/issue/8968
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
On aarch64, custodia creates AVC when accessing /proc/cpuinfo.
According to gcrypt manual
(https://gnupg.org/documentation/manuals/gcrypt/Configuration.html),
/proc/cpuinfo is used on ARM architecture to read the hardware
capabilities of the CPU. This explains why the issue happens only
on aarch64.
audit2allow suggests to add the following:
allow ipa_custodia_t proc_t:file { getattr open read };
but this policy would be too broad. Instead, the patch is using
the interface kernel_read_system_state.
Fixes: https://pagure.io/freeipa/issue/8972
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
The DogtagInstance.spawn_instance() and uninstall() have
been modified to specify the paths of PKI installation
logs using --log-file option on PKI 11.0.0 or later.
This allows IPA to have a full control over the log files
instead of relying on PKI's default log files.
Fixes: https://pagure.io/freeipa/issue/8966
Signed-off-by: Endi Sukma Dewata <edewata@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: François Cami <fcami@redhat.com>
When a test is executed around midnight and is checking the
journal content with --since=date, it needs to specify the
whole date (with day and time) to avoid missing entries.
If for instance --since=23:59:00 is used and the current time is
now 00:01:00, --since=23:59:00 would refer to a date in the
future and no journal entry will be found.
Fixes: https://pagure.io/freeipa/issue/8953
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: François Cami <fcami@redhat.com>
- previously, fastlint make's target includes both the Pylint task
and pycodestyle one. The purpose of this target is a fast checking
only for changed Python files. This makes sense for pycodestyle, but
limits Pylint due to a context(file) checking. The clients which
call the code being linted are not checked at all. In Azure Pylint
(for the whole codebase) is run in the Lint task, this makes fastlint
extra for Azure.
- `Quick code style check` task used distro's Pylint, while `Lint`
task PyPI's one. This may cause different results and confuse a
user.
- `Build` task takes time longer than `Lint` one, so this change
doesn't lead to increased CI time.
- all Azure tests depend on Build and Lint tasks. Mostly it's no need
to run tests due to a probably broken code.
Fixes: https://pagure.io/freeipa/issue/8961
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: François Cami <fcami@redhat.com>
LDAP/HTTP/PKINIT certificates should be renewd on replica after
moving system date. Test was failing because ipa-cert-fix ran
while these cert was not renewd and it tried to fix it.
This test adds check for replication before calling ipa-cert-fix
on replica.
Fixes: https://pagure.io/freeipa/issue/8815
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
- the base testing of this command is made by ipaclient `schema`
remote plugin, but some specifics are not covered
- allow testing of the plugin in `development` mode(locked API).
Fixes: https://pagure.io/freeipa/issue/8955
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
If some Param defines several values for `exclude` or `include`
attributes then API schema hash will be unstable.
First, these Param's attributes are converted to frozenset
(ipalib/parameters.py), then `ipaserver.plugins.schema` plugin
converts `exclude` and `include` attrs to list. Set/frozenset in
turn, is unordered collection [0]. So, the end order of values is
undefined.
But due to the nature of sets:
> two sets are equal if and only if every element of each set is
contained in the other (each is a subset of the other)
the order of values can be ignored.
Note: other Param's attrs with type frozenset are not affected because
they are not processed by the schema plugin.
[0]: https://docs.python.org/3/library/stdtypes.html#set-types-set-frozenset
Fixes: https://pagure.io/freeipa/issue/8955
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
PKI has its own internal knowledge of servers and services
in its securitydomain. This has not been cleaned up in the
past but is becoming more of an issue as PKI now relies on its
securitydomain for more things, and it has a healthcheck that
reports inconsistencies.
Removing entries is straightforward using the PKI REST API.
In order to operate on the API access is needed. There was an
unused Security Domain Administrators group that I've added to
the resourceACLS we created for managing the securitydomain.
The ipara user is added as a member of this group. The REST
API binds to the CA using the IPA RA certificate.
Related commits are b3c2197b7e
and ba4df6449a.
These resourceACLS were originally created as a backwards
compatibility mechanism for dogtag v9 and later only created when a
replica was installed purportedly to save a restart. I don't see
any reason to not have these defined. They are apparently needed due
to the PKI database upgrade issues.
In any case if the purpose was to suppress these ACLS it failed
because as soon as a replica with a CA was installed they were as
well, and we need this ACL in order to manage the securitydomain.
https://pagure.io/freeipa/issue/8930
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The test configures an external bind server on the ipa-server
(not the IPA-embedded DNS server) that allows unauthenticated nsupdates.
When the IPA client is registered using ipa-client-install,
DNS records are added for the client in the bind server using nsupdate.
The first try is using GSS-TIG but fails as expected, and the client
installer then tries with unauthenticated nsupdate.
Related : https://pagure.io/freeipa/issue/8402
Signed-off-by: Anuja More <amore@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Tests in `test_ipaserver.test_secure_ajp_connector' assume that there
is pkiuser in OS, but this is not always true (for example, in systems
having minimum installed dependencies, in particular, without pki-server
RPM package). Since the tests already use the mock and pkiuser entity is
not the subject of testing the pwd.getpwnam has been mocked.
Fixes: https://pagure.io/freeipa/issue/8942
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Use whole date when calling journalctl --since
ipa-otpd don't flush its logs to syslog immediately,
so check with run_repeatedly.
Also list failed units when ldap connection is
timed out.
Related: https://pagure.io/freeipa/issue/6587
Signed-off-by: Anuja More <amore@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The install helpers used an invalid string check. ``('ubuntu')`` is
not a tuple. It's a string with superfluous parenthesis. A single-item
tuple would be ``('ubuntu',)``. It's recommended to use set literals to
avoid such mistakes.
Also check for 'debian' platform.
Fixes: https://pagure.io/freeipa/issue/8937
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
A user wrote their own plugin to add custom attributes which was
failing with an incorrect error that the attribute wasn't allowed.
It wasn't allowed because it wasn't being treated as case-insensitive
so wasn't being found in the schema.
https://pagure.io/freeipa/issue/8415
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
tasks.create_active_user can fail in a subtle way when there
are two IPA servers due to replication delays.
Using the debug-enabled version of create_active_user helps
determine whether there is another underlying issue and, in
general, prevents the above problem.
Fixes: https://pagure.io/freeipa/issue/8944
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
The error message if ipa-ca can't be resolved included the
undefined variable ${domain_name}. Since this is static anyway
change to a python format string and hardcode the string in
the resulting script as api.env.domain.
Discovered while working on https://pagure.io/freeipa/issue/8934
Related: https://pagure.io/freeipa/issue/8934
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
p11-kit-proxy in newer distributions handles loading the OpenSC
PKCS#11 library so don't try to add it to the NSS database in
/etc/pki/nssdb if it is already available in order to avoid a
potentially confusing error message.
https://pagure.io/freeipa/issue/8934
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
ipa-certupdate obtains host credentials to operate. If this
fails with a ccache error this can be confusing if the user
executing it already has admin credentails.
Include the principal being retrieved and the keytab being
used.
This basically intercepts the exception to log additional
information and lets the exception be handled at a higher
level.
https://pagure.io/freeipa/issue/8257
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Test various usages of the -s/--server option:
* -s is defined, use it as the server
* no -s, use the host value from /etc/ipa/default.conf
* -s is '_srv_', do DNS discovery
https://pagure.io/freeipa/issue/8478
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Make read_ipa_config and filter_keys static to avoid
"no previous prototype" warnings.
Use correct datatype of return value for ber_scanf to
correct different signedness comparision.
Fixed while working on https://pagure.io/freeipa/issue/8478
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The basic flow is:
- If server is provided by the user then use it
- If server the magic value '_srv', check for _ldap._tcp SRV records for
the domain in /etc/ipa/default.conf
- If no servers are found use the server from default.conf
https://pagure.io/freeipa/issue/8478
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Only the first key of orphaned automount keys was being
displayed.
tofiles was created because making sense of LDAP automount
information is a brain squeezer. The purpose is not to
display in a precise file format but to display it in
a sensible and understandable way.
https://pagure.io/freeipa/issue/7814
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
test_integration/test_installation.py::TestInstallMaster 's
timeout is 10800 on all nightlies but it timeouts in gating with a
timeout of 3600. Use 7200 in gating so that it has some chance of
completing.
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
test_cert and test_SubCAkeyReplication are randomly failing.
The suspect for test_SubCAkeyReplication is an nss bug:
https://bugzilla.redhat.com/show_bug.cgi?id=1985061
The reason for test_cert failures was not identified, the only
relevant line in the log contains:
2021-07-22T17:37:21.0873339Z tests: cert, result: 1, time: 30:08.98
2021-07-22T17:37:21.0874172Z Command exited with non-zero status 1
Disable these tests in gating until the NSS bug is fixed and
the related build is available in Fedora.
Related: https://pagure.io/freeipa/issue/8864
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Use the new role-based mechanism, one that doesn't rely
on direct communication to the server, to determine whether
the server being removed by `ipa server-del` contains the
last KRA server.
https://pagure.io/freeipa/issue/8397
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
It previously used a vault connection to determine if any
KRA servers were installed. This would fail if the last KRA
was not available.
Use server roles instead to determine if the last KRA server
is to be removed.
https://pagure.io/freeipa/issue/8397
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
A kinit immediately following a password change can fail.
Setting KRB5_TRACE and retrieving kdcinfo will help to understand
the cause of failure.
Fixes: https://pagure.io/freeipa/issue/8929
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
The webui tests for ID range evaluate a potentially free id range
by looking for existing ranges and picking a range = max value
+ 1 million.
With the addition of subuid range this algorithm produces values
over the limit because the subuid range goes from
2,147,483,648 to 4,294,836,224 and the max base id is 4,294,967,295.
Ignore the subuid range when picking a potential range.
Fixes: https://pagure.io/freeipa/issue/8919
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The index definition for memberOf is inconsistent:
dn: cn=memberOf,cn=index,cn=userRoot,cn=ldbm database,cn=plugins,cn=config
cn: member
nsIndexType: eq
nsIndexType: sub
nsSystemIndex: false
objectClass: top
objectClass: nsIndex
The cn attribute should be memberOf, not member. Fix the definition.
Fixes: https://pagure.io/freeipa/issue/8920
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
When a cert is revoked, a notification is displayed
and may obscure the buttons. Make sure to close the
notification before moving to the next step.
Fixes: https://pagure.io/freeipa/issue/8911
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
When adding a new host the principal cannot be determined because it
relies on either:
a) an entry to already exist
b) krbprincipalname be a component of the dn
As a result the full dn is being passed into ipapython.Kerberos
which can't parse it.
Look into the entry in validate_validate_auth_indicator() for
krbprincipalname in this case.
https://pagure.io/freeipa/issue/8206
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
KRA instance import depends on lib389 package, which is not always
installed and that results in failure. Furthermore, test_installation
utilizes krainstance import. This fix moves relevant parts from
krainstance to ipalib constants where those are subsequently imported
from.
Related: https://pagure.io/freeipa/issue/8795
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Tibor Dudlak <tdudlak@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The test TestSelfExternalSelf::test_switch_back_to_self_signed
is checking the content of the journal using journalctl --since ...
but provides only the time, not the whole date with year-month-day.
As a consequence, if the test is executed around midnight it may
find nothing in the journal because it's looking for logs after 11:50PM,
which is a date in the future.
Fixes: https://pagure.io/freeipa/issue/8918
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
It required support in dogtag which was added in 10.5.0.
This is only easily configurable during installation because
it will set ca.signing.defaultSigningAlgorithm to the
selected algorithm in CS.cfg
The certificate profiles will generally by default set
default.params.signingAlg=- which means use the CA default.
So while an existing installation will technically allow
SHA384withRSA it will require profile changes and/or
changing the defaultSigningAlgorithm in CS.cfg and
restarting (completely untested). And that won't affect
already issued-certificates.
https://pagure.io/freeipa/issue/8906
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
- add "Subordinate ID Statistics" page
- add button for generating subid in "Subordinate ids" tab of user details page
- allow to navigate directly to owner details from subordinate id page
- adjust i18n strings
Ticket: https://pagure.io/freeipa/issue/8361
Signed-off-by: Serhii Tsymbaliuk <stsymbal@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Subordinate ids are now handled by a new plugin class and stored in
separate entries in the cn=subids,cn=accounts subtree.
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
New LDAP object class "ipaUserSubordinate" with four new fields:
- ipasubuidnumber / ipasubuidcount
- ipasubgidnumber / ipasgbuidcount
New self-service permission to add subids.
New command user-auto-subid to auto-assign subid
The code hard-codes counts to 65536, sets subgid equal to subuid, and
does not allow removal of subids. There is also a hack that emulates a
DNA plugin with step interval 65536 for testing.
Work around problem with older SSSD clients that fail with unknown
idrange type "ipa-local-subid", see: https://github.com/SSSD/sssd/issues/5571
Related: https://pagure.io/freeipa/issue/8361
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
At the end of a server installation, the DNS records for
ipa-ca.$DOMAIN are created/updated with the IP addresses of the
new server.
The current code resolves the IP addresses of the new server
but doesn't check them. This can result in the addition of
a link-local address to ipa-ca record.
For each address, make sure that it's neither reserved nor a
link-local address.
Fixes: https://pagure.io/freeipa/issue/8810
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
The command
ipa stageuser-add --user-auth-type=xxx
is currently failing because the objectclass ipauserauthtypeclass
is missing from the created entry.
There is code adding the missing objectclass in the
pre_common_callback method of user_add, and this code should
be common to user_add and stageuser_add. In order to avoid code
duplication, it makes more sense to move the existing code to
pre_common_callback of baseuser_add, that is called by both
classes.
Fixes: https://pagure.io/freeipa/issue/8909
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Authentication indicators should not be added to internal IPA services,
since this can lead to a broken IPA setup. In case a client with
an auth indicator set in its host principal, promoting it to a replica
should fail.
Related: https://pagure.io/freeipa/issue/8206
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Authentication indicators should not be enforced against internal
IPA services, since not all users of those services are able to produce
Kerberos tickets with all the auth indicator options. This includes
host, ldap, HTTP and cifs in IPA server and cifs in IPA clients.
If a client that is being promoted to replica has an auth indicator
in its host principal then the promotion is aborted.
Fixes: https://pagure.io/freeipa/issue/8206
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Some plugins, notably dns, modifies a returned entry in order
to compare it to the user-provided info (e.g. dnsrecord-del).
This modification was done on the cached entry directly rather
than a copy which caused unexpected results, mostly
EmptyResult because the cached entry was changed directly so
the next get_entry returned the same modified entry.
In addition, on a hit in the LDAP cache the entire cached entry
was returned regardless of what attributes were requested.
The automember condition add/remove calls only request the
inclusive/exclusive rule attributes and loop over the returned
values to look for duplicates. This was failing because the queried
entry contains attributes that the candidate entry does not contain.
The automember code is:
old_entry = ldap.get_entry(dn, [attr])
for regex in old_entry.keys():
if not isinstance(entry_attrs[regex], (list, tuple)):
old_entry, returned from the cache, contained objectclass, cn,
description, etc. which don't exist in the candidate entry so
entry_attrs[regex] threw a KeyError.
To return a copy of the entry and requested attributes on a
search HIT.
Also be more careful when storing the attributes in the cache entry.
The returned attributes may not match the requested. So store the
attributes we actually have.
This issue was exposed by Ansible which maintains a larger and
longer-lived cache because commands are executed in the server context
one after another, giving the cache a chance to build up.
Adjust the expected test results as well. In test_get_testuser()
the first request asks for all attributes (default) so ensure
that is successful since a user_add gets all attributes in
the post_callback. Next request a subset of the attributes which
is also a hit and confirm that only those requested were returned.
https://pagure.io/freeipa/issue/8897
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Previously there were no tests for `ipalib.misc` module.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The plugin `plugins` iterates over the keys of API instance,
__iter__ of which is a generator of class.__name__ from
(Command, Object, Method, Backend, Updater). So, the allowed type
is str, not bytes.
Fixes: https://pagure.io/freeipa/issue/8898
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
If dynamic DNS updates are selected, sssd will use GSS-TSIG
by default for nsupdate.
When ipa-client-install notices that plain nsupdate is required,
switch sssd to use no authentication for dynamic updates too.
Fixes: https://pagure.io/freeipa/issue/8402
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
ipa-client-install invokes nsupdate with GSS-TSIG at client
enrollment time. If that fails, no retry is done.
Change that behavior to try again without GSS-TSIG.
Fixes: https://pagure.io/freeipa/issue/8402
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Big numbers are automatically translated to scientific notation in JavaScript.
It causes an issue with some certificate serial numbers.
The fix normalizes the notation base on original value from serial_number_hex.
The implementation works only for browsers that support BigInt.
It would not work for old browsers like Internet Explorer.
Ticket: https://pagure.io/freeipa/issue/8754
Signed-off-by: Serhii Tsymbaliuk <stsymbal@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
As of 0.9 freeipa-healthcheck requires running `dirsrv` service
for `IPAHostKeytab` check. So, previous assumption about the
triggering the GSSAPI error no longer works. For example, this can
be achieved by deletion of host's keytab.
Fixes: https://pagure.io/freeipa/issue/8889
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This allows for a consistent way to retrieve the value from
LDAP. The method is used to find an existing entry. It is not usable
to add or remove entries.
Moving it in the code allows the value to always be set in the
substitution dictionary and not rely on a specific caller.
It was moved to installutils.py to avoid circular import.
https://pagure.io/freeipa/issue/8885
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This test checks that if 'selftests.container.order.startup' directive
is missing from CS.cfg, ipa-cert-fix fails and throw proper error
message. It also checks that underlying command 'pki-server cert-fix'
should fail to renew the cert.
related: https://pagure.io/freeipa/issue/8721
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Trying to resolve the FQDN before command execution (during
pre-callback) helps detect cases where the host specified by the user
does not exist, saving execution time. Aside from this, resolving the
FQDN is useful when only the shortname of the host is passed, as this
would cause issues when trying to update the DNS records during
modification of the entry.
Fixes: https://pagure.io/freeipa/issue/8726
Fixes: https://pagure.io/freeipa/issue/8884
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When upgrading from a server with IPA CA before PKINIT was introduced
(4.5), PKINIT would not be enabled and there wasn't any way to enable it
since upgrade code only issues self-signed certificates when
certificates are missing. With these change there is a way to enable
PKINIT when coming from a IPA server with a pre-PKINIT version (4.4 and
before).
Fixes: https://pagure.io/freeipa/issue/8532
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The CA-less installation creates an external CA with the
subject CN=CA,O=Example Organization.
In order to test non-ascii subjects, use
CN=CA,O=Example Organization España
instead.
Related: https://pagure.io/freeipa/issue/8880
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
In a CA-less install, if the CA cert subject contains
non-ascii characters, ipa-server-install fails when
configuring SSL for httpd.
The issue happens when calling ipautil.run to extract the keys
from a p12file. The code is using the raw output of the command
and doesn't need to specify capture_output=True, as this option
breaks if the output contains non-ascii characters.
The raw_output contains bytes, the output is a str built by decoding
the raw_output and may fail if non-ascii characters are present.
Fixes: https://pagure.io/freeipa/issue/8880
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
`mod_auth_gssapi` exports delegated credentials into `/run/ipa/ccaches`
and pass down that path as `KRB5CCNAME` env variable to WSGI worker.
