| elodilles | hi, fyi, a lot of projects' openstack-tox-docs job started to fail due to the latest oslo.policy release (?) | 06:27 |
|---|---|---|
| elodilles | the error log is this usually: https://zuul.opendev.org/t/openstack/build/c1504d60e39f41b7ac022d952b913168/log/job-output.txt#1477-1480 | 06:28 |
| elodilles | this patch could be connected to the issue: https://review.opendev.org/c/openstack/oslo.policy/+/830514 | 06:28 |
| elodilles | any suggestion about how to fix these / any way forward? | 06:29 |
| elodilles | (i did not investigate the issue deeper yet) | 06:29 |
| hberaud | elodilles: without too digging that seems to be an oslo.config issue | 08:08 |
| hberaud | elodilles: BTW, thanks for the heads up | 08:08 |
| elodilles | hberaud: no problem :) i thought it's oslo.policy as that is in the last log lines o:) | 08:28 |
| hberaud | elodilles: well this new oslo.policy option seems to trigger the oslo.context error. When I follow step by step the execution it fail when we generate the config (with oslo.config) to generate the policy sample. | 08:43 |
| hberaud | I don't know yet why this option seems to be a problem for oslo.config | 08:44 |
| opendevreview | Rajat Dhasmana proposed openstack/oslo.policy master: Only pass exclude-deprecated when True https://review.opendev.org/c/openstack/oslo.policy/+/839711 | 09:18 |
| ralonsoh | elodilles, hberaud ^^ right, this seems to be the problem | 09:19 |
| hberaud | ralonsoh: I tested your patch and indeed it fix the glance tox docs target | 09:24 |
| hberaud | ralonsoh: kudos | 09:24 |
| ralonsoh | not mine | 09:24 |
| hberaud | oh | 09:24 |
| hberaud | ah yes indeed | 09:26 |
| elodilles | ralonsoh: thanks! looks promising! | 09:31 |
| gibi | ralonsoh: thanks for the fix, it works for the nova doc job too | 09:50 |
| ralonsoh | gibi, hehehe the fix is not mine | 09:50 |
| ralonsoh | but works! | 09:50 |
| gibi | ahh yes | 09:50 |
| gibi | thanks Rajat! | 09:51 |
| opendevreview | Rajat Dhasmana proposed openstack/oslo.policy master: Only pass exclude-deprecated when True https://review.opendev.org/c/openstack/oslo.policy/+/839711 | 09:57 |
| opendevreview | Jorhson Deng proposed openstack/oslo.cache master: Support SASL protocol for memcached https://review.opendev.org/c/openstack/oslo.cache/+/742060 | 10:50 |
| hberaud | stephenfin, damani: please can review this patch https://review.opendev.org/c/openstack/oslo.policy/+/839711 | 11:54 |
| hberaud | stephenfin, damani: Once released this patch will unlock many gates | 12:00 |
| hberaud | See the above discussion | 12:00 |
| damani | hberaud, done and merge | 12:02 |
| hberaud | thanks | 12:02 |
| sean-k-mooney | by the way removing a paramater form oslo.polies config generator is a backward incompatible change so that should have been a majory version bump when it was released | 13:05 |
| sean-k-mooney | so the next release shoudl proably do a major version bump when you ship the fix to signal that | 13:06 |
| sean-k-mooney | hum i guess that was just intoduce in the most recent release? | 13:08 |
| sean-k-mooney | with https://review.opendev.org/c/openstack/oslo.policy/+/830514 | 13:08 |
| sean-k-mooney | i guess there has not been time for peopel to start dependin on --exclude-deprecated false | 13:10 |
| sean-k-mooney | so we might be able to get away with a minor versios or even a patch fix eventhough strictly specing it woudl be a major bump | 13:10 |
| sean-k-mooney | hberaud: there is just one oslo-specs repo right so its the right one for oslo.messaging | 13:26 |
| hberaud | sean-k-mooney: yes | 13:26 |
| sean-k-mooney | cool just creating the spec for the oslo messaging unix socket notifiction driver i talked about a few months ago | 13:26 |
