Configures PKI to remove expired certificates and non-resolved
requests on a schedule.
This is geared towards ACME which can generate a lot of certificates
over a short period of time but is general purpose. It lives in
ipa-acme-manage because that is the primary reason for including it.
Random Serial Numbers v3 must be enabled for this to work.
Enabling pruning enables the job scheduler within CS and sets the
job user as the IPA RA user which has full rights to certificates
and requests.
Disabling pruning does not disable the job scheduler because the
tool is stateless. Having the scheduler enabled should not be a
problem.
A restart of PKI is required to apply any changes. This tool forks
out to pki-server which does direct writes to CS.cfg. It might
be easier to use our own tooling for this but this makes the
integration tighter so we pick up any improvements in PKI.
The "cron" setting is quite limited, taking only integer values
and *. It does not accept ranges, either - or /.
No error checking is done in PKI when setting a value, only when
attempting to use it, so some rudimentary validation is done.
Fixes: https://pagure.io/freeipa/issue/9294
Signed-off-by: Rob Crittenden rcritten@redhat.com
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Avoid reading platform specific `/etc/resolv.conf` in `TestDNSResolver`
unit tests. Systems (e.g. sandboxes) may not have `/etc/resolv.conf`
or this file may not contain any configured name servers.
`TestDNSResolver` unit tests check only customized `nameservers`
property and should not depend on existence of `/etc/resolv.conf`.
Resolver accepts `configure` option.
https://dnspython.readthedocs.io/en/latest/resolver-class.html :
> configure, a bool. If True (the default), the resolver instance is
configured in the normal fashion for the operating system the resolver
is running on. (I.e. by reading a /etc/resolv.conf file on POSIX
systems and from the registry on Windows systems.)
Fixes: https://pagure.io/freeipa/issue/9319
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
freeipa-healthcheck changed some messages related to ipa-ca
DNS record validation in IPADNSSystemRecordsCheck. Include support
for it and retain backwards compatibility.
Fixes: https://pagure.io/freeipa/issue/9291
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
freeipa-healthcheck 0.12 includes a SUCCESS message if the
ipa-ca records are as expected so a user will know they
were checked. For that version and beyond test that it
is included.
Related: https://pagure.io/freeipa/issue/9291
Signed-off-by: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
PKI can issue ACME certs only when the key type is rsa.
With version 2.0.0, certbot defaults to ecdsa key type,
and this causes test failures.
For now, force rsa when requesting an ACME certificate.
This change can be reverted when PKI fixes the issue
on their side (https://github.com/dogtagpki/pki/issues/4273)
Related: https://pagure.io/freeipa/issue/9298
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
https://pylint.pycqa.org/en/latest/user_guide/messages/convention/unnecessary-lambda-assignment.html:
> Used when a lambda expression is assigned to variable rather than
defining a standard function with the "def" keyword.
https://peps.python.org/pep-0008/#programming-recommendations:
> Always use a def statement instead of an assignment statement that
binds a lambda expression directly to an identifier:
def f(x): return 2*x
f = lambda x: 2*x
The first form means that the name of the resulting function object is
specifically ‘f’ instead of the generic ‘<lambda>’. This is more useful
for tracebacks and string representations in general. The use of the
assignment statement eliminates the sole benefit a lambda expression can
offer over an explicit def statement (i.e. that it can be embedded
inside a larger expression)
Fixes: https://pagure.io/freeipa/issue/9278
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
> Emitted when a local variable is accessed before its assignment took
place. Assignments in try blocks are assumed not to have occurred when
evaluating associated except/finally blocks. Assignments in except
blocks are assumed not to have occurred when evaluating statements
outside the block, except when the associated try block contains a
return statement.
Fixes: https://pagure.io/freeipa/issue/9278
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
PEP 632 deprecates the distutils module. Replace
- distutils.spawn.find_executable with shutil.which
- distutils.log with logging
Related: https://pagure.io/freeipa/issue/9278
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
The newer version of pylint has fixed false positives and
does not need anymore these suppressions:
- global-variable-not-assigned
- invalid-sequence-index
- no-name-in-module
- not-callable
- unsupported-assignment-operation
Related: https://pagure.io/freeipa/issue/9278
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
The test ipatests/test_integration/test_fips.py is faking
FIPS mode and calls "openssl md5" to ensure the algo is
not available in the fake FIPS mode.