GSSProxy in turn, protects these credentials from direct usage of
`ipa-api`. But the configuration of `service/ipa-api` (in particular,
'cred_store = client_keytab:/var/lib/ipa/gssproxy/http.keytab') and
default GSS name ('=None') dictates to refresh expired credentials
with the client's keytab overwriting the origin credentials with
initial credentials of keytab's principal.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Previously, `get_credentials` raises either ValueError or re-raises
GSSError. The former makes the handling of this function more difficult
without a good reason.
With this change:
- `get_credentials` no longer handles exceptions by itself, but delegates
this to the callers (which already process GSS errors).
- `get_credentials_if_valid` doesn't raise any expected exceptions, but
return valid credentials (on the moment of calling) or None. This makes
it consistent with docs.
Related: https://pagure.io/freeipa/issue/8873
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The RADIUS auth type is only supported with Kerberos currently.
This design proposes a way to make it work with LDAP binds
as well without relying ok workarounds.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ipa-cert-fix man page needs to explain that certmonger may
trigger a renewal right after ipa-cert-fix completes because
certmonger does not notice the updated certificates.
Also add a similar note at the end of ipa-cert-fix.
Fixes: https://pagure.io/freeipa/issue/8702
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Also disable in Fedora 33 as it also has the faulty version of sssd
which produces multi-gigabyte log file
Related to https://pagure.io/freeipa/issue/8877
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
During test run on Fedora 34 and 35 sssd produces multi-gigabyte log file
which causes test runners to run out of disk space.
Related to https://pagure.io/freeipa/issue/8877
Reviewed-By: Francois Cami <fcami@redhat.com>
The test
test_installation.py::TestInstallWithCA1::test_install_with_bad_ldap_conf
is uninstalling a replica by calling ipa-server-install --uninstall
directly, instead of deleting the replica first.
Use tasks.uninstall_replica instead of tasks.uninstall_master
to perform a proper uninstallation.
Fixes: https://pagure.io/freeipa/issue/8876
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In order to test FIPS mode, the test is faking a user-space
FIPS environment by creating a file /var/tmp/userspace-fips
and bind-mounting this file as /proc/sys/crypto/fips_enabled
The security context needs to be properly set otherwise
/proc/sys/crypto/fips_enabled inherits the security context
unconfined_u:object_r:user_tmp_t:s0 and cannot be read,
resulting in the test seeing fips_mode=false.
Fixes: https://pagure.io/freeipa/issue/8868
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Update instructions on how to build images starting with Fedora 34 using
kickstart files used by Fedora to build its cloud images.
Change vagrant provisioning steps to support both prebuilt and default
cloud images, removing the burden of maintaining boxes up-to-date, but
also providing a way to build fresh images without external packer
templates.
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Fedora ELN represents itself as a RHEL but it does not have
redhat-logos-ipa package. CentOS Stream does not have redhat-logos-ipa
but has centos-logos-ipa package. Both RHEL and CentOS Stream provide
system-logos-ipa so we can depend on it instead.
This allows to make IPA packages installable on CentOS Stream and on
Fedora ELN.
Fixes: https://pagure.io/freeipa/issue/8874
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
get_credentials() was changed to raise ValueError instead of
gssapi.exceptions.GSSError as part of the sweeper to clean up
expired credentials caches.
For WebUI users, this will prevent a 500 error if their
associated credentials cache is expired or missing.
https://pagure.io/freeipa/issue/8873
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Add 'Auto private groups' field on 'Add ID range' form with the following options: true, false, hybrid.
The field is optional and can be omitted.
Its value can be also modified on 'Range Settings' page after the range is added.
Ticket: https://pagure.io/freeipa/issue/8837
Signed-off-by: Serhii Tsymbaliuk <stsymbal@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
HTTP service uses different user for keytab ownership than the service
user. On Fedora this leads to http.keytab being owned by 'apache' user
after IPA deployment while it should be owned by 'root' to allow
GSSPROXY configuration to work correctly.
The situation is fixed during upgrade (ipa-server-upgrade) but it means
for new deployments there might be a period of unexplained Web UI
authentication failures.
Fixes: https://pagure.io/freeipa/issue/8872
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Currently translated at 7.2% (340 of 4672 strings)
Translated using Weblate (Finnish)
Currently translated at 6.8% (318 of 4672 strings)
Translated using Weblate (Finnish)
Currently translated at 6.2% (290 of 4672 strings)
Translated using Weblate (Finnish)
Currently translated at 6.2% (290 of 4672 strings)
Added translation using Weblate (Finnish)
Co-authored-by: Jan Kuparinen <copper_fin@hotmail.com>
Translate-URL: https://translate.fedoraproject.org/projects/freeipa/master/fi/
Translation: freeipa/master
ipa-otptoken-import.1 describes the -k option.
However, the long option --keyfile option is also available.
Therefore, add the --keyfile option to ipa-otptoken-import.1.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Update documentation now that we moved IRC channels #freeipa and #sssd
to Libera.chat network.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Since commit 1906afbeb3c8b7140601be7f9bee2f7fef5b0a5e, in order to fix
rhbz#1780082, pki defines AJP connectors using localhost4 and localhost6:
<Connector port="8009" protocol="AJP/1.3" redirectPort="8443" address="localhost4" name="Connector1" secret="..."/>
<Connector port="8009" protocol="AJP/1.3" redirectPort="8443" address="localhost6" name="Connector2" secret="..."/>
When /etc/hosts only defines the following:
127.0.0.1 localhost
::1 localhost
the connector initialization may fail with
java.net.BindException: Address already in use
The installer can add the following definitions to pkispawn cfg file:
pki_ajp_host_ipv4=127.0.0.1
pki_ajp_host_ipv6=::1
in order to force the value to an IP address instead of localhost4/6.
Fixes: https://pagure.io/freeipa/issue/8851
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Otherwise, 'Check for coredumps' task fails with:
```
Verifying : samba-debugsource-2:4.14.4-0.fc34.x86_64 20/20
[Errno 2] No such file or directory: '/var/lib/dnf/rpmdb_lock.pid'
Finishing: Check for coredumps
```
This is due to systemd-tmpfiles(not ready yet).
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
drill is currently broken on F34. Fortunately, there are another
tools for checking DNSSEC trust. One of them is `delv`:
> delv is a tool for sending DNS queries and validating the results,
using the same internal resolver and validator logic as named.
delv sends to a specified name server all queries needed to fetch and
validate the requested data; this includes the original requested query,
subsequent queries to follow CNAME or DNAME chains, queries for DNSKEY,
and DS records to establish a chain of trust for DNSSEC validation. It
does not perform iterative resolution, but simulates the behavior of a
name server configured for DNSSEC validating and forwarding.
Related: https://pagure.io/freeipa/issue/8793
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The list of installed packages may be useful for checking the
versions of packages for analysis. Previously, only the newly
installed packages can be observed on Build phase.
This is convenient for experienced users of PR-CI.
Note: the read-only access provided for non-master containers
to be able to execute Azure scripts. The logs are still collected
only on controller.
Only RPM-based collection is implemented for Fedora. By default
nothing is collected.
Users may want to override `installed_packages` function
in the corresponding `ipatests/azure/scripts/variables-DISTRO.sh`.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There are tons of useless information in test's runner log on
server uninstallation about list trust and certificates, such
as:
```
RUN ['trust', 'list']
pkcs11:id=%D2%87%B4%E3%DF%37%27%93%55%F6%56%EA%81%E5%36%CC%8C%1E%3F%BD;type=cert
type: certificate
label: ACCVRAIZ1
trust: anchor
category: authority
pkcs11:id=%F7%7D%C5%FD%C4%E8%9A%1B%77%64%A7%F5%1D%A0%CC%BF%87%60%9A%6D;type=cert
type: certificate
label: AC RAIZ FNMT-RCM
trust: anchor
category: authority
pkcs11:id=%52%D8%88%3A%C8%9F%78%66%ED%89%F3%7B%38%70%94%C9%02%02%36%D0;type=cert
type: certificate
label: Actalis Authentication Root CA
trust: anchor
category: authority
...
```
This improves the readability of test logs.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There are tons of useless warnings about missing files on collecting
logs, such as:
```
tar: /var/log/ipaserver-kra-install.log: Warning: Cannot stat: No such file or directory
tar: /var/log/ipaepn.log: Warning: Cannot stat: No such file or directory
tar: /etc/NetworkManager/NetworkManager.conf: Warning: Cannot stat: No such file or directory
tar: /var/log/ipabackup.log: Warning: Cannot stat: No such file or directory
tar: /var/log/iparestore.log: Warning: Cannot stat: No such file or directory
...
```
Since `--ignore-failed-read` option is passed to tar the caller
doesn't care about not readable(mostly missing) files and these warnings
may be filtered out.
This improves the readability of test logs.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
By default pytest reports in summary section about tests failures and errors.
It will be helpful to see skipped, xfailed and xpassed tests.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The DNS server may not process a query in a its internal timeout for
a some reason or don't answer for a query at all. This may indicate
a high load on DNS server. For example, if IPA DNS server is
configured with 'none' forward policy (read as resolver), then
SERVFAIL/Timeout errors will be normal until the hot cache for zones.
Resolver's timeout in turn, indicates that it queried a server, but
didn't received an answer in specified timeout.
Related: https://pagure.io/freeipa/issue/7397
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
- check only IPv6 address of local NS if specified
- increase request timeout(2sec is too small, BIND resolver's
default 10sec)
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Since the dns plugin's tests have no access to wild resolvers
nobody answer such requests but authoritative NS.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The tests in these envs make DNS requests to wild(internet) NSs,
though usually tests assume the opposite making requests to
`test.` zone. This makes CI unstable and dependent on wild
resolvers and logically wrong.
In future there can be tests which may want to check BIND as
resolver(cache) for external networks. In this case such tests
should be placed on not isolated mode.
By default, a test env is not isolated from internet(as it was
before), but it may be a good idea to change this default in
future.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The nonzero number of memory/memory+Swap usage hits limits may
indicate the possible env instability(crashes, random failures, etc.).
> memory.failcnt # show the number of memory usage hits limits
memory.memsw.failcnt # show the number of memory+Swap hits limits
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
WebUI unit tests fail with:
```
PhantomJS threw an error:ERROR
>> Auto configuration failed 0 [
>> 'Auto configuration failed',
>> '140613066520384:error:25066067:DSO support routines:DLFCN_LOAD:could not load the shared library:dso_dlfcn.c:185:filename(libssl_conf.so): libssl_conf.so: cannot open shared object file: No such file or directory',
>> '140613066520384:error:25070067:DSO support routines:DSO_load:could not load the shared library:dso_lib.c:244:',
>> '140613066520384:error:0E07506E:configuration file routines:MODULE_LOAD_DSO:error loading dso:conf_mod.c:285:module=ssl_conf, path=ssl_conf',
>> '140613066520384:error:0E076071:configuration file routines:MODULE_RUN:unknown module name:conf_mod.c:222:module=ssl_conf'
>> ]
...
Warning: PhantomJS exited unexpectedly with exit code 1. Use --force to continue.
Aborted due to warnings.
```
See https://github.com/wch/webshot/pull/93 for details.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Since https://github.com/systemd/systemd/pull/17902/commits/a4a9a6f7c6e9cd9e219c56d08434a04bc2f395ff
systemd improves the detection of Docker and Podman containers based
on the presence of files-markers.
```console
[slev@test systemd]$ git describe --contains --tags a4a9a6f7c6e9cd9e219c56d08434a04bc2f395ff
v248-rc1~155^2~1
```
Note: on Azure unit tests are run as non-privileged user in non-systemd
inited container.
This worked on F32 because:
```console
[root@6d2aad38f62c /]# rpm -q systemd
systemd-245.9-1.fc32.x86_64
```
So, actual comparison in test was `assert None == None`.
But F34 has:
```console
[root@1ff1325f5a61 /]# rpm -q systemd
systemd-248-2.fc34.x86_64
```
So, the test's expectations should be updated.
Unfortunately, this is incompatible with older versions of systemd
(< v248).
See https://github.com/systemd/systemd/pull/17902 for details.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The initial value of NS of resolv.conf is 127.0.0.11, this
is the embedded NS of docker-compose. The disabling of
this feature is not currently supported by Docker.
On startup systemd-resolved caches the /etc/resolv.conf
(docker-compose version), which is later modified by
setup_containers.py script.
This results in resolving error occurs:
```console
[root@replica1 /]# getent ahosts master1.ipa.test
... can't resolve
[root@replica1 /]# grep 'hosts:' /etc/nsswitch.conf
hosts: files myhostname resolve [!UNAVAIL=return] dns
[root@replica1 /]# resolvectl status
Global
LLMNR setting: resolve
MulticastDNS setting: no
DNSOverTLS setting: no
DNSSEC setting: no
DNSSEC supported: no
Current DNS Server: 127.0.0.11
DNS Servers: 127.0.0.11
Fallback DNS Servers: 1.1.1.1
8.8.8.8
1.0.0.1
8.8.4.4
2606:4700:4700::1111
2001:4860:4860::8888
2606:4700:4700::1001
2001:4860:4860::8844
```
According to docs:
https://www.freedesktop.org/software/systemd/man/systemd-resolved.service.html#/etc/resolv.conf
our case is 4(managed by other packages).
So, restart of systemd-resolved is enough for its re-initialization,
but not for services that already received DNS results. To speed up
the overall process and to no restart each service which wants
internet connection(or wait until service retries connection)
systemd-resolved is masked.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The calling of systemd's utils during systemd boot may lead to
unpredictable results. For example, if DBus(dbus-broker) service
is not started then DBus request goes nowhere and eventually will
be timeouted. So, it's safer to wait fully booted system.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This may speed up boot process.
For example, 'fedora:34' set graphical.target as default,
while multi-user one will be more appropriate.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
If an error occured while containers setup phase then no logs will
be collected and it is hard(impossible?) to debug such issues on
remote Azure host. With this change in case of such error all the
container's journals will be collected in `systemd_boot_logs`.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
We've seen some intermittent failures of this test with warnings
about passwords about to expire. We suspect there may be a time
difference between the client and server so set the passwords
on the server instead to be sure that time is correct.
https://pagure.io/freeipa/issue/8785
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This reverts commit aa1350384a.
The search for certificates is a complex, three-step process,
which filters results in subsequent searches. This filters out
non-relevant certificates when deleting a host or service.
This patch breaks that so deleting one service of a host will
revoke *all* certificates for that host.
Another attempt will be made separately to implement this.
https://pagure.io/freeipa/issue/7835
Signed-off-by: Rob Crittenden rcritten@redhat.com
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The -s option no longer exists, so remove it from the command usage.
Also, due to this fix, E128 occurred in pycodestyle, so the coding style was changed.
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
There are no arguments in the --entry option,
but DN for the managed entry definition must actually be specified.
Therefore, add MANAGED_ENTRY as an argument.
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
There are no arguments in the --schema-file option,
but the Schema file in LDIF format must actually be specified.
Therefore, add FILE.ldif as an argument
In addition, the --schema option no longer exists, so remove the description.
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
In order to call import_included_profiles the dogtag RA plugin
needs to have been loaded. Modify the requirements to also allow
the installer context along with the ra_plugin value.
This lets us add missing profiles during a replica installation.
This is needed for ACME when installing a new replica in a
cluster of older servers that don't have support for ACME.
https://pagure.io/freeipa/issue/8738
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
ipa-winsync-migrate.1 has an explanation of options, but no arguments.
Therefore, add the arguments for --realm and --server.
Also, add a short option -U for --unattended.
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The command usage and man-page options may not match.
In ipa-client-automount, fix to match usage and man-page.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The ipa-client-samba man-page describes the -d option, but the -d option cannot actually be used.
Fix ipa-client-samba to enable the -d option.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Normally it takes 50 minutes for test_commands test suite
to complete. But due to fluctuations in virtualized environment
sometimes it exceeds 60 minutes which produces random failures
in nightly runs.
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
update_dna_shared_config plugin now waits for presence of both
``dnaHostname=fqdn+dnaPortNum=0`` and
``dnaHostname=fqdn+dnaPortNum=389`` entries before it updates the
entries.
The commit also refactors the code of update_dna_shared_config for
subordinate id support.
Fixes: https://pagure.io/freeipa/issue/8831
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Read the FINAL lines from the Apache error log, optionally from
a start time, and calculate the total cache hits and misses and
calculate the average read savings.
https://pagure.io/freeipa/issue/8798
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rafael Guterres Jeffman <rjeffman@redhat.com>
Insert a class before LDAPClient to cache the return value
of get_entry() and certain exceptions (NotFound and
EmptyResult). The cache uses an OrderedDict for the cases
where a large cache might result an LRU model can be used.
The cache be enabled (default) or disabled using
ldap_cache=True/False.
This cache is per-request so is not expected to grow
particularly large except in the case of a large batch
command.
The key to the cache entry is the dn of the object
being requested.
Any write to or referencing a cached dn is evicted from
the cache.
The set of attributes is somewhat taken into consideration.
"*" does not always match everything being asked for by
a plugin so unless the requested set of attributes is a
direct subset of what is cached it will be re-fetched. Err
on the side of safety.
Despite this rather conserative approach to caching 29%
of queries are saved with ipatests/xmlrpc_tests/*
https://pagure.io/freeipa/issue/8798
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rafael Guterres Jeffman <rjeffman@redhat.com>
'install' was being used in some places.
The context can be used to limit what configuration is
used for a given request so having consistency is
valuable.
This affected the force_schema_updates value in LDAPClient
which looks for api.env.context in ('installer', 'updates')
Related: https://pagure.io/freeipa/issue/8798
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rafael Guterres Jeffman <rjeffman@redhat.com>
This can happen on upgrades from older deployments that lack
an ACME installation and don't meet the minimum requirements
to deploy one automatically.
Also don't consider missing ACME schema a total failure, just
log and skip it.
https://pagure.io/freeipa/issue/8832
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The `hbacrule` module performs a call to `selinuxusermap-find`
during entry deletion. This can be optimized by passing pkey_only=True
to the search, skipping the post-callback function. Passing the full
DN of the hbacrule and detecting it in the selinuxusermap find
also saves one call to hbacrule-show, further reducing the searches.
Related: https://pagure.io/freeipa/issue/8784
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Older augeas does not support new options provided by chrony:
sourcedir /run/chrony-dhcp
ntsdumpdir /var/lib/chrony
and is failing to update /etc/chrony.conf in ipa installer.
Bump augeas-libs version to require the fix:
1.12.0-6 on fedora 33+
1.12.0-3 otherwise
Fixes: https://pagure.io/freeipa/issue/8676
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Including execution time (etime) was added in commit
4d716d3fbc
This is a parser that will collect API executions and
average them by command.
If > 5 requests of the same type then the fastest and slowest
results will be dropped to try to smooth the average.
These averages will be used for two purposes:
1. Identify potential bottlenecks in API performance
2. Provide a baseline so that future performance changes can be
measured.
It is included in contrib because this is not going to be shipped
with a distribution but is useful to have with the code.
A sample execution is:
Successful commands:
Mean user_show: 12234152.5 of 2 executions
Mean command_defaults: 3284363.0 of 3 executions
Mean user_add: 594369554.5 of 2 executions
Exceptions:
Mean user_del: 232540327 ns of 2 executions
The parselog command was successful
Times are in nanoseconds.
https://pagure.io/freeipa/issue/8809
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
This is the experimental fix to workaround the issue with
PKI on Azure Pipelines. PKI is the most sensitive to slow
systems(at least, appropriate delays and timeouts should be
adjusted for such).
