In 2012, ldbm backend in 389-ds started checking entry modification
after running betxnpreop plugins by comparing a number of modifications
before and after. If that number didn't change, it is considered that
plugins didn't modify the list.
ipa-pwd-extop actually removed and re-added modification to ipaNTHash if
it contained 'MagicRegen' value. This did not work since commit
https://pagure.io/389-ds-base/c/6c17ec56076d34540929acbcf2f3e65534060a43
but we were lucky nothing in FreeIPA code actually relied on that except
some code paths in ipasam Samba passdb driver. However, Samba didn't
reach the point where the code was triggered -- until now.
With support to run Samba as a domain member in IPA domain, that code
path is triggered for Kerberos service principals of domain members
(cifs/client.example.test, ...) and NT hash extraction from Kerberos
keys does not work.
Fix ipa-pwd-extop to follow recommendations in
https://pagure.io/389-ds-base/issue/387#comment-120145 and
https://pagure.io/389-ds-base/issue/50369#comment-570696
Fixes: https://pagure.io/freeipa/issue/7953
Reviewed-By: Christian Heimes <cheimes@redhat.com>
In order to test a fix for https://pagure.io/freeipa/issue/7953,
we need to create a keytab with a particular encryption type
(arcfour-hmac) and attempt to request generation of ipaNTHash attribute
from Kerberos keys in LDAP.
Add a test case that performs this operation.
Related: https://pagure.io/freeipa/issue/7953
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Convert use_keytab() function into a context manager to allow additional
operations to be done as part of the test. Also pass proper credentials
cache file to the backend while connecting to LDAP so that right creds
are in use.
This is required to perform actual tests for use of the retrieved keys.
Related: https://pagure.io/freeipa/issue/7953
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Rewrite templates to make test job declarations simpler and easier to
work with.
A test job template can be instantiated this way:
- template: templates/test-jobs.yml
parameters:
jobName: Base
jobTitle: Base tests
testsToRun:
- test_cmdline
- test_install
- test_ipaclient
- test_ipalib
- test_ipaplatform
- test_ipapython
testsToIgnore:
- test_integration
- test_webui
- test_ipapython/test_keyring.py
taskToRun: run-tests
Both 'testsToRun' and 'testsToIgnore' accept arrays of test matches.
Wildcards also supported:
....
testsToRun:
- test_xmlrpc/test_hbac*
....
'taskToRun' specifies a script ipatests/azure/azure-${taskToRun}.sh that
will be executed in the test environment to actually start tests.
Parameters 'testsToRun' and 'testsToIgnore' define TESTS_TO_{RUN,IGNORE}
variables that will be set in the environment of the test script. These
variables will have entries from the parameters separated by a space.
Reviewed-By: Christian Heimes <cheimes@redhat.com>
ipa-run-tests expands arguments passed with their full paths. However,
it doesn't support expanding globs, so targets like 'test_ipa*' cannot
be specified.
Expand the code that replaces '--ignore foo' and 'foo' positional
arguments with support for '--ignore foo*' and 'foo*'.
This allows to reduce a number of additional steps in the CI pipeline
preparation.
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Order of master/replica was incorect which has been
corrected
Signed-off-by: Kaleemullah Siddiqui <ksiddiqu@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
Add a test checking that the entry cn=CAcert,cn=ipa,cn=etc,BASEDN
is properly updated when the CA is renewed
The test also checks that the entry
cn=DOMAIN IPA CA,cn=certificates,cn=ipa,cn=etc,BASEDN properly
contains ipaconfigstring: compatCA
ipaconfigstring: ipaCA
Related to https://pagure.io/freeipa/issue/7928
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
The FreeIPA uses its own classes for managing temp files and
directories for tests. One of its underlying low-level functions
is `mkdtemp`.
According to documentation for `mkdtemp`:
```
If dir is not None, the file will be created in that directory; otherwise, a
default directory is used. The default directory is chosen from a
platform-dependent list, but the user of the application can control the
directory location by setting the TMPDIR, TEMP or TMP environment variables.
```
It's actually the truth,
/usr/lib64/python3.7/tempfile.py:
```
def _candidate_tempdir_list():
"""Generate a list of candidate temporary directories which
_get_default_tempdir will try."""
dirlist = []
# First, try the environment.
for envname in 'TMPDIR', 'TEMP', 'TMP':
dirname = _os.getenv(envname)
if dirname: dirlist.append(dirname)
# Failing that, try OS-specific locations.
if _os.name == 'nt':
dirlist.extend([ _os.path.expanduser(r'~\AppData\Local\Temp'),
_os.path.expandvars(r'%SYSTEMROOT%\Temp'),
r'c:\temp', r'c:\tmp', r'\temp', r'\tmp' ])
else:
dirlist.extend([ '/tmp', '/var/tmp', '/usr/tmp' ])
```
For now, there is a hardcoded assertion of a temp directory (`/tmp`) in
IPA tests. But some systems use the mentioned environment variables
(for example, pam_mktemp https://www.openhub.net/p/pam_mktemp).
It's easy to check an actual temp dir via `gettempdir`.