The error message has been updated with openssl-3.0.5-5.
In the past the command used to return:
$ openssl md5 /dev/null
Error setting digest
140640350118336:error:060800C8:digital envelope routines:EVP_DigestInit_ex:disabled for FIPS:crypto/evp/digest.c:147:
And now it returns:
$ openssl md5 /dev/null
Error setting digest
00C224822E7F0000:error:0308010C:digital envelope routines:inner_evp_generic_fetch:unsupported:crypto/evp/evp_fetch.c:349:Global default library context, Algorithm (MD5 : 97), Properties ()
00C224822E7F0000:error:03000086:digital envelope routines:evp_md_init_internal:initialization error:crypto/evp/digest.c:252:
To be compatible with all versions, only check the common part:
Error setting digest
Mark the test as xfail since installation is currently not working.
Related: https://pagure.io/freeipa/issue/9002
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The test wants to check the error related to an
exception obtained inside a "with pytest.raises" instruction.
The object is an ExceptionInfo and offers a match method
to check the content of the string representation.
Use this match() method instead of str(excinfo) which now
returns
'<ExceptionInfo NoSuchElementException() tblen=10>'
Fixes: https://pagure.io/freeipa/issue/9282
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Mohammad Rizwan Yusuf <myusuf@redhat.com>
The IPA location system relies on DNS record priorities in order to give
higher precedence to servers from the same location. For Kerberos, this
is done by redirecting generic SRV records (e.g.
_kerberos._udp.[domain].) to location-aware records (e.g.
_kerberos._udp.[location]._locations.[domain].) using CNAMEs.
This commit applies the same logic for URI records. URI location-aware
record were created, but there were no redirection from generic URI
records. It was causing them to be ignored in practice.
Kerberos URI and TXT records have the same name: "_kerberos". However,
CNAME records cannot coexist with any other record type. To avoid this
conflict, the generic TXT realm record was replaced by location-aware
records, even if the content of these records is the same for all
locations.
Fixes: https://pagure.io/freeipa/issue/9257
Signed-off-by: Julien Rische <jrische@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The tests calling sssctl domain-status were marked xfail
because of SSSD issue #6331. Now that the issue is fixed
and freeipa bumped sssd required version, remove the xfail
annotation.
Related: https://pagure.io/freeipa/issue/9234
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
On fedora 37+ the dnssec tests were broken. The tests
launched for each pull request were disabled or marked
as xfail.
With the bump of bind version, they should now succeed
and can be re-enabled.
Related: https://pagure.io/freeipa/issue/9216
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The ``ipa-client-install`` command now supports PKINIT for client
enrollment. Existing X.509 client certificates can be used to
authenticate a host.
Also restart KRB5 KDC during ``ipa-certupdate`` so KDC picks up new CA
certificates for PKINIT.
*Requirements*
- The KDC must trust the CA chain of the client certificate.
- The client must be able to verify the KDC's PKINIT cert.
- The host entry must exist. This limitation may be removed in the
future.
- A certmap rule must match the host certificate and map it to a single
host entry.
*Example*
```
ipa-client-install \
--pkinit-identity=FILE:/path/to/cert.pem,/path/to/key.pem \
--pkinit-anchor=/path/to/kdc-ca-bundle.pem
```
Fixes: https://pagure.io/freeipa/issue/9271
Fixes: https://pagure.io/freeipa/issue/9269
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
There is a crash occurring that causes Keycloak to be unable to
communicate with ipa-tuura on the bridge server (replica0). This is
much more prevalent in Fedora 37 so we need to xfail that test case
until the crash is resolved.