Somehow Azure's test envs became slower then they were earlier
(for example, CA subsystem start changed
~(20-30)sec -> ~(45-60)sec). This triggered various issues with
subsystems of PKI in CI.
PKI honors `PKISPAWN_STARTUP_TIMEOUT_SECONDS` env variable (if
not set the timeout is 60sec) for pkispawn tool. The default
timeout is inconsistent with IPA's one (startup_timeout=120sec), which in
particular, is used in ipa-pki-wait-running tool).
Related: https://pagure.io/freeipa/issue/8830
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Francois Cami <fcami@redhat.com>
test_hostname_parameter: Test for issue 2692 ipa-server-install ignores --hostname:
check whether hostname provided in `--hostname` parameter is being taken into account and set as new hostname without prompting for it again
test_ad_subpackage_dependency: Test for issue 4011 ipa-server-install crashes when AD subpackage is not installed:
test if ipa-server installation succeeds without `freeipa-ipa-server-trust-ad` installed
test_backup_of_cs_cfg_should_be_created: Test for issue 4166 Backup CS.cfg before modifying it:
test if ipa-server installer backs up CS.cfg before modifying it
test_installer_wizard_should_prompt_for_DNS: Test for issue 2575 [RFE] Installer wizard should prompt for DNS:
test if installer is asking for DNS setup details if not provided as parameter
Related: https://pagure.io/freeipa/issue/2692
Related: https://pagure.io/freeipa/issue/4011
Related: https://pagure.io/freeipa/issue/4166
Related: https://pagure.io/freeipa/issue/2575
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Two, three times a year PR-CI starts failing because tox tests pull in a
newer version of pylint with new warnings. To reduce breakage this
change constraints pylint (and indirectly astroid) to latest tested
minor version. The constraint should be updated when FreeIPA starts to
support a new Fedora version with more recent pylint.
Related: https://pagure.io/freeipa/issue/8818
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This test is to check if ACME certificate renews upon
reaching expiry
related: https://pagure.io/freeipa/issue/4751
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
ipatests: remove skipif for minimum pki dependency
Remove skipif from ACME tests as required pki version
updated in freeipa.spec file (pki 10.10.5)
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
ipatests: move common code to separate method
Move the code from prepare_acme_client, TestACME/test_certbot_register
and TestACME/test_certbot_certonly_standalone to separate method so
that it can be re-used in TestACMERenew.
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
If multiple notifications of the same type are shown at the same
time, assertion works for only the first one. This change enables to
search for notification's content in all shown notifications.
Fixes: https://pagure.io/freeipa/issue/8641
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Serhii Tsymbaliuk <stsymbal@redhat.com>
The Dogtag KRA backend now uses CryptographyCryptoProvider instead of
NSSCryptoProvider for KRAClient connections. The
CryptographyCryptoProvider uses PyCA cryptography to provide wrapping
and unwrapping. The change will allow Dogtag to remove the
NSSCryptoProvider and drop python-nss as a dependency.
The code in ipaserver.plugins.dogtag creates a Certificate object to
work around a bug in Dogtag. Dogtag supports paths but passes the wrong
type to PyCA cryptography.
Fixes: https://pagure.io/freeipa/issue/8814
See: https://github.com/dogtagpki/pki/issues/3499
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When LDAP server disconnects ipa-otpd client connection due to a
timeout, ipa-otpd instance would stop and report an issue. This confuses
systemd service reporting, so for these situations we better to shut
down gracefully.
Fixes: https://pagure.io/freeipa/issue/6587
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Simo Sorce <ssorce@redhat.com>
This test checks if ipa-cert-fix renews the certs on replica
after cert renewal on master.
related: https://pagure.io/freeipa/issue/7885
ipatests: refactor expire_cert_critical fixture
Defined method to move the date and refactor
expire_cert_critical fixture using it
ipatests: PEP8 fixes
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This utilizes the new status option so that we only retrieve
VALID certificates when revoking certificates issued for a
specific host or service.
ae74d348c3 made a special case
in cert_find when searching for hosts and services so that if only
one host/service was searched on do a subject search. It only
works when there is exactly one option requested.
Since we want to restrict to only VALID certificates, pass
the hostname as subject directly when revoking certs when deleting
an entry to limit the number of certificates to consider.
https://pagure.io/freeipa/issue/7835
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Scenarii:
- idrange-add prevents --auto-private-groups with a local id range
- idrange-mod prevents --auto-private-groups with a local id range
- auto-private-groups accepts only true/false/hybrid/empty values
Related: https://pagure.io/freeipa/issue/8807
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This saves at least one search per user-mod because the current
set of objectclasses are verified to ensure they are complete
on each update.
So always retrieve them in get_either_dn(). They are used by
every call but there is negligible overhead in retrieving
this from LDAP.
https://pagure.io/freeipa/issue/8801
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The pkey of services is krbprincipalname. The host plugin passes
this full value to service_del and service_disable if the service
hostname matches the requested host.
This limits the amount of data and post-processing done
when host_del and host_disable call service_find. It also saves
a presence query for keytab in each service found.
https://pagure.io/freeipa/issue/8787
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
During client installation, the installer calls sudo -V
to check if sudo command is installed. The output must not
be displayed in stdout.
Related: https://pagure.io/freeipa/issue/8767
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
ipa-client-install is checking if the sudo command is available
by calling 'sudo -V'. The call is currently using subprocess.popen
which redirects the output to the default stdout.
Use ipautil.run instead of subprocess.popen as this does not
capture stdout (the command output is just logged in the debug file).
Fixes: https://pagure.io/freeipa/issue/8767
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
systemd-resolved in Fedora 34+ returns SERVFAIL for reverse zone that
does not yet exist when we attempt to look it up before installation.
Assume that this is OK -- we are going to create the zone ourselves
during installation.
Fixes: https://pagure.io/freeipa/issue/8794
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
* move WSGI app code to main code base so it can be used with other
WSGI servers that expect a Python package.
* populate LDAP schema early to speed up first request by ~200ms
* gc.collect() and gc.freeze() to improve memory handling and GC
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The test test_replica_promotion.py::TestRenewalMaster::
test_automatic_renewal_master_transfer_ondelete is calling
ipa-server-install --uninstall directly without performing first
ipa server-del. This can lead to incomplete uninstallation and
test failures.
Call tasks.uninstall_replica instead of tasks.uninstall_master.
This is equivalent to ipa-replica-manage del + uninstall
(ipa-replica-manage del works in DL0 and DL1 and internally calls
ipa server-del in DL1).
Fixes: https://pagure.io/freeipa/issue/8792
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The test TestIpaHealthCheck has been updated with commit
e86ff48 and now needs 1 master, 1 replica and 1 client
in order to execute.
Update the nightly definitions accordingly.
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
openssl-1.1.1i introduced a regression preventing WebUI
login when the server is installed with --no-pkinit option.
On fedora 32/33/34/rawhide openssl-1.1.1k-1 is now available.
On RHEL8, openssl-1.1.1g is still shipped and doesn't have the
issue.
Fixes: https://pagure.io/freeipa/issue/8632
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The testcase checks that when ipa-healthcheck tool is run
on ipa-client it will display warning message "IPA is not configured"
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This test assumes that IPA server is installed and configured.
But test_jsplugins is the subtask of fasttest which is designed
(unittest) to be run in server less environment.
Note: `needs_ipaapi` is not completely suitable because there is
no direct IPA API usage. `xmlrpc_setup` fixture is also not
suitable because it assumes XMLRPC.
Fixes: https://pagure.io/freeipa/issue/8781
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Combining the existence check with the sudoorder handling
allows to reduce the number of searches during a sudorule
modification by removing a call to sudorule-show.
Related: https://pagure.io/freeipa/issue/8780
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Lint task uses PyPI to get the latest Pylint, which may pull in
any other packages as dependencies. For distro isolation the Lint
job should not produce any meaningful artifacts or they should not
be used in subsequent jobs. So, this job have to be isolated from
the others.
Fixes: https://pagure.io/freeipa/issue/8772
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Ubuntu 20.04's docker meets the requirement(4+).
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The security option 'apparmor:unconfined' tells Docker to not
apply AppArmor profiles for containers at all. This will not
replace or remove any existing profile. For example, this happens
on Ubuntu 20.04 which switched to chrony and brings its AppArmor
profile. Container's chronyd get blocked by AppArmor:
fv-az26-252 audit[11304]: AVC apparmor="DENIED" operation="capable" profile="/usr/sbin/chronyd" pid=11304 comm="chronyd" capability=2 capname="dac_read_search"
fv-az26-252 audit[11304]: AVC apparmor="DENIED" operation="capable" profile="/usr/sbin/chronyd" pid=11304 comm="chronyd" capability=1 capname="dac_override"
So, any of AppArmor profiles can block container's processes by
matching executable name. There are two ways:
1) prepare custom AppArmor unconfined profile, load it on Host and
reference it in container's configuration. This requires the
knowledge of profile syntax at least, not to difficult, but
potentially hard to maintain.
2) disable conflicting profile on Host;
Azure will warn about AVC in either case.
The second one was chosen as more simple.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Azure's VM distro is Ubuntu, which has enabled AppArmor.
This security module sometimes interferes with CI Docker containers,
but to be completely disabled it requires reboot(this is impossible,
at least for now). So, Azure will warn about AVC records in Host's
journal as a possible clue.
It will be equally important to be warned about SECCOMP records to
see possible blocked syscalls(requires SCMP_ACT_LOG as defaultAction in
seccomp profile).
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The journal of Host is useful for AVC/SECCOMP analyzing.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The syncing time stuff is required by IPA NTP tests.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The distros may use different sphinx builder paths,
for example, by exporting of SPHINXBUILD env var.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Collect disk usage information may be helpful, for example, for
debugging code required free space such as healthcheck tests.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Some tests require its specific Pytest args. With this change
they can be specified in tests definitions.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The test TestIpaHealthCheck::test_ipahealthcheck_ds_riplugincheck
is expecting a specific message for the RIPluginCheck
but the message has been updated to fix
4656 - Remove problematic language from UI/CLI/lib389
("enable referint on all suppliers" instead of
"enable referint on all masters").
Shorten the expected msg so that it fits both situations.
Fixes: https://pagure.io/freeipa/issue/8779
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
While "ktutil" does require a pseudo-terminal on particular systems to
operate, majority of programs do not need it.
At the same time invoking `ssh` with forced pseudo-terminal allocation
interferes with sessions multiplexing feature and increases connection
time. The increase can be as large as 10 seconds in certain cases which
leads to unexpected EOFs of pexpect utility.
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Distributions may want to run comprehensive fastcheck or lint tasks,
but rpmlint tool is mandatory for these targets while some platforms
don't have it at all.
With this change the rpmlint becomes optional for fastcheck, devcheck
and lint make targets.
Note: rpmlint option is disabled by default.
To enable: ./configure --enable-rpmlint
To explicitly disable: ./configure --disable-rpmlint
Fixes: https://pagure.io/freeipa/issue/8768
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
user-add now returns the `nsaccountlock` parameter as
a boolean instead of as a list of string, meaning tests
have to be adapted to expect the correct type.
Related: https://pagure.io/freeipa/issue/8743
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The `nsaccountlock` attribute was being returned as a
list of string ("TRUE"/"FALSE") instead of a boolean.
Use the convert function used in `user-find` and `user-mod`
for consistency, since these commands return the parameter as a boolean.
Fixes: https://pagure.io/freeipa/issue/8743
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Add test to ensure group creation fails when passing the --nonposix
option and a GID number at the same time. Failure shows a message
to warn the user that this is not allowed.
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Enhance error message when adding non-posix group
with a GID so the user knows that a GID should not
be passed when adding a group with the --nonposix option.
Fixes: https://pagure.io/freeipa/issue/8155
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Add test to ensure that IPA allows to introduce multiple
permitopen and permitlisten entries.
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
On RHEL8, ipa is using named-pkcs11.service but RHEL9 is based on
fedora34 and uses named.service instead. There is already some support
for this distinction in ipaplatform, and the patch relies on the
specific settings that can be configured in ipaplatform/xx/services.py
and ipaplatform/xx/constants.py
On RHEL9 ipa also needs to define NAMED_OPENSSL_ENGINE for named
to use openssl's okcs11 engine.
Fixes: https://pagure.io/freeipa/issue/8753
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
During development of a fix to workaround certmonger effectivly
hanging server uninstallation the test was re-worked to force
uninstall during the test execution itself.
https://pagure.io/freeipa/issue/8506
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Some certificates may have started renewal so returning to
present time can bind the server up with trying to renew.
certmonger fires off helpers when it's time to renew
certificates. This scenario puts the time within the renewal
window. If certmonger notices while the test is running it
will kick off renewal for all 12 certificates.
A lock is used to serialize things. The CA was shut down prior
to changing time so there is no chance of issuing new certs.
A fixture was used to ensure that things restarted when
the test was over. This was for chronyd and the CA. By restarting
the CA we allow the chance that it will be able to do some
work, versus returning a connection error and letting
certmonger just error out (CA_UNREACHABLE).
During uninstallation we call certmonger remove_request over
DBus (the equivalent to stop-tracking). As part of this
certmonger waits for any child (helper) processes to go away.
This used to do it via SIGKILL but that caused other problems
so it was changed to waitpid(). We know that it isn't going to
return for a while because the CA isn't up. DBus has a
hardcoded 25 second timeout. So we're guaranteed to get a
DBus timeout. We *could* try to play with it and change the
timeout, or retry a bunch of times, but it isn't worth the
hassle.
This is a contrived scenario that uninstalls immediately after
tweaking time forward. So rather than trying to make this
succesful, uninstall at the future time with the CA stopped
so that helpers won't be hanging around and certmonger can
remove the certs.
This is the last test so also the last time we need the replica
so to avoid replication bogging things down remove that prior
to executing the test. It's one less moving part during the
uninstall phase.
https://pagure.io/freeipa/issue/8506
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
There are failure observed in test_rekey_keytype_DSA(test_cert.py)
It is due to the fact that there is no guarantee that the request id
will match the filename that certmonger assigns.
This fix assigns the request id with -I option to command (and make
use of existing fixture) and get the file name by grepping the
certmonger's directory with specified req id.
fixes: https://pagure.io/freeipa/issue/8725
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
There may be both IPv4 and IPv6 AJP connectors. Test that both
are upgraded with the new tomcat attribute and the passwords are
kept in sync.
The Apache password will be updated if needed elsewhere in the
upgrade process.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This is needed so we can mock the DogtagInstance class
and have control over the version of tomcat.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
In this patch, we ensure we upgrade all AJP adapters with the same
secret value if any are missing. This ensures that both IPv4 and IPv6
adapters have the same secret value, so whichever httpd connects to
will be in sync. This is consistent with what Dogtag does when
provisioning them.
Notably missing from this patch is handling of multiple unrelated AJP
adapters. In an IPA scenario (and default PKI scenario) this shouldn't
be necessary. However, with external load balancing, this might happen.
This patch benefits IPA in the scenario when:
1. pkispawn runs on an older PKI version (pre-AJP secret, so ~8.2?)
2. pki gets upgraded to 10.10.1 before IPA can provision a secret,
resulting in split IPv4/IPv6 adapters -- this would only happen
on a direct migration from 8.2 -> 8.4
3. ipa upgrade script then runs to provision an AJP secret value for
use with both Dogtag and IPA.
Without this patch, only the first (IPv4) adapter would have a secret
value provisioned in the above scenario.
Signed-off-by: Alexander Scheel <ascheel@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
With this change, certbot test will be running on rhel.
certbot is not avilable on rhel through repository.
Plan is to install certbot using pip/epel on rhel and increase the
test coverage on rhel
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
This is a regression test for the bug in python-kdcproxy mentioned in
https://github.com/latchset/kdcproxy/pull/44
When the reply from AD is split into several TCP packets the kdc
proxy software cannot handle it and returns a false error message
indicating it cannot contact the KDC server.
This could be observed as login failures of AD user on IPA clients
when:
* IPA client was configured to use kdcproxy to communicate with AD
* kdcproxy used TCP to communicate with AD
* response from AD to kdcproxy was split into several packets
This patch also refactors and improves existing tests:
* switch to using pytest fixtures for test setup and cleanup steps to make
them isolated and reusable
* simulate a much more restricted network environment: instead of blocking
single 88 port we now block all outgoing traffic except few essential
ports
* add basic tests for using kdcproxy to communicate between IPA client
and AD DC.
Reviewed-By: Anuja More <amore@redhat.com>
The test test_dnssec.py::TestInstallDNSSECFirst::test_resolvconf
checks that /etc/resolv.conf points to the localhost and
fails on fedora33 because systemd-resolved is in place
(and /etc/resolv.conf contains 127.0.0.53).
The test logic needs to be adapted. When systemd-resolved is
used, the test can check the output of "resolvectl dns".
Fixes: https://pagure.io/freeipa/issue/8695
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
In the test test_sssd.py::TestSSSDWithAdTrust::test_is_user_filtered
the config file sssd.conf is modified with a parameter
filter_users written in the [domain/..] section but
the parameter should appear in [nss] section instead.
Fixes: https://pagure.io/freeipa/issue/8747
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Test was randomly failing if the query for the server role is
executed before the replication had time to replicate the
changes on cn=adtrust agents,cn=sysaccounts,cn=etc,dc=ipa,dc=test,
as the server role is read using this entry.
related: https://pagure.io/freeipa/issue/8553
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
- 389-ds fixes an information disclosure during unsuccessful LDAP BIND
operation, CVE-2020-35518, https://github.com/389ds/389-ds-base/issues/4609
- Dogtag PKI adopted to work with 389-ds with the fix,
https://github.com/dogtagpki/pki/issues/3458
FreeIPA needs to require new Dogtag and 389-ds versions on all Fedora
and RHEL versions.
RHEL 8 version is set to 1.4.3.16-12 which is the official build after
pki-core was fixed to work with the CVE fixes.
In order to avoid excessive %if/%endif conditionals in the spec file, I
have added a short Lua table with 389-ds versions for F32-33. F34 and
Rawhide will fallback to the same newer 389-ds 2.0.3 version. We do not
support building on F31 or older Fedora anymore as they are EOLed
already.
Fixes: https://pagure.io/freeipa/issue/8705
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
TestMaskInstall is a usual integration tests and should
install freeipa server during test run.
"ipaserver" template provides pre-install freeipa server and
is intended for use with webui and xmlrpc tests.
Reviewed-By: Francois Cami <fcami@redhat.com>
When dnssec role is removed from the local node, the uninstaller
creates a new replica key and marks the older replica keys as disabled
(both in the local HSM and in LDAP).
If ipa-ods-exporter runs in the middle of this operation, the old replica
key may be disabled in the local HSM but not yet in LDAP and
ipa-ods-exporter believes that it is a new replica key that needs to be
imported from LDAP to local hsm. The op fails as there is already the key
in the local HSM.
The error can be ignored, ipa-ods-exporter simply needs to log a warning.
Fixes: https://pagure.io/freeipa/issue/8654
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
When a master key is missing from the local HSM, ipa-ods-exporter crashes.
This can happen when the DNSSEC master role is moved from one node to
another with the following scenario:
- install server1 with dns + dnssec
- install server2 without dns
- disable dnssec from server1
- install dns + dnssec on server2
With the above scenario, server2 never had the opportunity to get
the master key (this happens only when the replica is already
configured as DNS server and has put its public replica key in LDAP +
the current DNSSEC master wraps its master key with the replica key).
ipa-ods-exporter can only log an error instead of crashing.
