*** markvoelker has quit IRC | 00:54 | |
*** markvoelker has joined #refstack | 00:54 | |
*** markvoelker has quit IRC | 00:59 | |
*** tosky has quit IRC | 01:17 | |
*** markvoelker has joined #refstack | 01:55 | |
*** markvoelker has quit IRC | 02:46 | |
*** markvoelker has joined #refstack | 02:46 | |
*** lifeless has joined #refstack | 03:10 | |
*** markvoelker has quit IRC | 08:08 | |
*** markvoelker has joined #refstack | 08:08 | |
*** markvoelker has quit IRC | 08:13 | |
*** kopecmartin|off is now known as kopecmartin | 08:28 | |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: Added python-tempestconf-tempest-devstack-py3 https://review.openstack.org/622865 | 08:39 |
---|---|---|
*** tosky has joined #refstack | 08:47 | |
tosky | kopecmartin: in https://review.openstack.org/#/c/627722/, did you mean "Why should *not* we use the older version?" | 09:22 |
kopecmartin | tosky, yes, your question means the same but from the other point of view .. the thing is, there is no description, no explanation of why the change is needed .. | 09:24 |
kopecmartin | oh wait, there is an answer, a moment | 09:24 |
kopecmartin | I see, i misunderstood the signs (< > =) :D so the change is updating to the new version | 09:26 |
tosky | :) | 09:27 |
kopecmartin | tosky, btw, do we need to have a top limit there? <1.2.0 | 09:27 |
tosky | it seems that RDO rocky has 1.1.0: https://cbs.centos.org/koji/buildinfo?buildID=23649 and Queens has 1.0.0+: https://cbs.centos.org/koji/buildinfo?buildID=21553 | 09:27 |
tosky | kopecmartin: I fear yes, because a new version may introduce stricter rule and introduce sudden failures | 09:28 |
tosky | which is not necessarily a bad thing, but I think that normally people are limiting the version of hacking | 09:28 |
kopecmartin | ok, makes sense | 09:28 |
tosky | for example, I had to apply some fixes and blacklist few parsing rules when I tried to bump the version on hacking in sahara recently: https://review.openstack.org/#/c/567164/ | 09:30 |
openstackgerrit | Merged openstack/python-tempestconf master: Replace tripleo-scenario002-multinode with scenario002-standalone https://review.openstack.org/628258 | 10:05 |
*** markvoelker has joined #refstack | 10:09 | |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: Fix diff in gates https://review.openstack.org/628180 | 10:09 |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: Added python-tempestconf-tempest-devstack-admin-py3 https://review.openstack.org/622865 | 10:09 |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: Added python-tempestconf-tempest-devstack-admin/demo-py3 https://review.openstack.org/622865 | 10:22 |
kopecmartin | arxcruz, chandankumar tosky about the functional tests https://review.openstack.org/#/c/626539/ do we want to have it? | 10:30 |
kopecmartin | I created tests for --append and --remove based on chandankumar request, suddenly, it turned out to be a role :D | 10:30 |
arxcruz | i have no idea | 10:31 |
kopecmartin | do you see benefits in having the role and the job? | 10:31 |
arxcruz | if we plan to have other tests, yes | 10:31 |
arxcruz | isn't py2/py3 jobs already cover it ? | 10:31 |
tosky | as I mentioned: IMHO it must be a separate role | 10:31 |
tosky | we can discuss whether using on the existing jobs or not | 10:31 |
kopecmartin | arxcruz, --remove and --append are not covered in the existing jobs | 10:32 |
kopecmartin | tosky, makes sense having it in a separate role | 10:33 |
kopecmartin | so maybe we could include the role in our existing jobs and in future if the role becomes bigger we can move it to a separate job? | 10:33 |
tosky | is there anything else that we test in the main job but where we don't touch all the use cases? Or is it anything else that we don't test of the public CLI (because the internal API should be covered also by unit tests) and that we should? | 10:35 |
kopecmartin | there are a few arguments which are not used in our jobs, e.g. --generate-profile, --image-disk-format, --network-id .. | 10:41 |
kopecmartin | but i have no idea if it's worth to test them too | 10:42 |
openstackgerrit | Merged openstack/python-tempestconf master: Update hacking version https://review.openstack.org/627722 | 10:47 |
tosky | maybe --generate-profile - at least we need some unit tests (there are for read) | 10:48 |
tosky | and what about the existing parameters? Are there combinations of parameters that we don't check? | 10:49 |
kopecmartin | tosky, i don't think more unit tests for generate-profile are needed, the functions which made sense to be tested by unit tests are already tested | 11:00 |
kopecmartin | tosky, which parameters you mean? | 11:00 |