Related: https://pagure.io/freeipa/issue/9264
Signed-off-by: Scott Poore <spoore@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Adding test case to test_sso.py to cover login to IPA client as Keycloak
user without relying on external IdP.
create_bridge.py:
- getkeytab in setup_scim_server to allow bridge to use IPA API.
- fix unintstall to remove plugin by version instead of main
test_sso.py:
- add keycloak_add_user function
- add test_ipa_login_with_sso_user
tasks.py:
- add set_user_password to only set password for ipa users
Fixes: https://pagure.io/freeipa/issue/9250
Signed-off-by: Scott Poore <spoore@redhat.com>
Reviewed-By: Anuja More <amore@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The webui test test_subid_range_deletion_not_allowed is
adding a new subid for the admin user but a previous
test already took care of that step.
Remove the call adding the subid.
2nd issue: a given record has to be selected in
order to check that there is no "delete" button.
Fixes: https://pagure.io/freeipa/issue/9214
Signed-off-by: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Test scenario:
configure sssd with ldap_group_name = info for the trusted domain,
so that the group name is read from the "info" attribute
of the AD group entry.
With this setting, it is possible to have a group and a user
that appear on IdM side with the same name.
Ensure that the conflict does not break IdM and that the id,
getent group and getent passwd commands work on an IdM client.
Related : https://pagure.io/freeipa/issue/9127
Signed-off-by: Anuja More <amore@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Previously the message was:
"\n\nIn Directory Server, we offer one hash suitable for this "
"(PBKDF2_SHA256) and one hash\nfor \"legacy\" support (SSHA512)."
"\n\nYour configuration does not use these for password storage "
"or the root password storage\nscheme.\n"
but now the message is:
\n\nIn Directory Server, we offer one hash suitable for this "
"(PBKDF2-SHA512) and one hash\nfor \"legacy\" support (SSHA512)."
"\n\nYour configuration does not use these for password storage "
"or the root password storage\nscheme.\n"
PBKDF2_SHA256 has been replaced with PBKDF2-SHA512
Pagure: https://pagure.io/freeipa/issue/9238
Signed-off-by: Sumedh Sidhaye <ssidhaye@redhat.com>
Reviewed-By: Florence Blanc-Renaud <flo@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Stanislav Levin <slev@altlinux.org>
OpenLDAP has made it explicit to use default CA store as provided by
OpenSSL in 2016:
branches 2.5 and later:
commit 4962dd6083ae0fe722eb23a618ad39e47611429b
Author: Howard Guo <hguo@suse.com>
Date: Thu Nov 10 15:39:03 2016 +0100
branch 2.4:
commit e3affc71e05b33bfac43833c7b95fd7b7c3188f8
Author: Howard Guo <hguo@suse.com>
Date: Thu Nov 10 15:39:03 2016 +0100
This means starting with OpenLDAP 2.4.45 we can drop the explicit CA
configuration in ldap.conf.
There are several use cases where an explicit IPA CA should be specified
in the configuration. These mostly concern situations where a higher
security level must be maintained. For these configurations an
administrator would need to add an explicit CA configuration to
ldap.conf if we wouldn't add it during the ipa-client-install setup.
RN: FreeIPA client installer does not add explicit TLS CA configuration
RN: to OpenLDAP's ldap.conf anymore. Since OpenLDAP 2.4.45, explicit CA
RN: configuration is not required as OpenLDAP uses the default CA store
RN: provided by OpenSSL and IPA CA is installed in the default store
RN: by the installer already.
Fixes: https://pagure.io/freeipa/issue/9258
Signed-off-by: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
While python-ldap is strict dependency of IPA in downstreams, it
is optional for IPA packages published on PyPI.
Openldap 2.6 no longer ships ldap_r-2, that makes
ipapython.dn_ctypes not working against such environments.
Thanks @abbra!
Fixes: https://pagure.io/freeipa/issue/9255
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Rob Crittenden
Stanislav Levin
Filip Dvorak
Florence Blanc-Renaud
Julien Rische
Christian Heimes
Mohammad Rizwan
Erik Belko
Scott Poore
Anuja More
Sumedh Sidhaye
Alexander Bokovoy