Related: https://pagure.io/freeipa/issue/8654
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
* Update template images to include updated packages
* Bump rawhide to use branched F34 template
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
As of 1.19.1 MIT krb changed the error returned if no valid
credentials could be obtained(GSS_S_CRED_UNAVAIL->GSS_S_NO_CRED).
To be compatible with previous versions of krb the new expected
error message has been added.
Fixes: https://pagure.io/freeipa/issue/8737
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This test creates healthcheck.log file in /var/log/ipa/healthcheck/
directory if its not present and then checks that when sosreport command
is run it collects the healthcheck log file by checking the console log
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
Do not check that host is resolvable.
systemd-resolved creates synthetic records for hosts in /etc/hosts.
If test hosts are listed in /etc/hosts on controller, no A records will
be created.
Related to https://pagure.io/freeipa/issue/8703
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
test_testconfig is using hardcoded hostnames which do not match ones
provided in real test config. This causes resolver factory to fail
when trying to detect resolver type of the host.
Related to https://pagure.io/freeipa/issue/8703
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
systemd-resolved enables positive and negative cache by default which
affects test scenarios where dns records are being created and deleted and
then verified using any tools that utilize default system resolver
(i.e. `dig` or `curl`).
Related to https://pagure.io/freeipa/issue/8703
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Set IPA master as nameserver on replica and client machines during default
installation. This will help to avoid manual configuration in test cases
which require members of IPA domain to be resolvable.
Related to https://pagure.io/freeipa/issue/8703
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Many test scenarios need to configure resolvers on test machines. Most
notable patterns are:
* using IPA master as DNS resolver on clients and replicas
* intentionally breaking name resolution
Now it is done by directly editing /etc/resolv.conf file. While being
simple this approach has following issues:
* NetworkManager restores this file periodically and on specific events
* This is not how users are expected to manage resolvers on modern
systems with NetworkManager and systemd-resolved.
This patch introduces three classes for main types of resolvers management:
* plain file
* NetworkManager
* systemd-resolved
For each resolver manager the native way of configuring of nameserves is
used: direct editing for /etc/resolv.conf or drop-in config files for
NM and resolved.
The type of resolver is automatically detected for each host and an
appropriate instance is added to Host object.
The Resolver class (and it's subclasses) provide convenience functions
for changing nameservers and restoring the original config.
During all operations (backup, modify, restore) it checks that resolver
configuration has not been altered unexpectedly and raises exception if it
was. This helps to detect unexpected changes in resolvers.
Related to https://pagure.io/freeipa/issue/8703
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The usage of the existing gssproxy service(`service/ipa-api`) leads
to undesirable for this case side effects such as auto renew of
expired credentials.
Fixes: https://pagure.io/freeipa/issue/8735
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
If nsds5replicalastupdateend is not yet present,
ipa-replica-manage will backtrace as it tries to retrieve that
attribute unconditionally.
Gracefully handle that situation.
Fixes: https://pagure.io/freeipa/issue/8605
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The test test_commands.py::TestIPACommand::test_ssh_key_connection
is checking the content of the journal using journalctl --since ...
but provides only the time, not the whole date with year-month-day.
As a consequence, if the test is executed around midnight it may
find nothing in the journal because it's looking for logs after 11:50PM,
which is a date in the future.
The fix provides a complete date with year-month-day hours:min:sec.
Fixes: https://pagure.io/freeipa/issue/8728
Reviewed-By: Francois Cami <fcami@redhat.com>
Add prototype to the exported function
Replace few tabs by spaces and mark static code as static.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
On Fedora 33 GCC defaults to -Wmissing-prototypes and emits warnings
about function prototypes missing. If -Werror is specified, this breaks
compilation.
We also default to -Werror=implicit-function-declaration
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
There are few fields in KDB structures that have 'conflicting' types but
need to be compared. They come from MIT Kerberos and we have no choice
here.
In the same way, SID structures have own requirements.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Calling to ipadb_get_connection() will remove LDAP context if any error
happens. This means upper layers must always verify that LDAP context
exists after such calls.
ipadb_get_user_auth() may re-read global configuration and that may fail
and cause IPA context to have NULL LDAP context.
Fixes: https://pagure.io/freeipa/issue/8681
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
'pki-server cert-fix' has a known and expected failure when the DS
certificate is expired. 'ipa-cert-fix' handles this by
optimistically ignoring the CalledProcessError and continuing when
the DS certificate was up for renewal.
This heuristic is a bit too optimistic. If 'pki-server cert-fix'
fails due and returns nonzero due to some other, more serious error
(as has been seen in the wild[1]), 'ipa-cert-fix' continues then
fails later with a more confusing error, for example:
[Errno 2] No such file or directory:
'/etc/pki/pki-tomcat/certs/27-renewed.crt'
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1930586
Improve the heuristic by also checking whether output files
corresponding ot all of the "extra" certificate that we asked
'ipa-cert-fix' to renew, do indeed exist and are X.509 certificates.
Fixes: https://pagure.io/freeipa/issue/8721
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
In order to get the fix for sync_repl, the following versions
are required:
on fedora32: 1.4.3.19-1
on fedora33 and above: 1.4.4.12-1
on rhel 8.4: 1.4.3.16-11
Note: the fix is not available yet on fedora32 as the build has
been marked as obsolete due to a pkispawn regression
(https://github.com/dogtagpki/pki/issues/3458).
The version will need to be updated in a later commit.
Fixes: https://pagure.io/freeipa/issue/8496
Reviewed-By: Francois Cami <fcami@redhat.com>
Moved fixture to use across multiple classes. Added capability
to install the KRA to the fixture
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
Test moves system date to expire certs. Then calls ipa-cert-fix
to renew them. This certs include subsystem, audit-signing,
OCSP signing, Dogtag HTTPS, IPA RA agent, LDAP and KDC certs.
related: https://pagure.io/freeipa/issue/7885
Signed-off-by: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
pylint 2.7.0 now emits inconsistent-return-statements if one of
try/except statement is not returning explicitly while the other do.
Fixes: https://pagure.io/freeipa/issue/8720
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The output from pki-spawn is already displayed to the user
as well as a short traceback so re-displaying the CalledProcess
error provides no value and only provokes confusion,
particularly because it is condensed and includes embedded
newlines.
Re-raise the exception from None so that the traceback is
removed and while there is still an immense traceback from
the admintool class it is significantly shorter than before
and removes:
"During handling of the above exception, another exception occurred"
The handling is in fact expected.
This changes the user-facing installer output from:
[1/28]: configuring certificate server instance
Failed to configure CA instance: CalledProcessError(Command ['/usr/sbin/pkispawn', '-s', 'CA', '-f', '/tmp/tmpr5x2l0gm', '--debug'] returned non-zero exit status 1: 'INFO: Connecting to LDAP server at ldap://ipa.example.test:389\nINFO: Connecting to LDAP server at ldap://ipa.example.test:389\nDEBUG: Installing Maven dependencies: False\nERROR: KeyError: \'CA\'\n File "/usr/lib/python3.9/site-packages/pki/server/pkispawn.py", line 575, in main\n raise KeyError(\'CA\')\n\n')
See the installation logs and the following files/directories for more information:
/var/log/pki/pki-tomcat
[error] RuntimeError: CA configuration failed.
CA configuration failed.
The ipa-server-install command failed. See /var/log/ipaserver-install.log for more information
to
[1/28]: configuring certificate server instance
Failed to configure CA instance
See the installation logs and the following files/directories for more information:
/var/log/pki/pki-tomcat
[error] RuntimeError: CA configuration failed.
CA configuration failed.
The ipa-server-install command failed. See /var/log/ipaserver-install.log for more information
The output is similarly reduced in the installer log. There is no
reason to acknowledge that a CalledProcessError was raised since
the output is already available and it's just an intermediary.
Hopefully this will encourage users to focus on the logs rather than
the malformed traceback.
https://pagure.io/freeipa/issue/8565
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
If the CA fails to deploy then the CRL directory will not exist
but will report an error that it has failed to be removed.
There is no need to try to navigate a directory if it doesn't exist.
Related: https://pagure.io/freeipa/issue/8565
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This file was missed in the first pass to remove support for
csrgen.
This was never feature complete and currently has issues and
we lack the resources to maintain it.
Drop it for now. It can be revived from git history in the
future if we see the need.
https://pagure.io/freeipa/issue/8669
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
If the Apache, 389-ds or KDC certificate was issued by
a third party there is nothing we can do, regardless of
whether it is expired or not.
Report which certificates will not be renewed so the
admin can manually do do (likely in the event of a
third-party certificate).
https://pagure.io/freeipa/issue/8600
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The nickname of the 389-ds certificate was hardcoded as
Server-Cert which failed if the user had installed a
third-party certificate using ipa-server-certinstall.
Instead pull the nickname from the DS configuration and
retrieve it based on that.
https://pagure.io/freeipa/issue/8600
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
With commit ec6698f , the error message has changed from
Unable to communicate with CMS (503)
to
Request failed with status 503: Non-2xx response from CA REST API: 503. (503)
Related: https://pagure.io/freeipa/issue/8704
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Check that sudocmd-add fails when trying to add
a command containing a trailing dot.
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Trailing dots aren't permitted in sudo commands, as
enforced explicitly in `get_dn`. Performing this check
before adding the command prevents the user from
entering invalid commands, which would otherwise trigger
errors when accessing them afterwards.
RN: ipa sudocmd-* commands now validate SUDO command name to not end with a dot.
RN: Previously a trailing dot was stripped away in when addressing a SUDO command's LDAP object.
RN: As a result, a SUDO command was created but it was not possible to refer to it in other IPA commands.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1925410
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ACME uses nonce values to prevent replay attacks. Since the
ipa-ca name can go to any of the IPA servers in order to verify the
nonce the servers need to know the value that was set which
relies on replication. Sometimes the client is faster than
replication so a request can fail.
This change returns the baseURL to the client as the name of the
ACME server during discovery which should pin all requests to this
one IPA server and alleviate the replication issue.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
https://pagure.io/freeipa/issue/8712
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
`update-crypto-policies` tool from RPM package `crypto-policies-scripts`
is required for tests.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
IPA server's AAAA records at embedded DNS mode depend on result of
`get_server_ip_address` function(`ipaserver.install.installutils`),
which in turn, relies on NSS.
In case of Azure Pipelines, there are neither IPv6 records in
'/etc/hosts' nor external DNS, which may provide such. This leads to
the missing AAAA records for master and missing AAAA records for `ipa-ca`
pointing to master in embedded DNS.
In particular, tests `test_ipa_healthcheck_no_errors`,
`test_ipa_dns_systemrecords_check` fail with:
```
[
{
"source": "ipahealthcheck.ipa.idns",
"check": "IPADNSSystemRecordsCheck",
"result": "WARNING",
"uuid": "b979a88a-6373-4990-bc83-ce724e9730b4",
"when": "20210120055054Z",
"duration": "0.032740",
"kw": {
"msg": "Got {count} ipa-ca AAAA records, expected {expected}",
"count": 1,
"expected": 2
}
}
]
```
where `ipa-ca` record exists only for replica.
Note: since the most of the code in setup_containers was touched it has
been reformatted.
Fixes: https://pagure.io/freeipa/issue/8683
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Add test for prune option of ipa-cacert-manage.
After a certificate is installed, a jump in time
is performed to a date where the certificate is expired,
and then it is pruned.
Related: https://pagure.io/freeipa/issue/7404
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
As part of the effort for reducing differences between
upstream and downstream releases, product naming in WebUI
About dialog is changed from FreeIPA to IPA.
Related: https://pagure.io/freeipa/issue/8669
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Add tests for the ipa cert-remove-hold command.
Scenario 1:
add host entry, request cert, revoke cert with "hold" reason, remove hold
Scenario 2:
call ipa cert-move-hold with a non-existent cert ID and ensure that
the exception mentions 'Certificate ID .. not found'
Related: https://pagure.io/freeipa/issue/8704
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ipa cert-show, ipa cert-revoke and ipa cert-remove-hold do not
print meaningful info when called on a non-existent cert id:
Certificate operation cannot be completed: Unable to communicate
with CMS
Propagate the reason from the HTTP message in order to print
'Certificate ID 0x.. not found'
Fixes: https://pagure.io/freeipa/issue/8704
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
system_u:system_r:init_t:s0 cannot execute gpg=agent when SELinux is in
enforcing mode. Use SELinux context that allows this execution:
system_u:system_r:initrc_t:s0 and wrap the whole execution into a bash
run to make sure init_t -> initrc_t transition.
Fixes: https://pagure.io/freeipa/issue/8699
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The UI in RHEL has a different set of logos and different
background colors. Some direct adjustments were made that
are not buildable so apply them as a patch.
https://pagure.io/freeipa/issue/8669
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Some tests are individually stopping a service and call
ipactl status to ensure it is stopped. They need to use
run_command with raiseonerr=False as ipactl status now
exits with 3 when one of the IPA services is down
(since commit 928ab51).
Related: https://pagure.io/freeipa/issue/8588
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
The test is storing the initial file permissions obtained with 'stat',
then modifies them, calls ipa-healthcheck and reverts the permissions
to the original value.
When the file is a symlink, stat returns the permissions of the link,
not of the pointed-to file. But chmod modifies the permissions of the
pointed-to file, not of the link.
As a consequence, the fixture does not properly restore the original
file permissions.
The fix consists in calling 'stat -L' because the command follows
links.
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Handle package removal without taking dependencies into account.
E.g. add frontends for rpm -e --nodeps.
Related: ipatests/pytest_ipa/integration/tasks.py
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
On 10.10+ releases of Dogtag, the PKI installer will not depend
on sudo anymore. This opens the possibility of creating IPA servers
without a properly configured sudo.
In fact, even IPA clients should have sudo and libsss_sudo installed
in most cases, so add a weak dependency on both of them to the client
subpackage.
Also make sure libsss_sudo is installed if sudo is present.
Fixes: https://pagure.io/freeipa/issue/8530
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
When DNS is enabled for a zone, 2 DNSKEYs should be created:
one KSK and one ZSK.
Add a test ensuring that they can be queried on the master and
the replica.
Related: https://pagure.io/freeipa/issue/8647
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The database storing the keys with OpenDNSSEC 2.1 has a
different schema from OpenDNSSEC 1.4, and the keytype
(ZSK, KSK) is stored in a different table column: "role"
instead of "keytype".
With OpenDNSSEC 1.4, keytype can be 256 (ZSK) or 257 (KSK), while
with OpenDNSSEC 2.1, role can be 1 (KSK) or 2 (ZSK).
The schema migration can be seen in opendnssec source code:
enforcer/utils/1.4-2.0_db_convert/sqlite_convert.sql
INSERT INTO hsmKey
SELECT DISTINCT REMOTE.keypairs.id, 1, REMOTE.keypairs.policy_id,
REMOTE.keypairs.HSMkey_id, 2, REMOTE.keypairs.size,
REMOTE.keypairs.algorithm, (~(REMOTE.dnsseckeys.keytype)&1)+1,
CASE WHEN REMOTE.keypairs.generate IS NOT NULL THEN
strftime('%s', REMOTE.keypairs.generate)
ELSE strftime("%s", "now") END,
0,
1, --only RSA supported
REMOTE.securitymodules.name,
0 --assume no backup
FROM REMOTE.keypairs
JOIN REMOTE.dnsseckeys
ON REMOTE.keypairs.id = REMOTE.dnsseckeys.keypair_id
JOIN REMOTE.securitymodules
ON REMOTE.securitymodules.id = REMOTE.keypairs.securitymodule_id;
and the schema for the table is defined in enforcer/src/db/kasp.sqlite:
CREATE TABLE HsmKey (
locator VARCHAR(255) NOT NULL,
candidate_for_sharing TINYINT UNSIGNED DEFAULT 0,
bits INT UNSIGNED DEFAULT 2048,
policy VARCHAR(255) DEFAULT 'default',
algorithm INT UNSIGNED DEFAULT 1,
role VARCHAR(3) DEFAULT 'ZSK',
inception INT UNSIGNED,
isrevoked TINYINT UNSIGNED DEFAULT 0,
key_type VARCHAR(255),
repository VARCHAR(255),
backmeup TINYINT UNSIGNED DEFAULT 0,
backedup TINYINT UNSIGNED DEFAULT 0,
requirebackup TINYINT UNSIGNED DEFAULT 0,
id INTEGER PRIMARY KEY AUTOINCREMENT
);
Fixes: https://pagure.io/freeipa/issue/8647
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The default for this setting is 'auto', which may affect
IPA Samba configuration on future Samba versions. By explicitly
setting this parameter in the template, future manual
intervention is prevented.
Fixes: https://pagure.io/freeipa/issue/8452
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The KRA attempts to unregister itself from the security domain
which requires that IPA be running for this to succeed.
1. Move the KRA uninstall call prior to stopping all IPA
services
2. Try to start IPA if it isn't running and a KRA is configured
It isn't mandatory that IPA be running for the KRA uninstall to
succeed but it will suppress a pretty scary backtrace and error
message.
https://pagure.io/freeipa/issue/8550
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This is an informational message and clutters the installation
screen with no end-user benefit. Logging it as debug is
sufficient to know what is going on.
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Refactoring ipa-rmkeytab with commit
f3f9672d52 led to new error code 7 when
MIT Kerberos fails to iterate through the keys. It appears now in places
where in past error code 3 was returned.
Related: https://pagure.io/freeipa/issue/8658
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
During interactive session of `ipa-adtrust-install` the user needs to
answer several questions. This was done by sending all answers to
the processes stdin without analyzing the questions.
If the installation scenario changes at some point we can get on of the
following results:
* the test fails in the end and the root cause is not obvious
* if a new question was added
* test does not fail but answers are provided for wrong questions -
in this case scope of test case changes without being noticed
If we use `pexpect` for controlling the session, the test will fail
immediately when it encounters unexpected question.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
`ktutil` is a REPL-style utility that can be controlled only interactively.
The common approach of sending commands to stdin does not work with it on
systems where `readline` library has version less then 8.0 due to a bug
in that version.
With `pexpect` we avoid this bug because it emulates the terminal
when interacting with spawned process instead of simply sending all input
to stdin.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The pexpect module can be used for controlling and testing interactive
command-line programs. The wrapper adds testing-oriented features like
logging and automatic process termination and default check for process
exit status.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
[MS-NRPC] section 2.2.1.4.11 requires that the structure
NETLOGON_VALIDATION_SAM_INFO has the same values as defined in the
KERB_VALIDATION_INFO structure from [MS-PAC] section 2.5.
Samba's netr_SamBaseInfo.domain_name corresponds to
KERB_VALIDATION_INFO.LogonDomainName and must be a NetBIOS name of the
domain, not a DNS one.