tosky | the ones we already had | 11:00 |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: Added python-tempestconf-tempest-devstack-admin/demo-py3 https://review.openstack.org/622865 | 11:02 |
kopecmartin | tosky, CLI arguments? there are a few which are not tested, but they basically sets a value, nothing complicated | 11:02 |
kopecmartin | not tested by gates | 11:03 |
tosky | uhm | 11:06 |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: [DNM] testing os_tempest https://review.openstack.org/627482 | 11:31 |
arxcruz | chandankumar: tosky regarding https://review.openstack.org/#/c/622865/ we should create story to remove py2 job in near future | 12:39 |
arxcruz | or will have support unofficial to python 2 in openstack? | 12:39 |
*** lifeless has quit IRC | 12:42 | |
chandankumar | arxcruz: we donot know when upstream is going to remove python2 support but a story would be ok to create for tracking | 12:44 |
arxcruz | chandankumar: https://pythonclock.org/ is my guess | 12:45 |
openstackgerrit | Chandan Kumar proposed openstack/python-tempestconf master: [DNM] testing os_tempest https://review.openstack.org/627482 | 12:46 |
openstackgerrit | Merged openstack/python-tempestconf master: Fix diff in gates https://review.openstack.org/628180 | 13:03 |
chandankumar | arxcruz: kopecmartin tosky https://review.openstack.org/#/c/622865/ can we merge py3 stuff? | 13:04 |
*** markvoelker has quit IRC | 13:33 | |
*** markvoelker has joined #refstack | 13:33 | |
openstackgerrit | Martin Kopec proposed openstack/refstack-client master: Fix generate-tempestconf-file role https://review.openstack.org/628446 | 14:30 |
openstackgerrit | Merged openstack/python-tempestconf master: Added python-tempestconf-tempest-devstack-admin/demo-py3 https://review.openstack.org/622865 | 14:52 |
chandankumar | \o/ | 14:54 |
*** arxcruz is now known as arxcruz|brb | 16:11 | |
openstackgerrit | Martin Kopec proposed openstack/refstack-client master: Fix generate-tempestconf-file role https://review.openstack.org/628446 | 16:13 |
openstackgerrit | Martin Kopec proposed openstack/python-tempestconf master: DNM - Testing refstack gate https://review.openstack.org/628639 | 17:25 |
chandankumar | kopecmartin: Hey | 17:31 |
kopecmartin | chandankumar, what's up? | 17:31 |
chandankumar | kopecmartin: We are seeing a lot of repeatition in our python-temepstconf roles | 17:32 |
chandankumar | kopecmartin: the install parts are copied everywhere | 17:32 |
chandankumar | kopecmartin: I think we can move it to a single role | 17:32 |
kopecmartin | what were? do you have any example? | 17:32 |
chandankumar | kopecmartin: one min | 17:33 |
chandankumar | kopecmartin: https://github.com/openstack/python-tempestconf/blob/master/roles/generate-tempestconf-file-cloud/tasks/main.yaml#L2 to l26 | 17:33 |
chandankumar | kopecmartin: https://github.com/openstack/python-tempestconf/blob/master/roles/generate-tempestconf-file/tasks/main.yaml#L5 to l29 | 17:34 |
chandankumar | kopecmartin: did you get that? | 17:34 |
kopecmartin | yes , but is it worth it to create a new role for only these two cases? | 17:35 |
chandankumar | kopecmartin: https://review.openstack.org/#/c/626539/7/roles/functional-tests/tasks/main.yaml | 17:36 |
tosky | do you mean a new job? | 17:36 |
chandankumar | kopecmartin: what do you think repeating same code at three places | 17:36 |
chandankumar | tosky: nope | 17:37 |
chandankumar | tosky: just moving the install part in a single role | 17:37 |
tosky | ok, but then why are those role different? | 17:37 |
tosky | if they do almost the same thing? | 17:38 |
kopecmartin | chandankumar, three places if we merge that patch :D but i see your point | 17:38 |
chandankumar | tosky: because they do some different stuff while testing python tempestconf named args | 17:38 |
tosky | ok, I agree that the common parts deserves a role | 17:39 |
chandankumar | kopecmartin: hehe | 17:40 |
kopecmartin | let's create a storyboard for that | 17:40 |
kopecmartin | *a story in our storyboard | 17:40 |
chandankumar | kopecmartin: will create a tomorrow | 17:40 |
kopecmartin | ok | 17:40 |
chandankumar | kopecmartin: one more thing with whom you are talking with redhat certification work? | 17:42 |
chandankumar | kopecmartin: generally I go to office, I will catch them to understand their problem | 17:42 |
chandankumar | kopecmartin: is it Anshul or somebody else? | 17:42 |
tosky | chandankumar, kopecmartin maybe in private? | 17:46 |
chandankumar | tosky: yes done! | 17:46 |
*** kopecmartin is now known as kopecmartin|off | 17:58 | |
*** tosky has quit IRC | 18:09 | |
*** lifeless has joined #refstack | 20:12 | |
*** openstack has joined #refstack | 23:46 | |
*** ChanServ sets mode: +o openstack | 23:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!