| sean-k-mooney | huh looks like no one has added a driver since mitaka | 13:28 |
| sean-k-mooney | at least not with a spec | 13:28 |
| sean-k-mooney | do you somethime just use specless blueprints | 13:29 |
| sean-k-mooney | i proably shoudl ask that before i write a spec | 13:29 |
| hberaud | sean-k-mooney: AFAIK nope, we usually try to write specs (my experience is more related to oslo.cache) | 13:38 |
| sean-k-mooney | ack no worreis | 13:43 |
| opendevreview | Merged openstack/oslo.policy master: Only pass exclude-deprecated when True https://review.opendev.org/c/openstack/oslo.policy/+/839711 | 13:50 |
| hberaud | sean-k-mooney: concerning the oslo.policy stuff I'm not sure to see which backward incompatible change you mean? the `--exclude-deprecated false`? | 14:22 |
| sean-k-mooney | it is being chagne form a optional parmater that took True|False to a parmater that takes no argurments | 14:24 |
| sean-k-mooney | so --exclude-deprecated false woudl no longer be valid correct | 14:25 |
| sean-k-mooney | or did i missunderstand | 14:25 |
| hberaud | If yes, IMO this a new feature with a default value so I'm not sure we could consider that as an incompatible change. IMO the problem is more on the sample generator and not related to the option itself | 14:25 |
| hberaud | The option was not changed it was only added | 14:25 |
| sean-k-mooney | it was added in one patch and then modifed in the more recent one that just merged | 14:26 |
| sean-k-mooney | and presumable there was a release in between becasue it broke ci jobs | 14:26 |
| hberaud | the option remains the same between those patches, only its usage in the sphinx plugin have been changed | 14:27 |
| sean-k-mooney | ok so you can still pass false if you want too | 14:28 |
| hberaud | yes | 14:28 |
| sean-k-mooney | and that will have the expected behavior? if so how did the sphix usage break things | 14:28 |
| hberaud | because the sphinx plugin passed the bool value to oslo.config that tried to use it as a string | 14:29 |
| sean-k-mooney | ah so type error as a result | 14:31 |
| hberaud | My point is that I planed to release a bugfix version because I think these changes are backward compat | 14:32 |
| hberaud | does it make sense for you? | 14:32 |
| sean-k-mooney | yep | 14:32 |
| sean-k-mooney | i tought ye were removing the ablity to pass the bool on the cli | 14:32 |
| sean-k-mooney | so i just missread the commit message | 14:32 |
| hberaud | For the first patch I didn't released a major version because from my POV only a new option with a defaut value was added | 14:32 |
| hberaud | np you did well to talk about it | 14:33 |
| hberaud | I prefered to double check with you | 14:33 |
| sean-k-mooney | yep the first patch is just a feature release not a major and the second is not removign the value form the cli so also fine | 14:36 |
| sean-k-mooney | its just a patch release | 14:36 |
| sean-k-mooney | so a z bump in x.y.z numbering | 14:36 |
| hberaud | yes | 14:36 |
| hberaud | sean-k-mooney: https://review.opendev.org/c/openstack/releases/+/839775 | 14:36 |
| opendevreview | sean mooney proposed openstack/oslo-specs master: [oslo.messaging] add unix socket notifcation driver https://review.opendev.org/c/openstack/oslo-specs/+/839819 | 19:31 |
| sean-k-mooney | stephenfin: hberaud: ^ im not really sure what else to add to that. i was going to see if i could create WIP verions early next week | 19:32 |
| sean-k-mooney | maybe the week after actully since monday is a bank holiday and i have some doctors appointes mid week | 19:33 |
| opendevreview | sean mooney proposed openstack/oslo-specs master: [oslo.messaging] add unix socket notifcation driver https://review.opendev.org/c/openstack/oslo-specs/+/839819 | 19:36 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!