Failure to provide NetBIOS name here actually breaks
netr_LogonSamLogonEx call issued by IPA-enrolled Samba domain member
which is confused by the returned value:
[2021/01/30 16:36:36.636010, 0, pid=1633, effective(0, 0), real(0, 0), class=winbind]
../../source3/winbindd/winbindd_util.c: 175(add_trusted_domain)
add_trusted_domain: SID [S-1-5-21-3342930694-1632731913-1318603033]
already used by domain [INTERNAL], expected [internal.example.test]
[2021/01/30 16:36:36.636050, 10, pid=1633, effective(0, 0), real(0, 0), class=winbind]
../../source3/winbindd/winbindd_util.c:362(add_trusted_domain_from_auth)
add_trusted_domain_from_auth: Adding domain [internal.example.test]
with sid [S-1-5-21-3342930694-1632731913-1318603033] failed
[2021/01/30 16:36:36.636060, 0, pid=1633, effective(0, 0), real(0, 0), class=winbind]
../../source3/winbindd/winbindd_pam_auth_crap.c:169(winbindd_pam_auth_crap_done)
winbindd_pam_auth_crap_done: add_trusted_domain_from_auth failed
[2021/01/30 16:36:36.636079, 10, pid=1633, effective(0, 0), real(0, 0), class=winbind]
../../source3/winbindd/winbindd.c:814(process_request_done)
process_request_done: [smbd(1650):PAM_AUTH_CRAP]: NT_STATUS_LOGON_FAILURE
Fixes: https://pagure.io/freeipa/issue/8636
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Sumit Bose <sbose@redhat.com>
This reverts commit 69d480003b.
this test will be moved to existing test_uninstallation.py file
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The dns parameter of request_and_wait_for_cert() must be a string of
hostnames.
* Enforce list/tuple type so that API misuse no longer passes silently.
* Add commonNameToSANDefaultImpl to KDCs_PKINIT_Certs profile
* Explicitly pass hostname for service certs
Fixes: https://pagure.io/freeipa/issue/8685
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The string was updated to include the directory the for the database
but this was not reflected in the test and not picked up because
the tests were executed on Fedora 32 which supports dbm so the
test wasn't executed.
https://pagure.io/freeipa/issue/8675
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The original test had some issues:
* it was doing many actions not related to the tested issue which obscured
actual test scenario
* subordinate suffix was hard coded in the test which prevented the test
from checking original issue in case AD domain name did not match this
hard coded value
* Invocation of commands on AD controller was failing in some environments
Other improvements:
* added docstring with test details
* added guard assertions for test preliminary conditions
Related to https://pagure.io/freeipa/issue/8554
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The job definition for fedora-previous/test_installation_client
is missing the .py in the test file:
test_suite: test_integration/test_installation_client
should be instead:
test_suite: test_integration/test_installation_client.py
Reviewed-By: Armando Neto <abiagion@redhat.com>
If NSSDatabase() throws a ValueError it means we can't open it
to look for an existing ipaCert to migrate. Chances are there is
no certificate to migrate at this point in Fedora so don't let
it blow up the entire installation/upgrade. Warn the user and let
them figure it out.
We have no real path forward on this and by proceeding it could
lead to more errors (like no RA) but it is extremely unlikely and
would require a user to upgrade from very old Fedora to very
new Fedora in one step.
https://pagure.io/freeipa/issue/8675
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
If the system lacks DBM support and an older database is
opened then an exception is raised. Include the directory in
the exception so it is clearer which database cannot be opened.
https://pagure.io/freeipa/issue/8675
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
One of the files that are collected after each test is /etc/resolv.conf.
In Fedora 33 this file is actually a symlink. `tar` does not follow
symlinks by default which results in either a broken link in test
artifacts or a symlink pointing to local file on the tests controller
machine.
Fixed by instructing `tar` to resolve the symlinks, so that actual file
pointed by symlink is stored in test artifacts.
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
sshd logs are not displayed immediately in journalctl, this results
in the tests checked the corresponding system logs are racy. I can't
find a way to flush ones. So, the best is the periodical reading of
the system log.
Related: https://pagure.io/freeipa/issue/8682
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
IPA traditionally allowed to add names not found in IPA LDAP to external
attributes. This is used to allow, for example, a local system user or
group be present in a SUDO rule.
With membership validator, we can actually check validity of the names
against both IPA users/groups and users/groups from trusted domains.
If in future we decide to reject a local system's objects, then all it
would take is to switch reject_failures to True.
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Samba 4.13+ in Fedora 33+ and RHEL 8.4+ defaults to Kerberos
authentication. This means user name used for authentication must be
mapped to a target realm.
We have to remove trust on AD side first before removing it locally or
otherwise MIT Kerberos might not be able to locate DCs from AD as
removal of the trust information would cause SSSD to clear the details
for a KDC locator plugin as well.
For the test that modifies AD DNS zone on IPA side to inject unreachable
DCs addresses, the configuration has to be reverted first, to allow
plain 'kinit' during removal of trust to reach AD DCs directly.
Fixes: https://pagure.io/freeipa/issue/8678
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Tests test_integration/test_trust.py::TestTrust::test_sudorules_ad_*
check that a user from a trusted AD domain can perform SUDO
authentication without a password for any command based on a direct user
reference or on indirect AD group reference. The test suite also ensures
an AD user and group can be used for runAsUser/runAsGroup settings.
Due to https://github.com/SSSD/sssd/issues/5475 anything added to
'ipaSudoRunAsExtUserGroup' attribute will be prefixed with '%' and thus
any relying on the value of this attribute displayed by 'sudo -l'
command will fail. The test only validates that a proper group name
appears in the 'sudo' output, so we handle both prefixes in the
corresponding test check. It is not possible to differ by the SSSD
version as a fix to the issue is only a patch on top of 2.4.0 in RHEL.
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Allow specifying AD users and groups from trusted Active Directory
forests in `ipa sudorule-add/remove-runasuser/runasgroup` family of
commands.
IPA provides 'ipasudorunasextuser' and 'ipasudorunasextusergroup' LDAP
attributes to record 'external' objects referenced in SUDO rules for
specifying the target user and group to run the commands allowed in the
SUDO rule.
Use member type validators to 'ipa sudorule-add/remove-runasuser/runasgroup'
family of commands and rely on member type validators from 'idviews'
plugin to resolve trusted objects.
Referencing fully qualified names for users and groups from trusted
Active Directory domains in IPA SUDOERs schema attributes is supported
in SSSD 2.4 or later.
RN: IPA now supports users and groups from trusted Active Directory
RN: domains in SUDO rules to specify runAsUser/runAsGroup properties
RN: without an intermediate non-POSIX group membership
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Allow specifying AD users and groups from trusted Active Directory
forests in `ipa sudorule-add/remove-user` family of commands.
SSSD uses single attribute 'externalUser' for IPA to pull 'external'
objects referenced in SUDO rules. This means both users and groups are
represented within the same attribute, with groups prefixed with '%',
as described in sudoers(5) man page.
Add member type validators to 'ipa sudorule-add/remove-user' family
commands and rely on member type validators from 'idviews' plugin to
resolve trusted objects.
Referencing fully qualified names for users and groups from trusted
Active Directory domains in 'externalUser' attribute of SUDO rules is
supported in SSSD 2.4 or later.
RN: IPA now supports adding users and groups from trusted Active
RN: Directory domains in SUDO rules without an intermediate non-POSIX
RN: group membership
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Register extended validator for users from trusted domains to be called
through add_external_pre_callback() in sudorules and other plugins.
The callbacks allow to validate user names as following:
- if user name passes basic user name validator it is accepted, otherwise
- if user name can be resolved to any user in IPA or in a trusted
domain, it is accepted
- otherwise the name is rejected
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
It was possible to add external members without any validation. Any
object that was not found in IPA LDAP was considered an external object
and a command such as sudorule could have added it to the list of values
for externalUser attribute.
With member validator support, real external members from trusted
domains can be differentiated from the objects that were not found in
IPA and in trusted domains.
Use information from the ID Views plugin to treat external objects
accordingly. Not found objects will be part of the error messaging
instead.
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
baseldap.py:add_external_pre_callback() allows to redefine validators
used to validate member names. Originally this was done to allow
hostname validation and reused default validators associated with other
parameter types.
Provide extension of the validator callbacks to allow fine grained
validation strategy. This is helpful in case we want to apply an
alternative validation strategy in case default validator fails.
New validators can be added to 'member_validator' registry in a similar
way to how API objects are registered:
from .baseldap import member_validator
@member_validator(membertype='foo')
def my_new_validator(ldap, dn, keys, options, value):
<validate value here>
Arguments passed to the validator are arguments passed to the
add_external_pre_callback() augmented with the value to validate.
Fixes: https://pagure.io/freeipa/issue/3226
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Signed-off-by: Peter Keresztes Schmidt <carbenium@outlook.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
With ipa-healthcheck 0.8, the test ipahealthcheck.ipa.files is able
to return a list of possible owners/groups as a comma-separated string
instead of a single owner/group (see commit 930ec5f).
The test output needs to be fixed accordingly.
Fixes: https://pagure.io/freeipa/issue/8662
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Operations in FIPS mode make impossible use of NTLMSSP when
authenticating to trusted Active Directory domain controllers because
RC4 cipher is not allowed. Instead, Kerberos authentication have to be
used. We switched to enforce Kerberos authentication when communicating
with trusted domains' domain controllers everywhere.
Kerberos library uses system wide configuration which in IPA defaults to
resolving location of KDCs via DNS SRV records. Once trust is
established, SSSD will populate a list of closest DCs and provide them
through the KDC locator plugin. But at the time the trust is established
performing DNS SRV-based discovery of Kerberos KDCs might fail due to
multiple reasons. It might also succeed but point to a DC that doesn't
know about the account we have to use to establish trust.
One edge case is when DNS SRV record points to an unreachable DC,
whether due to a firewall or a network topology limitations. In such
case an administrator would pass --server <server> option to
'ipa trust-add' or 'ipa trust-fetch-domains' commands.
'ipa trust-fetch-domains' runs a helper via oddjobd. This helper was
already modified to support --server option and generated custom
krb5.conf overlay to pin to a specific AD DC. However, this
configuration was removed as soon as we finished talking to AD DCs.
With switch to always use Kebreros to authenticate in retrieval of the
topology information, we have to use the overlay everywhere as well.
Convert the code that generated the overlay file into a context that
generates the overlay and sets environment. Reuse it in other
trust-related places where this matters.
Oddjob helper runs as root and can write to /run/ipa for the krb5.conf
overlay.
Server side of 'ipa trust-add' code calls into ipaserver/dcerpc.py and
runs under ipaapi so can only write to /tmp. Since it is a part of the
Apache instance, it uses private /tmp mounted on tmpfs.
Fixes: https://pagure.io/freeipa/issue/8664
Related: https://pagure.io/freeipa/issue/8655
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Changes in https://pagure.io/freeipa/issue/8655 made it impossible
to use AD admin name without domain part in "ipa trust-add" command to
establish external trust with an AD tree domain.
Also use fully qualified admin name by default in all trust related tests
to reduce abiguity
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
winbindd expects forest topology information blob to find out child
domains. We should store it in LDAP and let ipasam to retrieve it.
In fact, ipasam already supports updating and loading this information
but during 'ipa trust-fetch-domains' we didn't provide it.
Make sure the blob is preserved after it was retrieved and also updated
when we fetch forest topology information.
Fixes: https://pagure.io/freeipa/issue/8576
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
[MS-ADTS] 6.1.6.7.13 defines 'trustPartner' attribute as containing a
FQDN of the trusted domain. In practice, for a subdomain of a forest, it
would be FQDN of the subdomain itself in the trusted domain entry in the
parent domain. This is reflected as ipaNTTrustPartner attribute in
FreeIPA.
Remove ipaNTTrustPartner from the searches that use NetBIOS name. We
match cn of that entry already.
Use RDN value of the entry to derive DNS domain name in case
ipaNTTrustPartner is missing.
For subdomains, set trust attributes to 0 and trust flags to mark them
as being within the forest. This will trigger winbindd to not ask for
credentials to reach those domain controllers directly.
Fixes: https://pagure.io/freeipa/issue/8576
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The context is hanging off a talloc memory context passed into the
function so it will eventually be freed. It is better, though, to free
it immediately when we exit from the fill_pdb_trusted_domain() function.
Related: https://pagure.io/freeipa/issue/8576
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
lookup_name() in Samba may call PASSDB API to search by a UPN (e.g.
username@suffix). Support this call by detecting '@' in the passed name
and setting up filter to be
(&(objectClass=ipaNTUserAttrs)(objectClass=krbPrincipalAux)(krbPrincipalName:caseIgnoreIA5Match:=%s))
instead of
(&(objectClass=ipaNTUserAttrs)(uid=%s))
The result of the search would still contain a proper user entry as we
always have krbPrincipalName in LDAP entries of IPA users. Note that the
match must be case-insensitive because otherwise krbPrincipalName is
matched with exact case in the schema. We use the same matching override
in KDB driver already.
Fixes: https://pagure.io/freeipa/issue/8661
Signed-of-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
ipasam already implemented retrieval of groups for MS-SAMR calls.
However, it did not have implementation of a group retrieval for the
path of lookup_name() function in Samba. The lookup_name() is used in
many places in smbd and winbindd.
With this change it will be possible to resolve IPA groups in Windows UI
(Security tab) and console (net localgroup ...). When Global Catalog
service is enabled, it will be possible to search for those groups as
well.
In Active Directory, security groups can be domain, domain local, local
and so on. In IPA, only domain groups exposed through ipasam because
SID generation plugin only supports adding SIDs to POSIX groups and
users. Thus, non-POSIX groups are not going to have SIDs associated and
will not be visible in both UNIX and Windows environments.
Group retrieval in Samba is implemented as a mapping between NT and
POSIX groups. IPA doesn't have explicit mapping tables. Instead, any
POSIX group in IPA that has a SID associated with it is considered a
domain group for Samba.
Finally, additional ACI is required to ensure attributes looked up by
ipasam are always readable by the trust agents.
Fixes: https://pagure.io/freeipa/issue/8660
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When MS-PAC structure is created, we get passed the time of
authentication from KDC. Use this to record logon time in MS-PAC
structure.
Set allow password change time to the last password change. We need to
refer to the actual password policy here in future.
Also use INT64_MAX to represent the resulting value for logoff
and kickoff times according to MS-PAC 2.6.
Fixes: https://pagure.io/freeipa/issue/8659
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
- Build is using a prio of 100 while tests use 50, use consistent
values
- fix the requires for test_installation_client
Reviewed-By: Armando Neto <abiagion@redhat.com>
df uses 1024 bytes as its default display value, but this can be
tweaked by environment variables or a CLI knob.
Force the output unit to 1024 bytes using the CLI and parse it
accordingly.
Fixes: https://pagure.io/freeipa/issue/8674
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
This was never feature complete and currently has issues and
we lack the resources to maintain it.
Drop it for now. It can be revived from git history in the
future if we see the need.
https://pagure.io/freeipa/issue/8669
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
In order to simplify the build process between upstream FreeIPA
and downstream builds (such as CentOS Stream) we are changing
some file references from FreeIPA to IPA (and Identity Management).
https://pagure.io/freeipa/issue/8669
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
389ds is combining the value set in dse.ldif and the current crypto
policy to evaluate the min TLS version that it will be using.
The test needs to change the crypto policy to LEGACY in order to allow
TLS 1.0, because the DEFAULT policy prevents TLS 1.0 on fc33+.
Fixes: https://pagure.io/freeipa/issue/8670
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Kaleemullah Siddiqui <ksiddiqu@redhat.com>
- change log level for replication debugging
According to the docs:
```
default level of logging(16384) used for critical errors and other
messages that are always written to the error log. Messages at this
level are always included in the error log, regardless of the log
level setting.
```
- always flush the access logs to filesystem
During the testing access logs may be written with delay, this
results in logs are not collected by this test node, but for example,
the next one.
- as of now, the changes on `cn=config` are made after the installation
of server or replica. If an error occurs during these stages, then the
actual log level will be the default and not as expected.
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Checking that IPA cert has been added to trust
store is needed to verify that installation
of the server is correct. This cert should also
be removed on uninstall to prevent failures
on further installations.
Related: https://pagure.io/freeipa/issue/8614
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Test that IPA certs are removed after the server uninstall
process. This is needed since if these certs are not
removed from the system store, further installations
will fail.
Related: https://pagure.io/freeipa/issue/8614
Signed-off-by: Antonio Torres <antorres@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
389ds does not return any more additional information
about a failing bind (to avoid leaking information).
As a consequence, when ipa ping is executed with an AD user
the error message contains less info as in the past and needs to be fixed.
Fixes: https://pagure.io/freeipa/issue/8668
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Sergey Orlov <sorlov@redhat.com>
The return value of functions managing the cursor in the keytab
were not checked or reported in a consistent way. This should
assure a reasonable error message in case something goes wrong.
https://pagure.io/freeipa/issue/8658
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This makes it clearer what the return value means.
Replace closing of keytab based on the numeric return value
and do it based on whether the keytab was opened at all.
https://pagure.io/freeipa/issue/8658
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This reverts commit 81c97bb992.
This is to make IPA installable again with older versions of dogtag
so it will install on CentOS 8 Stream.
ACME will not be deployed but on upgrade, if pki 10.10.x is available
then it will be.
https://pagure.io/freeipa/issue/8634
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
It is already confirmed that item_data is not NULL so there
is no need to check it again.
Discovered by coverity.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
ipadb_get_principal() allocates client_actual. Call
ipadb_free_principal to release it.
Rather than spreading the free() amongst the code introduce
done as a target to match behavior in similar functions.
Discovered by coverity.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The purpose is to set a default if the RID doesn't match
expectations.
Discovered by coverity
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Modify the test scenario in order to be independant from PKI
behavior. The aim of the test is to ensure that the KRA
detection is not based on the presence of the directory
/var/lib/pki/pki-tomcat/kra/.
Previously the test was calling ipa-server-upgrade but this cmd
may fail even with the kra detection fix because of an issue in
pki (https://github.com/dogtagpki/pki/issues/3397).
Instead of exercising the whole ipa-server-upgrade command, the
test now checks the output of the API kra.is_installed() to validate
KRA detection mechanism.
Fixes: https://pagure.io/freeipa/issue/8653
Related: https://pagure.io/freeipa/issue/8596
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
We want to always use SMB encryption if it is possible on LSA pipe as we
are going to pass what accounts to a plain-text content within
CreateTrustedDomainEx2 call.
The catch is that older Samba version might not have a way to enforce
this and we need fall back to work with existing connection then.
Fixes: https://pagure.io/freeipa/issue/8655
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
In FIPS mode we cannot rely on NTLMSSP at all, so we have ensure
Kerberos is used by Samba Python libraries. This is achieved by
requiring credentials objects to always use Kerberos authentication.
Additionally, we have to normalize the principal used to authenticate.
In case it was passed without realm, add forest root domain as a realm.
In case it was passed with NetBIOS domain name, remove it and replace
with a realm. Since we only know about the forest root domain as a
realm, require that for other domains' users a real Kerberos principal
is specified.
Fixes: https://pagure.io/freeipa/issue/8655
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When establishing trust to Active Directory forest, RC4 is used to
encrypt trusted domain object credentials as an application-specific
material in a secure channel based on AES session key.
In FIPS mode it is not possible to use RC4 directly.
Samba 4.14 and backports to 4.13 in Fedora 33+ and RHEL 8.4+ now
provide a helper that wraps LSA RPC call CreateTrustedDomainEx2.
This helper ensures that in FIPS mode we first check that LSA session
key is AES before allowing RC4 use internally in Samba bindings. Thus,
it becomes possible to establish trust to Active Directory forest in
FIPS mode.
Adopt FreeIPA code to use the helper provided by Samba when it is
available. If neither the helper nor unprotected arcfour_encrypt utility
is available from Samba bindings, fail import of the ipaserver.dcerpc
module.
Fixes: https://pagure.io/freeipa/issue/8655
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Per the ssh_config(5) man page under ProxyCommand:
"The command string extends to the end of the line, and is
executed using the user's shell ‘exec’ directive to avoid a
lingering shell process."
