According to our records, this should be resolved by python-manila-tests-tempest-1.4.0-2.20210629204822.f083c71.el8ost. This build is available now.
Tested using: python3-manila-tests-tempest-1.8.0-0.20220427151611.ee15891.el9ost.noarch There are still 2 tests that generate random names. There is already a patch that was merged upstream but need to be back-ported downstream: https://review.opendev.org/c/openstack/manila-tempest-plugin/+/840282
(In reply to lkuchlan from comment #2) > Tested using: > python3-manila-tests-tempest-1.8.0-0.20220427151611.ee15891.el9ost.noarch > > There are still 2 tests that generate random names. > There is already a patch that was merged upstream but need to be back-ported > downstream: > https://review.opendev.org/c/openstack/manila-tempest-plugin/+/840282 Even using that patch, there are a few tests which inject a full random ssh key in the test signature whose signature will continue to change. No patch for that at this point, probably a discussion is needed to find out whether exposing that information in the test signature is relevant or not.
(In reply to Luigi Toscano from comment #3) > (In reply to lkuchlan from comment #2) > > Tested using: > > python3-manila-tests-tempest-1.8.0-0.20220427151611.ee15891.el9ost.noarch > > > > There are still 2 tests that generate random names. > > There is already a patch that was merged upstream but need to be back-ported > > downstream: > > https://review.opendev.org/c/openstack/manila-tempest-plugin/+/840282 > > Even using that patch, there are a few tests which inject a full random ssh > key in the test signature whose signature will continue to change. No patch > for that at this point, probably a discussion is needed to find out whether > exposing that information in the test signature is relevant or not. Oh, i'm not aware of any test needing an SSH key as data. Do you mean this one? https://github.com/openstack/manila-tempest-plugin/blob/86f2432476ba9128bb2c26a78ae3b3f53327d4bd/manila_tempest_tests/tests/api/test_access_rules_metadata_negative.py#L71-L79
I found 2 tests that generate random names because one of the parameters generates a random value: https://github.com/openstack/manila-tempest-plugin/blob/86f2432476ba9128bb2c26a78ae3b3f53327d4bd/manila_tempest_tests/tests/api/admin/test_share_types.py#L117 https://github.com/openstack/manila-tempest-plugin/blob/149acc51adf0c17b22fe690a2e9018596bcc5bbd/manila_tempest_tests/tests/api/test_share_types_negative.py#L91 In the new version(1.6.0) of ddt there is an option to define a name for the test without depending on the given values: https://github.com/datadriventests/ddt/blob/2125da16830c69c87f237640e617729ffcbf235a/ddt.py#L416 This can be a solution for these cases.
Tested using: python3-manila-tests-tempest-1.4.0-2.20210629204822.f083c71.el8ost.noarch There is one more patch that didn't land downstream: https://review.opendev.org/c/openstack/manila-tempest-plugin/+/855813
(In reply to lkuchlan from comment #10) > Tested using: > python3-manila-tests-tempest-1.4.0-2.20210629204822.f083c71.el8ost.noarch > > There is one more patch that didn't land downstream: > https://review.opendev.org/c/openstack/manila-tempest-plugin/+/855813 That patch can't land downstream because we use a different version of ddt. ddt 1.6.0 is only available upstream from stable/zed.
(In reply to Luigi Toscano from comment #11) > (In reply to lkuchlan from comment #10) > > Tested using: > > python3-manila-tests-tempest-1.4.0-2.20210629204822.f083c71.el8ost.noarch > > > > There is one more patch that didn't land downstream: > > https://review.opendev.org/c/openstack/manila-tempest-plugin/+/855813 > > That patch can't land downstream because we use a different version of ddt. > ddt 1.6.0 is only available upstream from stable/zed. Is it make sense to close it and move it to VERIFIED even though the randomness isn't completely resolved ? Because I don't see any other way to solve it.
(In reply to lkuchlan from comment #12) > (In reply to Luigi Toscano from comment #11) > > (In reply to lkuchlan from comment #10) > > > Tested using: > > > python3-manila-tests-tempest-1.4.0-2.20210629204822.f083c71.el8ost.noarch > > > > > > There is one more patch that didn't land downstream: > > > https://review.opendev.org/c/openstack/manila-tempest-plugin/+/855813 > > > > That patch can't land downstream because we use a different version of ddt. > > ddt 1.6.0 is only available upstream from stable/zed. > > Is it make sense to close it and move it to VERIFIED even though the > randomness isn't completely resolved ? > Because I don't see any other way to solve it. That last patch can't work on 16.x (and maybe not even in 17.x), so - either the bug stays on 16.2.4 but another solution is needed, - or the scope of the bug, still on 16.2.4, is reduced, - or the bug is retargeted to 18, or 17 if ddt can be bumped there - that upstream patch may be problematic for 17.x too.
There are still two test cases that generate random signatures. Fixing them requires version 1.6.0 of ddt, however, in OSP 16.2 the packaged version is 1.2.0. Since it's an old version, we move this bz to VERIFIED and clone it to OSP 18.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Release of components for Red Hat OpenStack Platform 16.2.4), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2022:8794