Fixes: https://pagure.io/freeipa/issue/7956
Signed-off-by: Stanislav Levin <slev@altlinux.org>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
The test runs rpm -V in order to check that the file
permissions are consistent with the expectations set
in the spec file. The file mode, owner and group are
checked.
Related to https://pagure.io/freeipa/issue/7934
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
The ipatests/test_integration/ package only uses ipaserver in a few
places. Copy some simple constants to decouple the packages.
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
If TESTS_TO_RUN contains a single parameter that cannot be expanded,
bash will not perform brace elimination. Remove braces manually.
For example, TESTS_TO_RUN='test_xmlrpc/test_*.py' will not expand
outside of ipatests and the script would generate
tests_to_run=-k{test_xmlrpc/test_*.py}
Braces then will prevent actual ipa-run-tests execution from matching
any of XMLRPC tests.
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Sets up a pipeline to run FreeIPA build and tests in Azure Pipelines.
Azure Pipelines provides 10 parallel free runners for open source projects.
Use them to run following jobs:
- Build: build RPMs and Fedora 30 container with them
- Lint: run linting of the source code
- Tox: run py36,pypi,pylint tests using Tox
- Web UI unit tests: run Web UI unit tests with Grunt/QUnit/PhantomJS
- XMLRPC tests: install FreeIPA server and run XMLRPC tests against it
All jobs are running in Fedora 30 containers. Build, Lint, Tox, and Web
UI unit tests run inside f30/fedora-toolbox container. Build job
generates a container with pre-installed FreeIPA packages using official
fedora:30 container. All containers are picked up from
registry.fedoraproject.org.
Artifacts from the build job are pushed to a pipeline storage and reused
in the XMLRPC tests. They also are accessible in the 'Summary' tab to
download.
XUnit and QUnit outputs from the tests that produce it are reported in
the 'Tests' tab.
Logs from individual steps from each job are available for review in
the 'Logs' tab. They also can be downloaded.
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
This commit tests the--idmap-domain knob with the following behavior:
- if not present, default to IDM domain (current behavior)
- if equal to DNS (magic value), set nothing and let idmapd autodetect domain
- otherwise set Domain in idmap.conf to the value passed by this parameter
Related to: https://pagure.io/freeipa/issue/7918
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
Under some conditions, ipa config-show and several other commands were
failing with error message:
ERROR: invalid 'PKINIT enabled server': all masters must have IPA master role enabled
Amongst others the issue can be caused by a broken installation, when
some services are left in state 'configuredServices'. The problem even
block uninstallation or removal of replicas. Now configured servers are
also consider valid providers for associated roles.
A new test verifies that config-show works with hidden and configured HTTP
service.
Remark: The original intent of the sanity check is no longer clear to me. I
think it was used to very that all services can be started by ipactl.
Since ipactl starts hidden, configured, and enabled services, the new
logic reflect the fact, too.
Fixes: https://pagure.io/freeipa/issue/7929
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
python-ldap is a Python package with heavy C extensions. In order to
build python-ldap, not only OpenLDAP development headers are necessary,
but also OpenSSL, Cyrus SASL, and MIT KRB5 development headers.
A fully functional ipaclient doesn't need an LDAP driver. It talks JSON
RPC over HTTPS to a server. python-ldap is only used by ipapython.dn.DN
to convert a string to a DN with ldap_str2dn(). The function is simple
and can be wrapped with ctypes in a bunch of lines.
Related: https://pagure.io/freeipa/issue/6468
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Look for SELinux violation after installing a master with CA, KRA, and
DNS with DNSSEC. The test does not fail yet, because there are known
SELinux violations.
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The p11-kit configuration injects p11-kit-proxy into all NSS databases.
Amongst other p11-kit loads SoftHSM2 PKCS#11 provider. This interferes
with 389-DS, certmonger, Dogtag and other services. For example certmonger
tries to open OpenDNSSEC's SoftHSM2 token, although it doesn't use it at
all. It also breaks Dogtag HSM support testing with SoftHSM2.
IPA server does neither need nor use SoftHSM2 proxied by p11-kit.
Related: https://pagure.io/freeipa/issue/7810
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
The object logger methods been deprecated for about two years since release
4.6.0. The log manager used to moneky-patch additional log methods like
info(), warning(), and error() into API plugin objects. The methods have
been replaced by calls to module logger objects in 4.6.0.
Remove monkey-patch logger methods, log manager, and its root logger from
ipapython.ipa_log_manager.
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Thomas Woerner <twoerner@redhat.com>
On non-usrmerge systems (e.g., Debian), bash, mv, cp, cat, tail,
keyctl, and gzip live in /bin, not /usr/bin.
On usrmerge systems, /bin is a symlink to /usr/bin (or vice versa), so
this has no effect.
Signed-off-by: Robbie Harwood <rharwood@redhat.com>
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
With the latest NFS changes:
* systemd NFS-related unit files
* configuration from /etc/sysconfig/nfs to /etc/nfs.conf
testing NFS client {manual, ipa-client-automount} configuration
has become paramount.
This extends the existing automount location test and must be
run nightly.