<shell> -c <proxy command>
Some older versions of nologin (RHEL/CentOS) do not support
the -c option so will still fail but since nologin doesn't
actually execute properly it doesn't include the output
'This account is currently not available' so don't assert
in that case. The returncode of 1 is sufficient to know
that the login is denied.
https://pagure.io/freeipa/issue/7676
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
On a failed permission update if the generated ACI is
invalid then the updated permission is rolled back.
Add the virtual relationship attributes to list of attributes
to be ignored when rolling back the entry.
This relies on the current order in the LDAPObject
relationships field where member and memberof are the first
two values.
https://pagure.io/freeipa/issue/8646
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Gating tests are still on Fedora 32 because some tests are failing.
Rest is being updated to unblock nightly runs:
- "previous" updated to Fedora 32
- "latest" updated to Fedora 33
- 389ds, testing and pki definitions updated to Fedora 33
Signed-off-by: Armando Neto <abiagion@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Upstream dnspython 2.1.0 introduced additional error checking
on SRV values and now rejects invalid priorities.
Remove the sorting test for priority of -1.
https://pagure.io/freeipa/issue/8650
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
In selinux enforcing mode, the command ipa trust-add fails
to establish a one-way trust, during the step fetching the remote
domains.
This step calls a script over DBus and oddjob, that is executed
with oddjob_t context. The policy must allow noatsecure.
Currently the optional_policy is defined in selinux-policy
repo but is ineffective as ipa_helper_noatsecure is not defined
in this repo. When the optional_policy is defined in our own
module, it is taken into account and ipa trust-add succeeds.
Fixes: https://pagure.io/freeipa/issue/8508
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
When the CSR for an expired cert is not found in
/etc/pki/pki-tomcat/{ca|kra}/CS.cfg, ipa-cert-fix fails to
renew the certificate and repair the installation.
The CSR can be found using certmonger as it is stored in
/var/lib/certmonger/requests/<ID> in the "csr" attribute.
Prior to calling pki-server cert-fix, make sure that the
CSR is present in CS.cfg, or update CS.cfg with the content
found using certmonger.
Fixes: https://pagure.io/freeipa/issue/8618
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Add a new test for ipa-cert-fix issue 8618. When the CSR for one
of the certs to be renewed is missing from /etc/pki/pki-tomcat/{ca|kra}/CS.cfg
ipa-cert-fix fails to renew the certificates.
Test scenario:
move the date in the future to expire PKI system certificates (+3 years)
delete the directive ca.sslserver.certreq from CS.cfg
call ipa-cert-fix and ensure that the CSR was found
Related: https://pagure.io/freeipa/issue/8618
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The CA tracking request was modified to drop the profile which
was caught by ipa-healthcheck. Run ipa-certupdate then
ipa-healthcheck to confirm that no problems are introduced.
https://pagure.io/freeipa/issue/8644
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The CA tracking request is modified (it calls renew but it doesn't
actually do a renewal) as part of ipa-certupdate and it dropped
the profile. ipa-healthcheck discovered this condition.
https://pagure.io/freeipa/issue/8644
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
dns_lookup was set to false during ipa client installation which prevented
searches for SRV records for Kerberos servers.
Since https://pagure.io/freeipa/issue/6523 is fixed, dns_lookup is always True
now and the fixture is not needed anymore.
Reviewed-By: Francois Cami <fcami@redhat.com>
The tasks module provides a method to clear sssd cache,
but the method does not remove the file /var/lib/sss/mc/initgroups.
Update the method to also remove this file.
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
TestIpaHealthCLI::test_input_file test was skipped due to
bz1866558. Removed the below statement so that the test can
now run as the bug is fixed.
@pytest.mark.xfail(reason='BZ 1866558', strict=False)
Also changed the assert statement to search text in
stdout_text rather than sdterr_text
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This testcase checks when trust between IPA-AD is established
successfully, IPATrustControllerPrincipalCheck displays
result as SUCCESS
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
When using a fixture, we get a temporary directory created and then
removed by pytest. Pytest uses `shutil.rmtree` call which collects all
files in the directory being removed and then removes them one by one.
At the point of removal of our GNUPGHOME directory, gpg daemon is being
shut down and there might still be an agent UNIX domain socket. The
removal actually overlaps in time with shut down of the gpg daemon, thus
causing `shutil.rmtree()` to fail when an agent UNIX domain socket is
removed by the daemon.
Change the way how we run the gpg agent to use a temporary systemd
service. Stop the service in the finalizer method so that systemd would
send SIGTERM signal and the gpg agent would clean itself up.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
389ds does not return any more additional information
about a failing bind (to avoid leaking information).
As a consequence, when ipa-nis-manage is provided a
wrong password, the error message contains less info
as in the past and needs to be fixed.
Fixes: https://pagure.io/freeipa/issue/8631
Reviewed-By: Kaleemullah Siddiqui <ksiddiqu@redhat.com>
test_acme is not stable and often needs to be
launched multiple times. Remove the test from gating
until the issue is fixed
Related: https://pagure.io/freeipa/issue/8602
Reviewed-By: Kaleemullah Siddiqui <ksiddiqu@redhat.com>
Move actual resolution of UID/GID values for 'ods' entities to the code
that needs them. This prevents failures when uninstalling IPA server set
up without DNS feature. In particular, 'ods' user and group are created
when 'opendnssec' package is installed and if 'opendnssec' package is
not installed, uninstall fails in OpenDNSSEC Exporter instance
constructor.
We use common pattern of checking the service during uninstall:
svc = SVCClass()
if svc.is_configured():
svc.uninstall()
Thus, service class constructor must not do UID/GID resolution
Fixes: https://pagure.io/freeipa/issue/8630
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Move actual resolution of UID/GID values for 'ods' and 'named' entities
to the code that needs them. This prevents failures when uninstalling
IPA server set up without DNS feature. In particular, 'named' group is
created when 'bind' package is installed and if 'bind' package is not
installed, uninstall fails in OpenDNSSEC instance constructor.
We use common pattern for all services during uninstall:
svc = SVCClass(..)
if svc.is_configured()
svc.uninstall()
This requires that the class constructor should not rely on artifacts
that only exist when the service is configured.
Fixes: https://pagure.io/freeipa/issue/8630
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Move actual resolution of UID/GID values for 'ods' and 'named' entities
to the code that needs them. This prevents failures when uninstalling
IPA server set up without DNS feature. In particular, 'named' group is
created when 'bind' package is installed and if 'bind' package is not
installed, uninstall fails in OpenDNSSEC instance constructor.
We use common pattern for all services during uninstall:
svc = SVCClass(..)
if svc.is_configured()
svc.uninstall()
This requires that the class constructor should not rely on artifacts
that only exist when the service is configured.
Fixes: https://pagure.io/freeipa/issue/8630
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
User name for trail space key was using the name for lead space key.
As a result, when both tests were transformed, second one was
unsuccessful as the original user was already created.
Fix the user name data according to the test.
Fixes: https://pagure.io/freeipa/issue/8629
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Only checking for ID to equal "rhel" causes build failures on CentOS
Stream. Instead check both ID and ID_LIKE. This should also work later
on when rebuilds like CentOS Linux get this update.
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
- Create /var/named/dyndb-ldap/ipa/master/ early
- Assume that /var/named/dyndb-ldap/ipa/master/ exists in BINDMgr.sync()
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
'gpg2' is a convenience symlink on Debian, provided by a package that
will go away eventually. 'gpg' is available everywhere.
Signed-off-by: Timo Aaltonen <tjaalton@debian.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
- order permissions alphabeticaly
- do not use semicollon after interfaces
- gen_require should only be used in interfaces
-- to resolve this issue, corresponding changes have to be made in
distribution policy instead of ipa module - disabling check
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Add test to ensure that strings with trailing or leading
whitespaces are allowed as valid passwords.
Signed-off-by: Antonio Torres Moríñigo <atorresm@protonmail.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Previously, dns_lookup_kdc was only set to True if DNS
discovery worked or if the KDC was not specified on the
command-line.
Setting dns_lookup_kdc to False would result in a hardcoded
configuration which is less reliable in the long run.
For instance, adding a trust to an Active Directory forest
after clients are enrolled would result in clients not being
able to authenticate AD users. Recycling FreeIPA servers
could prove problematic if the original hostnames are not
reused too.
Change summary:
Always set dns_lookup_kdc to True on client enrollment.
With this change, DNS SRV search will always be performed
before looking into /etc/krb5.conf realm entries.
Fixes: https://pagure.io/freeipa/issue/6523
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Previously, dns_lookup_kdc was only set to True if DNS
discovery worked or if the KDC was not specified on the
command-line.
Make sure dns_lookup_kdc is always set to true.
Fixes: https://pagure.io/freeipa/issue/6523
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Convert configuredService to either enabledService or hiddenService
depending on the state of the server role. This is to fix situations
when deployment has happened before introduction of hidden replicas
as those services will stay as configuredService and will not get
started after upgrade, rendering the system non-functioning.
Fixes: https://pagure.io/freeipa/issue/8623
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
In case we've got a principal name as '*', we don't need to specify
the principal itself, use pre-defined filter for a wild-card search.
Previously, we had to escape the '*' as specifying it with an explicit
matching rule would have violated RFC 4515 section 3. However, since we
don't really need to specify a different matching rule for a wild-card
search, we can remove this part completely.
Use this change as an opportunity to simplify the code and reduce
number of duplicated filter constants -- if extra filter is NULL, we can
simply pass "" and use _EXTRA filter constants to format the final
filter.
Fixes: https://pagure.io/freeipa/issue/8624
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Robbie Harwood <rharwood@redhat.com>
Currently the TestIpaHealthCheckWithADtrust trust required
only one root AD Domain for testing.
Replaced the existing topology with adroot_adchild_adtree_master_1client
so that trust tests can be run with child/tree root AD domains.
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This testcase checks that when external trust is configured
between IPA and AD subdomain, IPATrustDomainsCheck
doesnot display ERROR
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Python code does detection of the system encoding based on the locale
settings. On RHEL 8.4 development images we somehow get LANG=en_US which
defaults to iso8859-1 _inside_ the systemd-started service, even though
the whole environment defaults to LANG=en_US.UTF-8.
When instrumented with ExecStartPre=/usr/bin/locale, the following
output can be seen:
locale[45481]: LANG=en_US
locale[45481]: LC_CTYPE="en_US"
locale[45481]: LC_NUMERIC="en_US"
locale[45481]: LC_TIME="en_US"
locale[45481]: LC_COLLATE="en_US"
locale[45481]: LC_MONETARY="en_US"
locale[45481]: LC_MESSAGES="en_US"
locale[45481]: LC_PAPER="en_US"
locale[45481]: LC_NAME="en_US"
locale[45481]: LC_ADDRESS="en_US"
locale[45481]: LC_TELEPHONE="en_US"
locale[45481]: LC_MEASUREMENT="en_US"
locale[45481]: LC_IDENTIFICATION="en_US"
locale[45481]: LC_ALL=
ipactl[45483]: Unexpected error
ipactl[45483]: SystemEncodingError: System encoding must be UTF-8, 'iso8859-1' is not supported. Set LC_ALL="C.UTF-8", or LC_ALL="" and LC_CTYPE="C.UTF-8".
systemd[1]: ipa.service: Main process exited, code=exited, status=1/FAILURE
Set the environment to explicit LC_ALL=C.UTF-8 to please the Python
code. FreeIPA server side only cares about actual encoding, not the
language itself. We already use LC_ALL=C.UTF-8 in httpd service snippet.
Fixes: https://pagure.io/freeipa/issue/8617
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Thomas Woerner <twoerner@redhat.com>
With commit c6644b8566 we default to
create unique credential caches in /run/ipa/ccaches for every client
that connects to IPA with a new session. On F34, mod_auth_gssapi process
running as 'apache' cannot create the ccache in /run/ipa/ccaches because
it has no access rights.
The core of the problem is that we have two different paths to obtaining
a ccache: one where 'apache' running httpd process creates it directly
and one where an internal redirect from 'ipaapi' running httpd process
is happening.
Use SUID and SGID to 'ipaapi'/'ipaapi' and allow 'apache' group to write
to '/run/ipa/ccaches'. This fixes the problem.
Note that we cannot completely remove 'GssapiDelegCcachePerms'. If we'd
do so, mod_auth_gssapi will do redirects and fail.
Fixes: https://pagure.io/freeipa/issue/8613
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
The test test_user_plugin is using a hardcoded date for
password expiration and started failed since we passed this date.
Replace the hardcoded date with now + 1 year.
Fixes: https://pagure.io/freeipa/issue/8616
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
Add a new upgrade test. Scenario:
- create an empty /var/lib/pki/pki-tomcat/kra directory
- call ipa-server-upgrade
With issue 8596, the upgrade fails because it assumes KRA is
installed. With the fix, ipa-server-upgrade completes successfully.
Related: https://pagure.io/freeipa/issue/8596
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The dogtaginstance.is_installed() method currently relies on
the presence of the directory /var/lib/pki/pki-tomcat/{ca|kra},
even if it is empty.
An unwanted consequence is ipa-server-upgrade wrongly assuming the KRA
is installed and crashing when trying to upgrade a not-installed
component.
The fix relies on the command "pki-server subsystem-show {ca|kra}" to
detect if a subsystem is installed. The command does not require PKI
to be running (hence can be called anytime) and is delivered by
the pki-server package which is already required by ipa server pkg.
Fixes: https://pagure.io/freeipa/issue/8596
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
freeipa-selinux subpackage is used by both client and server but
requires freeipa-server subpackage unconditionally. This needs to be
removed.
Originally, upstream spec file did not have this bug. It was brought
in with unification of the specfiles.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1883005
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
mod_md requires its own SELinux policy which is only available
in the upstream github. It is beyond the scope of this test to
maintain SELinux policy only for the scenario so skip it
if the client is in enforcing.
Note that no check needs to be done on OS because that is
already handled by the outer skipif for skip_mod_md_tests.
https://pagure.io/freeipa/issue/8514
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The associated service doesn't need to be enabled. Enabling the
timer is suffient for it to execute. It requires the timers
service so will be ready automatically to run on the configured
period.
https://pagure.io/freeipa/issue/8589
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
- Force wipe all existing ccaches
- Set the ticket policy to a short value, 30 seconds.
- Do a series of kinit, ipa command, kdestroy to generate ccaches
- sleep(30)
- Run the sweeper
- Verify that all ccaches are gone
https://pagure.io/freeipa/issue/8589
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
Rather than having a shared ccache per user, configure
mod_auth_gssapi to create a unique one. This requires cleanup
to remove expired caches. A new script is added,
ipa-ccache-sweeper to do this. It will be invoked by a
new service, ipa-ccache-sweep, which will be executed every
12 hours by an equally-named timer.
https://pagure.io/freeipa/issue/8589
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Francois Cami <fcami@redhat.com>
In order to reduce maintenance burden and to be able to use automatic
build tools, bring up the differences between RPM spec files in
upstream, RHEL, and Fedora to a minimum.
This gives us an opportunity to:
- start using proper conditional macros (%bcond_with/%bcond_without)
- remove old cruft where Fedora 31+ and RHEL8+ are already the same
- remove Group lines which already deprecated in Fedora packaging
policy
- remove buildroot cleanup
- support release candidate designations: mostly affects downstreams but
it is better to have macro support in the common spec file.
There is also a special handling of the %SOURCE1 (detached tarball
signature). In developer builds we wouldn't have the signature generated
but RPM needs all files mentioned as sources and patches to exist. The
solution is to filter out detached signature if the final component of
the IPA_VERSION starts with 'dev'. This should cover both in-source
builds (also used in Azure CI and COPR) and PR CI.
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Thomas Woerner <twoerner@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
The test test_subordinate_suffix is failing when configuring the DNS
for the trust, because the dnsforwardzone already exists. It was
configured during the previous test for nonposix trust.
At the end of the tests for nonposix trust, unconfigure the DNS
and the trust before calling the subordinate_suffix test, and add
a test cleaning up subordinate_suffix test.
Fixes: https://pagure.io/freeipa/issue/8601
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There is still some exposure to killing in a bad place. This was
reproduced by killing the process in the parser.parse() call within
__restore_config (line 230) so the values were restored from the
backup but the new dse.ldif never written or copied. But the values
had already been restored from the state file.
I'm not sure this can ever be 100% bullet-proof since it can be
externally killed but if rather than calling restore_state() on the
values in __restore_config we use get_state() which will peek at the
values in the state file without removing them. Then the last step
is to pop upgrade-in-progress and then the rest.
If the values have been restored and the new ldif written and copied
then it's only upgrade-in-progress that really matters. The rest will
be overwritten.
https://pagure.io/freeipa/issue/7534
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
On upgrading a server without ACME to one with ACME
the RA Agent DN needs to be added as a member of the
ACME Enterprise Users group. This was previously
done as part of the creation of that entry.
So on upgrade the RA Agent wouldn't be a member so
ipa-acme-manage didn't have access to operate against
the CA REST API.
In order to add the RA Agent to this group during installation
the ACME provisioning has to come after that step so it is
moved from the middle of an installation to the end and
the group addition moved into the setup_acme() method.
https://pagure.io/freeipa/issue/8603
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
This method was added temporarily while the required packages
were still under development and not available in stable
repositories.
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
Any request other than the FQDN is redirected with a permanent
move (301). Allowing ipa-ca as a valid name saves a round-trip.
This is only allowed on /ca, /kra, /pki, /acme and /ipa/crl.
https://pagure.io/freeipa/issue/8595
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The KRA was previously configured with Internal CA profiles
which did not work with the IPA RA.
Use public, common profiles to manage renewal of the KRA
subsystem certificates.
https://pagure.io/freeipa/issue/8545
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Internal profiles were assigned which prevented rewewals.
dogtag is providing a new profile for the audit signing cert,
caAuditSigningCert.
There are existing profiles for the transport (caTransportCert)
and storage (caStorageCert) certificates.
https://pagure.io/freeipa/issue/8545
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
The variable was None by default and set to /etc/ipa/dnssec/openssl.cnf
for fedora only because the code is specific to the support of pkcs11
engine for bind. As a consequence ipa-backup had a "None" value in the
list of files to backup and failed on Exception.
ipa-backup code is able to handle missing files, and the code using
the pkcs11 engine is called only when NAMED_OPENSSL_ENGINE is set
(only in fedora so far). It is safe to always define a value for
DNSSEC_OPENSSL_CONF even on os where it does not exist.
The fix also improves the method used to verify that a path exists.
Fixes: https://pagure.io/freeipa/issue/8597
Reviewed-By: Alexander Bokovoy <abbra@users.noreply.github.com>
Commit 8b6d1ab854 added support for
subordinate UPN suffixes but missed the case where subordinate UPN is a
subdomain of the forest root domain and not mentioned in the UPN
suffixes list.
Correct this situation by applying the same check to the trusted domain
name as well.
Fixes: https://pagure.io/freeipa/issue/8554
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This test adds new UPN Suffix on the AD side
within the ad.test subtree i.e new.ad.test and this
UPN is then assigned to aduser and then try to
kinit using aduser along with the UPN set, to ensure
that the kinit succeeds
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
0d67180f7d introduced the with_admin
fixture using class scope, which caused test failures as pytest
instantiated it before the multihost fixture.
It additionally failed to account for jitter - the issued ticket becomes
within a window of the expected lifetime, so we need to include the
ticket lifetime jitter into that calculation.
Finally, the PKINIT test could not have ever worked because PKINIT is
not set up as part of policy testing.
Related: https://pagure.io/freeipa/issue/8590
Also-authored-by: Rob Crittenden <rcritten@redhat.com>
Signed-off-by: Robbie Harwood <rharwood@redhat.com>
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
TestACMEwithExternalCA subclasses TestACME which subclasses
CALessBase.
CALessBase is necessary to generate the certificates for the
test_third_party_certs() test.
This means that the TestACME install classmethod needs to be
called by its subclasses. But the install actually does the
installation of the servers as well so needs to be aborted
at that point in the case of a subclass.
https://pagure.io/freeipa/issue/8581
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
This directory defines the location for ACME-related files
used by mod_md. Specify and create it rather than relying
on defaults to both fix a test failure and to make the
files accessable for debugging purposes.
https://pagure.io/freeipa/issue/8581
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Mohammad Rizwan <myusuf@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Reviewed-By: Armando Neto <abiagion@redhat.com>
Test for PF#7137: [RFE]: Able to browse different links
from IPA web gui in new tabs
Test verifies whether opening target link in new tab
navigates to target (desired behaviour) compared to creation of copy of
current state of page on new tab (old behaviour).
WebUI: Add method for opening link in a new tab
Add `navigate_to_row_record_in_new_tab` method.
`navigate_to_row_record_in_new_tab` works as
`navigate_to_row_record` but opens link in new tab instead and focuses
on it.
Related: https://pagure.io/freeipa/issue/7137
Signed-off-by: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Michal Polovka <mpolovka@redhat.com>
Reviewed-By: Serhii Tsymbaliuk <stsymbal@redhat.com>
test_nsscheck_cert_expiring is moved under test_ipa_healthcheck_expiring
This patch checks that the 'ipahealthcheck.ds.nss_ssl' check in
healthcheck tool reports the correct status for the "Server-Cert"
about to expire and already expired respectively.
Signed-off-by: Sudhir Menon <sumenon@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
This command doesn't require any permissions that a host doesn't
already have and it simplifies overall credential handling.
It also corrects the case where the server API cache is out of
date and there are no credentials available to refresh it which
can lead to a confusing error message.
Also switch to MEMORY-based ccache rather than file to avoid
having to clean up the filesystem.
https://pagure.io/freeipa/issue/8531
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Fraser Tweedale <ftweedal@redhat.com>
Commit 26b9a69784 did not fully fix the
client-only build as util/ipa_pwd.c unconditionally includes
pwquality.h.
Make sure we define USE_PWQUALITY in the full server configuration and
if that one is defined, include libpwquality use.
Fixes: https://pagure.io/freeipa/issue/8587
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
# Try to load yml/yaml files via safe_load, which recognizes only standard
# YAML tags and cannot construct an arbitrary Python object.
# There are Jinja yaml templates, which differ from reqular ones. These
# files should be placed on skip list (YAML_TEMPLATE_FILES), otherwise
# safe_load fails.
# Also check PRCI definitions yaml files jobs format and content with
# prci_checker script
.PHONY:yamllint
yamllint:
YAML_TEMPLATE_FILES="\
@@ -326,7 +361,10 @@ yamllint:
echo$${YAML};\
$(PYTHON) -c "import yaml; f = open('$${YAML}'); yaml.safe_load(f); f.close()"||{echo Your YAML file: $${YAML} has a wrong syntax or this is a Jinja template. In the latter clause, consider to add your YAML file to the YAML_TEMPLATE_FILES list in Makefile.am.;exit 1;}\
@@ -78,7 +78,7 @@ Specifies the insecure CA end user port. The default is 8080.
The time to wait for a certmonger request to complete during installation. The default value is 300 seconds.
.TP
.B context <context>
Specifies the context that IPA is being executed in. IPA may operate differently depending on the context. The current defined contexts are cli and server. Additionally this value is used to load /etc/ipa/\fBcontext\fR.conf to provide context\-specific configuration. For example, if you want to always perform client requests in verbose mode but do not want to have verbose enabled on the server, add the verbose option to \fI/etc/ipa/cli.conf\fR.
Specifies the context that IPA is being executed in. IPA may operate differently depending on the context. The current defined contexts are cli, server and dns. Additionally this value is used to load /etc/ipa/\fBcontext\fR.conf to provide context\-specific configuration. For example, if you want to always perform client requests in verbose mode but do not want to have verbose enabled on the server, add the verbose option to \fI/etc/ipa/cli.conf\fR.
.TP
.B debug <boolean>
When True provides detailed information. Specifically this set the global log level to "debug". Default is False.
@@ -113,6 +113,15 @@ Specifies whether values should be prompted for or not. The default is True.
.B kinit_lifetime <time duration spec>
Controls the lifetime of ticket obtained by users authenticating to the WebGUI using login/password. The expected format is a time duration string. Examples are "2 hours", "1h:30m", "10 minutes", "5min, 30sec". When the parameter is not set in default.conf, the ticket will have a duration inherited from the default value for kerberos clients, that can be set as ticket_lifetime in krb5.conf. When the ticket lifetime has expired, the ticket is not valid anymore and the GUI will prompt to re-login with a message "Your session has expired. Please re-login."
.TP
.B ldap_cache <boolean>
Enable a per-request LDAP cache. The default is True.
.TP
.B ldap_cache_size <integer>
The maximum number of entries cached if ldap_cache is True. Since this cache is per-request it is not expected to be very large. The default is 100. Setting the value < 1 effectively disables the cache regardless of the ldap_cache setting
.TP
.B ldap_cache_debug <boolean>
Log details on hits, misses, etc. for the LDAP cache if the cache is enabled.
.TP
.B ldap_uri <URI>
Specifies the URI of the IPA LDAP server to connect to. The URI scheme may be one of \fBldap\fR or \fBldapi\fR. The default is to use ldapi, e.g. ldapi://%2fvar%2frun%2fslapd\-EXAMPLE\-COM.socket
.TP
@@ -154,6 +163,9 @@ Specifies the mode the server is running in. The currently support values are \f
.B mount_ipa <URI>
Specifies the mount point that the development server will register. The default is /ipa/
.TP
.B oidc_child_debug_level <debuglevel>
Specifies the debug level of \fBoidc_child\fR, a helper process used by \fBipa-otpd\fR for OIDC/OAuth2 authentication. Level can be between 0 and 10, the higher the more details. If the level is 6 or higher HTTP debug output is added as well.
.TP
.B prompt_all <boolean>
Specifies that all options should be prompted for in the IPA client, even optional values. Default is False.
.TP
@@ -166,6 +178,9 @@ Specifies the Kerberos realm.
.B replication_wait_timeout <seconds>
The time to wait for a new entry to be replicated during replica installation. The default value is 300 seconds.
.TP
.B schema_ttl <seconds>
The number of seconds for the ipa tool to cache the IPA API and help schema. Reducing this value during development is helpful so that API changes are seen sooner in the tool. Setting this on a server will define the TTL for all client versions > 4.3.1. Client versions > 4.3.1 that connect to IPA servers older than 4.3.1 will use the client-side configuration value. The default is 3600 seconds. 0 disables the cache. A change in the ttl will not be immediately recognized by clients. They will use the new value once their current cache expires.
It is also possible to define context\-specific configuration files. The \fBcontext\fR is set when the IPA api is initialized. The two currently defined contexts in IPA are \fBcli\fR and \fBserver\fR. This is helpful, for example, if you only want \fBdebug\fR enabled on the server and not in the client. If this is set to True in \fIdefault.conf\fR it will affect both the ipa client tool and the IPA server. If it is only set in \fIserver.conf\fR then only the server will have \fBdebug\fR set. These files will be loaded if they exist:
It is also possible to define context\-specific configuration files. The \fBcontext\fR is set when the IPA api is initialized. The currently defined contexts in IPA are \fBcli\fR, \fBserver\fR and \fBdns\fR. This is helpful, for example, if you only want \fBdebug\fR enabled on the server and not in the client. If this is set to True in \fIdefault.conf\fR it will affect both the ipa client tool and the IPA server. If it is only set in \fIserver.conf\fR then only the server will have \fBdebug\fR set. These files will be loaded if they exist:
.TP
.I /etc/ipa/cli.conf
system\-wide IPA client configuration file
.TP
.I /etc/ipa/server.conf
system\-wide IPA server configuration file
.SH "EXAMPLES"
.TP
An example of a context-specific configuration file is \fB/etc/ipa/dns.conf\fR to be used to increase debug output of the IPA DNSSEC daemons.
@@ -89,6 +89,9 @@ Specifies the From: e-mail address value in the e-mails sent. The default is nor
.B notify_ttls <list of days>
This is the list of days before a password expiration when ipa-epn should notify a user that their password will soon require a reset. If this value is not specified then the default list will be used: 28, 14, 7, 3, 1.
.TP
.B msg_subject <subject>
Specifies the subject of the e-mails sent. The default is "Your password will expire soon."
.TP
.B msg_charset <type>
Set the character set of the message. The default is utf8. This will result in he body of the message being base64-encoded.
The automount configuration consists of three files:
The automount configuration consists of two files:
.PP
.IPo
/etc/nsswitch.conf
.IPo
/etc/sysconfig/autofs
.IPo
/etc/autofs_ldap_auth.conf
.TP
By default this will use DNS discovery to attempt to determine the IPA server(s) to use. If IPA servers are discovered then the automount client will be configured to use DNS discovery.
@@ -42,9 +40,9 @@ The default automount location is named default. To specify a different one use
.TP
The IPA client must already be configured in order to configure automount. The IPA client is configured as part of a server installation.
.TP
There are two ways to configure automount. The default is to use sssd to manage the automount maps. Alternatively autofs can configured to bind to LDAP over GSSAPI and authenticate using the machine's host principal.
SSSD is configured to manage the automount maps.
.TP
The nsswitch automount service is configured to use either sss or ldap and files depending on whether SSSD is configured or not.
The nss automount service is configured to use sss and files.
.TP
NFSv4 is also configured. The rpc.gssd and rpc.idmapd are started on clients to support Kerberos\-secured mounts.
.SH"OPTIONS"
@@ -54,10 +52,7 @@ Set the FQDN of the IPA server to connect to.
\fB\-\-location\fR=\fILOCATION\fR
Automount location.
.TP
\fB\-S\fR, \fB\-\-no\-sssd\fR
Do not configure the client to use SSSD for automount.
NFS domain for idmapd.conf. If unset, defaults to the IPA domain. If set to DNS, let idmapd or nfsidmap determine the domain from DNS (see idmapd(8) or nfsidmap(5) for details). If set to anything else, set idmapd.conf's Domain entry to that value.
.TP
\fB\-d\fR, \fB\-\-debug\fR
@@ -71,21 +66,12 @@ Restore the automount configuration files.
.SH"FILES"
.TP
Files that will be always be configured:
Files that will be configured:
/etc/nsswitch.conf
.TP
Files that will be configured when SSSD is the automount client (default):
/etc/sssd/sssd.conf
.TP
Files that will be configured when using the ldap automount client:
ipa\-client\-samba \- Configure Samba file server on an IPA client
.SH"SYNOPSIS"
@@ -18,7 +18,7 @@ During the configuration process, the tool will perform following steps:
1. Discover details of IPA domain: realm, domain SID, domain ID range
2. Discover details of trusted Actvide Directory domains: domain name, domain SID, domain ID range
2. Discover details of trusted Active Directory domains: domain name, domain SID, domain ID range
3. Create Samba configuration file using the details discovered above.
@@ -34,12 +34,12 @@ The tool does not start nor does it enable Samba file services after the configu
systemctl enable --now smb winbind
.SS"Assumptions"
The ipa\-client\-samba script assumes that the machine has alreaby been enrolled into IPA.
The ipa\-client\-samba script assumes that the machine has already been enrolled into IPA.
.SS"IPA Master Requirements"
At least one IPA master must hold a \fBTrust Controller\fR role. This can be achieved by running ipa\-adtrust\-install on the IPA master. The utility will configure IPA master to be a domain controller for IPA domain.
IPA master holding a \fBTrust Controller\fR role has also to have support for a special service command to create SMB service, \fBipa service-add-smb\fR. This command is available with FreeIPA 4.8.0 or later release.
IPA master holding a \fBTrust Controller\fR role has also to have support for a special service command to create SMB service, \fBipa service-add-smb\fR. This command is available with IPA 4.8.0 or later release.
ipa\-epn provides a method to warn users via email that their IPA account password is about to expire.
It can be used in dry\-run mode which is recommmended during setup. The output is always JSON in this case.
It can be used in dry\-run mode which is recommended during setup. The output is always JSON in this case.
It can also be launched daily by its systemd timer.
In this case it will parse its configuration file epn.conf(5) and send an email to users whose passwords are expiring within the defined future date ranges.
@@ -45,10 +45,10 @@ Together, these two CLI options can be used to determine how many emails would b
The \fB\-\-to\-nbdays\fR CLI option implies \fB\-\-dry\-run\fR.
.TP
\fB\-\-from\-nbdays\fR\fI<number of days>\fR
See \fB\-\-to\-nbdays\fR for an explanation. This option must be used in conjonction with \fB\-\-to\-nbdays\fR.
See \fB\-\-to\-nbdays\fR for an explanation. This option must be used in conjunction with \fB\-\-to\-nbdays\fR.
.TP
\fB\-\-dry\-run\fR
The \fB\-\-dry\-run\fR CLI option is intented to test ipa\-epn's configuration.
The \fB\-\-dry\-run\fR CLI option is intended to test ipa\-epn's configuration.
For instance, if notify_ttls is set to 21, 14, 3, \fB\-\-dry-run\fR would display the list of users whose passwords would expire in 21, 14, and 3 days in the future.
SSSD provides an option to tune its behavior regarding private groups: *auto_private_groups*. Currently, this option is designed to work with **ldap** or **ad** providers and doesn’t behave properly when the provider is **ipa** and a trust is configured between IPA and AD.
The goal of this RFE is to propose the same setting at the trust level.
References:
* [BZ 1688267](https://bugzilla.redhat.com/show_bug.cgi?id=1688267) (ipa) [RFE] IPA to allow setting a new range type
* [BZ 1649464](https://bugzilla.redhat.com/show_bug.cgi?id=1649464) (sssd) auto_private_groups not working as expected with posix ipa/ad trust
* [SSSD #4216](https://github.com/SSSD/sssd/issues/4216) [RFC] IPA: allow switching off user private groups for trusted AD users
## High level overview
### Current auto private group behavior
The *auto_private_group* setting can be configured in the **DOMAIN** section of *sssd.conf*.
The *auto_private_groups* setting can take any of these 3 values:
* **true**: always create a user private group from the user’s UID number. If a GID number already exists, it is ignored.
* **false**: always use the users’ primary GID number. A group with this GID number must already exist.
* **hybrid**: a primary group is generated if the entry has UID = GID but there is no group with this GID. If UID and GID are different, a group with this GID number must already exist.
### Transposition to AD trust environments
#### Scope
When an IPA provider is used, with an AD trust, the UIDs and GIDs belong to different ID ranges:
* The users defined in IPA contain a uidNumber and gidNumber attribute, which must fit in the *ipa-local* range
* There are 2 possibilities for the users defined in AD:
* If they contain a uidNumber and gidNumber, the trust is established with a *ipa-ad-trust-posix* range and uidNumber and gidNumber values are taken from the user entry on AD
* If they don’t define uidNumber/gidNumber, the trust is established with a *ipa-ad-trust* range and SSSD computes uidNumber=gidNumber from the value of the user SID.
As there can be multiple ranges defined for a single ipa provider (for instance if multiple trusts are set), the *auto_private_groups* setting cannot have a domain scope and needs to be set at the **id range** level.
=> The setting needs to be added as an option to *ipa-ad-trust-posix* and *ipa-ad-trust* ranges.
#### Default behavior
When the *auto_private_groups* option is not explicitly set, it uses a default value:
* For *ipa-ad-trust-posix* range: default value = **false**. This means that the uidNumber and gidNumber of the AD entry are always used, and a group with the gidNumber must already exist.
* For *ipa-ad-trust* range: default value = **true**. This means that the uidNumber is mapped from the entry SID, the gidNumber is always set to the same value and a private group is always created.
### User stories
#### User story #1: AD trust with posix attributes, gidNumber corresponding to an existing group
As an administrator, I want to establish a trust between my IPA domain and my AD domain. My AD users contain posix attributes (uidNumber and gidNumber) that I want to use on IPA side, and the gidNumber corresponds to an existing AD group.
On any IPA client, the *id* command must return the uidNumber and gidNumber stored in AD:
If the gidNumber does not correspond to an existing group, the entry is not resolved.
#### User story #2: AD trust with posix attributes, gidNumber not corresponding to any group
As an administrator, I want to establish a trust between my IPA domain and my AD domain. My AD users contain posix attributes (uidNumber and gidNumber) but there is no group with the corresponding gidNumber.
On any IPA client, the id command must return the uidNumber stored in AD:
#### User story #3: AD trust with SID mapping, private group created with the uidNumber
As an administrator, I want to establish a trust between my IPA domain and my AD domain. My AD users are not posix, and their SID is mapped to an uidNumber and gidNumber on IPA. The gidNumber is equal to the uidNumber and a corresponding private group is generated.
On any IPA client, the id command must return the value computed from the SID stored in AD:
#### User story #4: AD trust with SID mapping, primary group reusing the primaryGroupID
As an administrator, I want to establish a trust between my IPA domain and my AD domain. My AD users are not posix, and their SID is mapped to an uidNumber on IPA. I want to compute the gidNumber from the RID of the AD entry’s primaryGroupID.
On any IPA client, the id command must return the value computed from the SID stored in AD:
When the range has the *ipa-ad-trust-posix* type, the uidNumber and gidNumber are taken from the AD entry. Currently, if the gidNumber does not correspond to an existing group, the user cannot be resolved.
This situation can be fixed by setting
* *auto_private_groups=true*: a private group is created with gidNumber = uidNumber.
* *auto_private_group=hybrid*: if uidNumber=gidNumber but there is no group with this gidNumber, a private group is created with the gidNumber
auto_private_groups | true | false | hybrid
------------------- | ---- | ----- | ------
AD User entry with uidNumber=1000 gidNumber=missing | uid=1000 gid=1000 +private group | Not resolvable | Not resolvable
AD user entry with uidNumber=1000 gidNumber=2000 Group does not exist | uid=1000 gid=1000 +private group | Not resolvable | Not resolvable
AD User entry with uidNumber=1000 gidNumber=1000 Group does not exist | uid=1000 gid=1000 +private group | Not resolvable | uid=1000 gid=1000 +private group
AD User entry with uidNumber=1000 gidNumber=2000 Group exists | uid=1000 gid=1000 +private group | uid=1000 id=2000 | uid=1000 gid=2000
##### ipa-ad-trust
When the range has the *ipa-ad-trust* type, the uidNumber and gidNumber are computed from the entry’s SID. They currently both take the same value.