Fixes: https://pagure.io/freeipa/issue/7805
Signed-off-by: François Cami <fcami@redhat.com>
Reviewed-By: Peter Cech <pcech@redhat.com>
Reviewed-By: Rob Crittenden <rcritten@redhat.com>
IPA used to write a custom /etc/resolv.conf. On Fedora and RHEL,
NetworkManager is typically maintaining resolv.conf. On reboot or
restart of the service, NM overwrites the custom settings.
On systems with NM enabled, the DNS server installer now drops a config
file into NM's global config directory and delegates resolv.conf to NM.
On systems without NM, fall back to create /etc/resolv.conf directly.
Fixes: https://pagure.io/freeipa/issue/7900
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Thomas Woerner <twoerner@redhat.com>
Vault and KDC proxy are neither critical subsystems nor are they likely to
fail. They have been pretty stable and don't see any major development.
It's sufficient to run them in nightly tests only.
The removal speed up gating a bit. Especially vault tests are slow and
usually take more than 30 minutes to complete
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Florence Blanc-Renaud <frenaud@redhat.com>
389-DS 1.4.0.22 was pushed to Fedora over the weekend. The new versin
breaks test_find_orphan_automember_rules. Skip the test case for now
until we have more time to investigate the issue.
Related: https://pagure.io/freeipa/issue/7902
Signed-off-by: Christian Heimes <cheimes@redhat.com>
Reviewed-By: Christian Heimes <cheimes@redhat.com>
Tests in test_trust.py were organized in ten classes, one for each trust type,
requiring ten cycles of ipaserver installation/uninstallation and the full test
run lasted for about 5500 seconds.
There is no need in reinstallation of ipaserver between establishing different
types of trust.
I moved all tests to sinle class, preserving test logic.
Additional changes:
* TestEnforcedPosixADTrust was totally removed as it was duplicate of
TestPosixADTrust
* code of repeated checks was moved to methods
* A task was cretated for cleaning up DNS configuration changes made for
establishing trust
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
AD hosts can now be extracted from list in respective class attributes and host
domain names -- from properties provided by multihost plugin (host.domain.name).
Also removed conditional skips of tests when test configuration contains only
part of required AD machines as this feature never worked:
multihost plugin removes all machines from config which are not explicitly
requested.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Tests for AD trust can use three types (roles) of AD machines:
forest root, subdomain and tree domain.
All those machines were placed in one domain object of multihost configuration,
though they all have different domain names.
This is bad as we can not use domain attributes provided by multihost plugin
like host.domain.name and host.domain.basedn and others and need to reimplement
them, evaluating domain name from host.hostname.
And if we accidently used those properties it would lead to difficult to locate
errors (we would use same domain name for all AD hosts).
I modified multihost fixture function mh() to allow creating several AD domains.
As multihost plugin does not support requesting multiple domains with the same type,
I had to introduce new domain types: AD_SUBDOMAIN and AD_TREEDOMAIN.
Also there was a error in mh() which forced user to provide all three AD
machines when only one was needed (value from test class property num_ad_domains
was applied to subdomains and treedomains requirement).
I changed this behavior and now additional AD machines are specified with
properties num_ad_subdomains and num_ad_treedomains.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
When synchronizing time with windows server using chronyd I often see
error "No suitable source for synchronisation". This happens because chronyd
with default options refuses to use time servers with big jitter and delay.
For some reasons Windows time server does have big jitter. In some test setups
delay also can be rathe big.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Test was expecting that when trust is established with forest root, than all
three AD domains should be found when quering trust-find for that domain.
Actually only root domain and its subdomain should be returned, without
the tree domain.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
TestExternalTrustWithRootDomain was inherited from ADTrustSubdomainBase
This caused that external trust was checked two times with subdomain
and was not checked with root domain.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
Before establishing trust with AD it is recommended in documentation
(and for many setups necessary) to create add DNS forwarder for AD domain.
Bind config supplied by ipa server has dnssec validation enabled.
If Windows server DNS does not have DNSSEC enabled with valid certificate,
then bind will not be able to use it as forwarder and trust will not be
established.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
It was changed in f487233df0
for unknown reason. It did not influence test runs as
configure_dns_for_trust was made no-op in previous commit
1d9e1521c5. As now this commit is reverted,
configure_dns_for_trust is restored, invocation parameters also need to
be changed to initial values.
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>
This reverts commit 1d9e1521c5.
The reverted commit message states:
"Since DNS configuration is no longer needed for running trust tests,
this method's contents are removed."
In fact tests can run without DNS configuration only in case if test setup
has a DNS server with DNSSEC support and there are A records for Windows
machines and SRV records Windows AD services and this DNS server is used
as forwarder by bind. If one of these in not true
then tests fail when trying to establish trust (ipa trust-add) as --server
option is not used and ipa can not find the AD machine. If we specify
--server option and add Windows hosts to /etc/hosts, then trust will be
established, but then sssd will fail to find the host to talk for getting users
from AD. So for general case we should setup DNS forwarders prior to
establishing trust, as stated in
https://www.freeipa.org/page/Active_Directory_trust_setup
Related to https://pagure.io/freeipa/issue/7889
Reviewed-By: Alexander Bokovoy <abokovoy@redhat.com>