There are situations where the customer doesn’t want this behavior but would prefer to use the primaryGroupID set in the AD entry as gidNumber. This can be achieved by setting *auto_private_groups=false* and would solve the ticket [SSSD #4216](https://github.com/SSSD/sssd/issues/4216).
auto_private_groups | true | false | hybrid
------------------- | ---- | ----- | ------
AD User entry with SID mapped to 1000 primaryGroupID mapped to 2000 | uid=1000 gid=1000 +private group | uid=1000 gid=2000 | uid=1000 gid=2000
## Design
### IPA part
#### Schema
The LDAP schema for ID ranges needs to be modified:
* Creation of a new attribute to store the *autoprivategroups* option (with a DirectoryString type, single valued)
* Modification of the *ipaTrustedADDomainRange* objectclass in order to allow the new attribute as OPTIONAL
objectClasses: (2.16.840.1.113730.3.8.12.17 NAME 'ipaTrustedADDomainRange' SUP ipaIDrange STRUCTURAL MUST ( ipaBaseRID $ ipaNTTrustedDomainSID ) MAY ( ipaAutoPrivateGroup ) X-ORIGIN 'IPA v3' )
```
#### ID-range-related commands
*ipa idrange-add* and *ipa idrange-mod* need to provide a new option allowing to set *autoprivategroups* for the range if the range type is either *ipa-ad-trust* or *ipa-ad-trust-posix*.
*ipa idrange-find* will also allow to use *auto-private-groups* as a search criteria.
Note that *ipa-local* range should not accept the option.
```shell
# ipa idrange-add --help
[...]
--auto-private-groups=[‘true’, ‘false’, ‘hybrid’]
Automatic creation of private groups, one of allowed values.
[...]
# ipa idrange-mod --help
…
--auto-private-groups=[‘true’, ‘false’, ‘hybrid’]
Automatic creation of private groups, one of allowed values.
*ipa idrange-show* must display the value, if any is set.
```shell
# ipa idrange-show AD-DOM.COM_id_range
[...]
Range name: AD-DOM.COM_id_range
First Posix ID of the range: 1136400000
Number of IDs in the range: 200000
First RID of the corresponding RID range: 0
Domain SID of the trusted domain: S-1-5-21-3005052257-2375221410-442149667
Range type: Active Directory domain range
Auto private groups: true
[...]
```
#### Trust establishment
*ipa trust-add* command also internally creates an idrange object but the decision is **NOT** to add new parameters for *auto-private-range* at this level. The rationale is that *ipa trust-add* currently provides only a subset of the idrange options (for instance it doesn’t allow to set *--rid-base*) and the finer configuration options are set using *ipa idrange-add* or *ipa idrange-mod*.
When the sysadmin wants to set *--auto-private-range*, he needs to do
* *ipa trust-add [options] AD-REALM*: this command also creates the idrange *AD-REALM_id_range*
There is no need to handle the new option during upgrade: SSSD must be able to assign a default value to *auto_private_groups* if none is set in the id range. The default value is picked according to the range type, in accordance with the content of sssd.conf man page:
* For *ipa-ad-trust-posix* range: default value = **false**.
* For *ipa-ad-trust* range: default value = **true**.
The schema upgrade does not require specific actions: the new attribute is optional in the existing object class and the schema is replicated to the other replicas in the topology as soon as one of them is upgraded with the new schema.
# IPA and an external identity provider integration - idp objects
IPA needs to store and manage IdP references. As a new IPA object, IdP reference needs:
* creation of an LDAP object class and LDAP attribute types used to store IdP
information
* creation of a new IPA API managing the IdP references
* definition of access control lists protecting access to the IdP references
* upgrade and backward compatibility.
## LDAP Schema
### New attribute types
The external IdP reference object needs to contain the following information:
* name (string)
* authorization endpoint (URI mapped as a case-sensitive string),
* device authorization endpoint (URI mapped as a case-sensitive string), for instance https://oauth2.googleapis.com/device/code
* token endpoint (URI mapped as a case-sensitive string), for instance https://oauth2.googleapis.com/token
* userinfo endpoint (URI mapped as a case-sensitive string)
* JWKS endpoint (URI mapped as a case-sensitive string)
* OIDC URL (URI mapped as a case-sensitive string)
* `client_id` (case-sensitive string)
* `client_secret` (optional, may be an empty string, must be protected as a password)
* scope (case sensitive string)
* user subject attribute (case-sensitive string)
Additionally, an idp object needs to be referenced from a user object, thus a link attribute has to be defined. This translates into the following attribute type definitions:
objectClasses: (2.16.840.1.113730.3.8.24.7 NAME 'ipaIdpUser' SUP top AUXILIARY DESC 'User from an external Identity Provider ' MAY ( ipaIdpConfigLink $ ipaIdpSub ) X-ORIGIN 'IPA v4.9' )
```
In the current implementation, a user can be connected to a single IdP
reference. This is similar to RADIUS proxy references already used in IPA. For
both RADIUS and IdP cases IPA needs to know a particular user identity within
the context of the referred resource (IdP or RADIUS server). Without associated
IdP user information authentication of the IPA user will not be triggered.
### LDAP indices
Indices need to be defined for all the attributes that can be used in
`ipa idp-find` command.
- The `cn` attribute is already indexed.
- The `ipaidpdevauthendpoint`, `ipaidpauthendpoint`, `idaidptokenendpoint`, and `idpidpscope` attributes
need indices (equality and substring indices).
## IdP reference API
### Command Line Interface
```
idp-add Add new external IdP server.
idp-del Delete an external IdP server.
idp-find Search for external IdP servers.
idp-mod Modify an external IdP server.
idp-show Display information about an external IdP server.
```
Following common options defined for the external IdP server object:
* memberOf (will be reconstructed after the migration)
* userCertificate if issued by the remote IPA server
#### Attributes to update to new REALM
If a REALM change is part of migration then these will need to be updated
to reflect the new installation.
* krbPrincipalKey
* krbCanonicalName
#### Attributes to change basedn
Prior to writing a new value any DN syntax attributes will need to be
examined to see if they contain the remote baseDN if it is different. If
so then the following need to be updated:
* mepManagedEntry
* mepManagedBy
#### Migrating custom schema
Schema is stored in LDAP so we *should* be able to use our ldap library
to discover differences. We probably want to do only ADD and bail if we
determine something will be a MOD or DEL.
#### Excluding specific entries
The ignore user/group options of migrate-ds were introduced so that non-compliant
entries could be skipped to not block the migration. We can ignore these for now.
#### Performance considerations
Some benefits would be likely if common entry types were added in
batches rather than individually. This would not easily allow for a
"stop on failure" approach but could be considerably faster.
The basic idea is that added entries of the same type (user, group, etc)
would be accumulated and added together. The result would be iterated
through and logged.
The memberOf plugin should be disabled. A fixup task can be run post-migration
to calculate the memberships.
#### Idempotency
The migrate-ds plugin manages idempotency because it skips over entries
that already have been migrated (or exist). This new migration will handle
it by merging the remote and local entries. This may not be desirable in
all cases.
#### Migration steps
A simplistic view of the steps
Set migration mode=True
Disable compat
Disable memberof
Migrate schema
Loop through the list of remote objects:
- fix up any DN values or other syntaxes
- remove conflicting attributes/objectclasses
- write new entry
- if entry already exists, merge them if possible preferring remote side
Enable memberOf
Create memberOf fixup task
Enable compat
Restart world
Run ipa-server-upgrade
### pure LDAP migration
The current code generally works minus a few bugs and RFEs, some of which are resolved
by doing an IPA-to-IPA migration instead. It should be maintained for now and migrated
to the standalone client in the future.
These bugs should be considered for the existing plugin.
* https://pagure.io/freeipa/issue/3096 - error when migrating unknown schema
* https://pagure.io/freeipa/issue/3100 - Check for userPassword in migration
* https://pagure.io/freeipa/issue/4738 - [RFE] ipa migrate-ds should provide option for creating UPG from posixGroup objectClass
* https://pagure.io/freeipa/issue/5020 - migrate-ds: does not show migrated users if an error happened during group migration
* https://pagure.io/freeipa/issue/5693 - Passwords become "expired" when migrating from directory server to IPA
* https://pagure.io/freeipa/issue/6105 - migrate-ds is not completely ignoring attributes.
* https://pagure.io/freeipa/issue/6360 - ipa migrate-ds does not rename uniquemember/member attributes properly
* https://pagure.io/freeipa/issue/6380 - ipa migrate-ds should print warning for referrals
* https://pagure.io/freeipa/issue/7368 - ipa migrate-ds converts groupofuniquenames objects to groupofnames, but leaves groupofuniquenames objectclass present
* https://pagure.io/freeipa/issue/7749 - `ipa migrate-ds` fails to migrate user and group data from directory server to IDM.
## Logging
By default the log file will be /var/log/ipa-migrate.log and will be appended to
and not overwritten. This is so it can reflect multiple runs if they are required.
At least the DN of all entries written should be logged (pkey may be sufficient).
Logging by object type could be handy and should natural since this is how the
objects will be sorted.
DEBUG logging may want to show gory details, particularly when merging entries.
## Implementation
The command will be named ipa-migrate. It must determine whether the remote server is
actually an IPA server or not. ipaclient/discovery.py::ipacheckldap may be re-usable.
The standalone client should use a unique context, migration. This will
allow for a separate configuration file.
## Feature Management
### UI
No UI option will be provided. This is command-line client only.
### CLI
Overview of the CLI commands.
#### IPA to IPA
Advance knowledge of the DIT substantially reduces the number of options
necessary for migration.
| Option | Description |
| --- | --- |
| --dry-run | try the migration without writing data |
| --force | ignore errors and keep going |
| --version | version of the tool |
| --quiet | output only errors |
| --log-file | log to the given file |
| --help | this message |
The DM password will be prompted for interactively.
| Argument | Description |
| url | ldap url for remote IPA server |
#### pure LDAP
Will remain unchanged unless one of the bug fixes requires it (perhaps for the UPG
ticket).
### Configuration
N/A
## Upgrade
N/A
## Test plan
There are currently no tests for migration.
Some simplisitic approaches for starting testing might include:
* Count the number of entries that will be migrated and ensure they were migrated, by type (hosts, groups, etc).
* Verify that the services enabled on the remote side are enabled after migration (NIS, ACME, etc).
* Double-check, perhaps spot-checking, memberOf
* Migrate a password to ensure it was imported properly
We have a data generation script in freeipa-tools that may be leveraged to generate the data but it currently generates a LOT of entries which is likely too much for automation.
## Troubleshooting and debugging
Include as much information as possible that would help troubleshooting:
- Does the feature rely on existing files (keytabs, config file...)
- Does the feature produce logs? in a file or in the journal?
- Does the feature create/rely on LDAP entries?
- How to enable debug logs?
- When the feature doesn't work, is it possible to diagnose which step failed? Are there intermediate steps that produce logs?
IPA previously allowed LDAP authentications past expiration to prevent a chicken-and-egg problem. Because all administratively-set passwords are expired a user must be given an opportunity to authenticate in order to reset their password. This allowed for unlimited authentications.
## Use Cases
Many IPA users have asked for a mechanism to deny LDAP authentications once expired.
## How to Use
A new option is added to password policies, --gracelimit. This specifies the number of allowed BINDs past expiration.
To set the global policy:
ipa pwpolicy-mod --gracelimit 3
This will allow 3 additional logins past expiration.
Setting to -1 disables the grace limit check.
Setting to 0 will do a grace limit check but will always fail because no further logins are allowed. The distinction between 0 and -1 is that with 0 a password policy control is returned.
## Design
An expired draft around LDAP password policies describes two methods for limiting account access past expiration, https://tools.ietf.org/id/draft-behera-ldap-password-policy-10.html
The RFE provides two options (which can be use simultaneously).
1. Limit the number of authentications past expiration
2. Limit by time past the expiration
This implements the first part only. 389-ds already implements this so we can reuse existing schema, using their alternative attribute names passwordGraceLimit (system) and passwordGraceUserTime (user).
A 389-ds plugin will be registered as a bind preop function which will compare the policy grace period with the user's grace period
The password policy plugin will have a new attribute added, passwordGraceLimit, to store the policy maximum.
The basic flow is:
- Register preop BIND plugin
- On BIND, get the DN
- Get the password policy for that DN
- If the policy grace limit == -1 then exit
- Compare the policy grace limit to the user's passwordGraceUserTime
- If the user's grace limit exceeds policy, fail
- If the grace limit == 0, fail
- Set a password policy control with the number of remaining logins
- Set the time in the policy to -1 to indicate no value
On successful password reset (by anyone) reset the user's passwordGraceUserTime to 0.
The default value on install/upgrade will be -1 to retail existing behavior.
The per-user attempts will not be replicated.
## Implementation
N/A
## Feature Management
### UI
The new option is not visible in the UI yet.
### CLI
Overview of the CLI commands. Example:
| Command | Options |
| --- | ----- |
| pwpolicy-mod | --gracelimit=<-1 to MAXINT> |
### Configuration
The plugin will be registered on new installations.
## Upgrade
The plugin will be registered on upgrades.
## Test plan
1. Set a grace limit on a custom policy, ldapsearch a few times, verify that the limit is set to what is expected
2. Keep going until authentication fails
3. Reset password, verify that authentication can happen
4. With two servers verify that an authentication one one server doesn't trigger a replication event to the other.
Previously PKI used range-based serial numbers in order to ensure uniqueness across multiple clones. Based on their experience they determined that this would not fit well into cloud environments with short-lived certificates.
The new version 3 implementation of serial numbers uses a 128-bit random value for the serial number which they compute collisions will be mathematically small.
See https://github.com/dogtagpki/pki/wiki/Random-Certificate-Serial-Numbers-v3 for more details.
Given the known issues reported this will be supported in IPA for new installations only.
## Use Cases
RSNv3 will be supported only for new installations. If enabled it is required to use RSNv3 on all PKI services including the CA and KRA in the deployment.
The PKI team currently discourages mixing ranged and random serial numbers and in practice creating a ranged clone from a RSNv3 server will fail. At some point in the future it may be possible to write a tool to flip values in an existing deployment to enable RSNv3 on existing servers prior to creating new replicas but it will be complex and require significant testing. Currently interoperability between ranged and RSNv3 is not a good experience.
## How to Use
Instructions for installing a CA: https://github.com/dogtagpki/pki/blob/master/docs/installation/ca/Installing-CA-with-Random-Serial-Numbers-v3.adoc
Instructions for installing a KRA: https://github.com/dogtagpki/pki/blob/master/docs/installation/kra/Installig-KRA-with-Random-Serial-Numbers-v3.adoc
A new option was added for ipa-server-install and ipa-ca-install to --random-serial-numbers to enable RSNv3. This sets the configuration values as described in the documentation above.
## Design
### SerialNumber class
The first limitation within IPA related to RSNv3 is the size of the serial numbers which can be up to 40-digit decimal values. This far exceeds the limits of XML-RPC and JSON integer values.
There has never been a reason to treat serial numbers as integers and the original CA implementation transmitted them as strings. A new class will be needed to handle the huge integers as strings with some basic validation.
The validation will include:
- Minimum length of 1, no empty serial numbers
- Maximum length of 40 (2^^128)
- No negative values
- Greater than 0
- If prefixed with 0x a valid hex value
The only "math" we do on serial numbers is conversion to and from hex. Python internally can handle these huge numbers, it is only in the transmission that there is a problem.
### Determining RSN version
There are two possible methods to handle the introduction of random serial numbers:
#### Implement as a Domain Level 2
There are existing checks which will limit interoperability that may be extended as needed. This was not chosen because with RSNv3 it is an all or nothing affair according to the PKI documentation. All existing servers would need to be moved at once, not something easily enforceable or doable in large installations.
#### Store the RSN version as a value in LDAP
Instead store the version number in LDAP which is easily searchable and a common practice in IPA. A non-existent value will be treated as 0 and therefore no RSN capability. This will allow for expansion into future RSN versions. Current CA ACLs can be extended to allow read/write to this for admin users.
### KRA impact
The PKI team strongly encouraged that if RSNv3 is enabled anywhere then it should be enabled everywhere. This includes request and key IDs and a KRA. A check will be done on KRA installation to automatically enable RSNv3 if it is enabled on the underlying CA.
## Implementation
The RSN version will be stored in the IPA CA entry which will exist on any IPA server with a CA.
The new attribute is named ipaCaRandomSerialNumberVersion and consists of:
attributeTypes: (2.16.840.1.113730.3.8.21.1.9 NAME 'ipaCaRandomSerialNumberVersion' DESC 'Random Serial Number Version' EQUALITY integerMatch ORDERING integerOrderingMatch SYNTAX 1.3.6.1.4.1.1466.115.121.1.27 SINGLE-VALUE X-ORIGIN 'IPA v4.9 RSNv3' )
This attribute is not yet merged into IANA pending design approval.
During replica install or ipa-ca-install a check will be made to determine if RSN is > 0 to determine whether it is allowed or not.
No special effort is needed to include externally signed IPA certificates.
CAless installations are unaffected. Promotion from CAless can include RSN and be held to the same standards as a new installation.
Unfortunately the PKI server provides no mechanism to determine whether the RSNv3 capability exists in the server so a straight version number check is required. It is only available in 11.2.0+.
## Feature Management
### UI
N/A
### CLI
A new option, --random-serial-numbers, is added to ipa-server-install and ipa-ca-install. If --setup-ca is provided to ipa-replica-install then RSN is determined automatically based on the capability of the remote CA.
### Configuration
N/A
## Upgrade
Not currently allowed. While theoretically possible to write a tool to enable it doing so is complex and prone to user error which could result in a non-functioning installation.
## Test plan
A RSNv3 installation varies only from a ranged installation in its CA/KRA configuration. Existing CA tests can be extended to pass the --random-serial-numbers option and from a user perspective, beyond getting certificates with potentially huge serial numbers, there is no difference.
A selected set of existing tests will be subclassed to add the RSNv3 installation option, including CA-less to CA-full installations.
Typically "is the CA ok" tests which do "ipa cert-show 1" can still work to show that communication is working (e.g. Not Found is a fine answer), existing tests will be extended to use the IPA CA serial number so we get a positive answer back.
During development the entire test suite was successfully executed with RSNv3 enabled. Doing so even on a nightly basis is an extremely heavy-weight option.
## Troubleshooting and debugging
Given this supports new installations only that greatly simplifies the possible conditions.
There are only two conditions which would disallow a server to be installed:
- PKI < 11.2.0
- The LDAP-stored RSNv3 value prevents ipa-ca-install or ipa-replica-install
It is possible that some client tools may not honor huge serial numbers but testing has not borne any out yet.
Loaded 100 of 707 files, more files were not shown because too many files have changed in this diff.
Show more
Reference in new issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
zpericic
Florence Blanc-Renaud
Thomas Woerner
Rob Crittenden
Sumedh Sidhaye
Carla Martinez
Sudhir Menon
David Pascual
Erik
Stanislav Levin
Armando Neto
Fraser Tweedale
Antonio Torres
Matthew Davis
Anuja More
Timo Aaltonen
Michal Polovka
Matthew Davis
Alexander Bokovoy
Francisco Trivino
Thorsten Scherf
Mohammad Rizwan
Sumit Bose
Christian Heimes
Brian Turek
Julien Rische
jh23453
François Cami
Petr Vobornik
Sergey Orlov
Vit Mojzis
Pavel Březina
Chris Kelley
Endi S. Dewata
Serhii Tsymbaliuk
Yuri Chornoivan
Emilio Herrera
Jan Kuparinen
Piotr Drąg
Weblate
Marcin Stanclik
Rafael Fontenelle
Hela Basa
Daniel Lara Souza
MIZUTA Takeshi
Carl George
Alexander Scheel
Jan Pazdziora
Troy Dawson
Kaleemullah Siddiqui
JoeDrane
Florence Blanc-Renaud
Peter Keresztes Schmidt
Robbie Harwood
Slava Aseev
Antonio Torres Moríñigo
Mark Reynolds