| *** jamielennox is now known as jamielennox|away | 00:58 | |
| *** jamielennox|away is now known as jamielennox | 01:12 | |
| *** gouthamr has quit IRC | 02:33 | |
| *** yfried has joined #openstack-shade | 06:00 | |
| yfried | hi, I'm having trouble with shade requirements for the last week | 06:30 |
|---|---|---|
| yfried | anyone seeing something similar? | 06:30 |
| yfried | anyone seeing issues with shade and pbr? | 06:42 |
| *** yfried has quit IRC | 08:27 | |
| *** yfried has joined #openstack-shade | 08:27 | |
| *** ioggstream has joined #openstack-shade | 08:37 | |
| *** yolanda has quit IRC | 08:37 | |
| *** yolanda has joined #openstack-shade | 08:42 | |
| *** cdent has joined #openstack-shade | 11:09 | |
| *** ioggstream has quit IRC | 11:23 | |
| *** Matias_ has quit IRC | 11:44 | |
| *** Matias_ has joined #openstack-shade | 11:45 | |
| *** gouthamr has joined #openstack-shade | 12:52 | |
| -openstackstatus- NOTICE: Gerrit is going to be restarted due to performance problems | 13:41 | |
| *** ChanServ changes topic to "Gerrit is going to be restarted due to performance problems" | 13:41 | |
| *** yfried has quit IRC | 13:45 | |
| -openstackstatus- NOTICE: Gerrit has been successfully restarted | 13:47 | |
| *** ChanServ changes topic to "Gerrit has been successfully restarted" | 13:47 | |
| *** ChanServ changes topic to "Gerrit is going to be restarted due to performance problems" | 13:53 | |
| -openstackstatus- NOTICE: Gerrit has been successfully restarted | 13:53 | |
| *** IRCFrEAK has joined #openstack-shade | 14:12 | |
| *** IRCFrEAK has quit IRC | 14:14 | |
| *** yfried has joined #openstack-shade | 15:33 | |
| clarkb | yfried: possibly you are caught in the release of pbr 2.0 if shade has/had a pin for <2.0 | 15:38 |
| clarkb | yfried: are you using latest shade? | 15:38 |
| yfried | clarkb: "pip install shade" | 15:44 |
| yfried | clarkb: I saw a recent fix was merged. when will it be released? | 15:44 |
| clarkb | yfried: https://git.openstack.org/cgit/openstack-infra/shade/tree/requirements.txt?h=1.16.0&id=6ff03db0a8f71781645111f2de664d2bfaf25aca ya that has the pin and thats the last commit that was tagged | 15:45 |
| clarkb | mordred: ^ you'll want to make a shade release soon/now to fix the pbr issue | 15:45 |
| yfried | clarkb: mordred: yes please | 15:45 |
| yfried | we had a pretty bad week due to this bug | 15:46 |
| yfried | I've been getting angry demands to drop shade from our project | 15:48 |
| clarkb | yfried: a simple workaround is to install pip<2.0 first | 15:49 |
| clarkb | er | 15:49 |
| clarkb | pbr<2.0 | 15:49 |
| yfried | clarkb: not helping | 15:50 |
| yfried | clarkb: clients will later kill the command line due to conflicts | 15:50 |
| clarkb | yfried: the clients should still work with older pbr though | 15:51 |
| * clarkb is testing locally really quick | 15:51 | |
| clarkb | yfried: yes worked fine locally | 15:51 |
| clarkb | pip install 'pbr<2.0' && pip install shade | 15:51 |
| yfried | clarkb: we have a CLI tool and it's entrypoint is validating deps | 15:51 |
| clarkb | yfried: yes pbr 1.x is a valid dep | 15:52 |
| clarkb | nothing has a pbr >2.0 pin | 15:52 |
| clarkb | (or if it does thats a new bug | 15:52 |
| yfried | clarkb: https://github.com/openstack/python-cinderclient/blob/master/requirements.txt#L4 | 15:53 |
| clarkb | ugh | 15:53 |
| * clarkb cries | 15:53 | |
| yfried | clarkb: we've had several like this | 15:53 |
| clarkb | I guess thats how we "fixed" things | 15:54 |
| yfried | clarkb: basically, any new openstack pkg release is now killing us | 15:54 |
| clarkb | dhellmann: FYI ^ that was probably the wrong way to fix pbr | 15:54 |
| clarkb | dhellmann: since it means there is no range that straddles the old and new working sets with pbr | 15:54 |
| dhellmann | ? | 15:54 |
| * dhellmann reads scrollback | 15:54 | |
| clarkb | dhellmann: tldr is when we had pbr 2.0 release we should only have removed the upper pin on PBR not also moved the lower pin to be >= 2.0 | 15:54 |
| dhellmann | yeah, we were just discussing whether we needed to remove the cap on pbr in stable/ocata | 15:55 |
| clarkb | dhellmann: yes the cap should be removed everywhere imo | 15:55 |
| clarkb | but the lower bound should never have been pushed up to >= 2.0 | 15:55 |
| yfried | clarkb: https://storyboard.openstack.org/#!/story/2000915 | 15:55 |
| dhellmann | I raised it in #openstack-requirements but most of that team seems to be offline atm and I didn't want to take any action without consulting them | 15:55 |
| clarkb | dhellmann: because now you either work with 1.x on old code XOR work with 2.x on new code | 15:55 |
| dhellmann | yeah, I don't think there's really any reason we needed to update to >=2.0 | 15:56 |
| clarkb | ideally we would hvae let new code work with 1.x too to avoid the problem yfried is running into | 15:57 |
| dhellmann | oh, except some projects wanted to start using the sphinx warning setting, and I'm not sure if earlier versions of pbr passed that through properly | 15:57 |
| clarkb | dhellmann: it doesn't matter for testing though | 15:57 |
| clarkb | since testing will always install latest | 15:57 |
| clarkb | (if this was a functional issue for installations then ya I would worry about it) | 15:57 |
| dhellmann | we can try resetting that lower bound. | 15:57 |
| dhellmann | we don't really have a good way to signal the expected version of pbr other than the g-r list, though, do we? it's not constrained. | 15:58 |
| clarkb | correct, you can't really constrain it because setup.py pulls it in as an install requires | 15:58 |
| clarkb | yfried: fwiw this issue affected all of openstack so dropping shade wouldn't have completely avoided the problem for you | 15:59 |
| yfried | clarkb: shouldn't shade req be managed by global req? | 15:59 |
| clarkb | yfried: I'm not sure but thats orthogonal, pbr isn't managed by constraints and isn't really even managed by g-r | 16:00 |
| clarkb | since pbr is an install requires it happens outside of all that managment | 16:00 |
| yfried | clarkb: I don't want to drop shade, it's key for our ansible works, but I have some opposition... | 16:00 |
| clarkb | (which is why we had problems with the 2.0 release) | 16:00 |
| dhellmann | can we not just release a version of shade that doesn't have the cap? | 16:02 |
| clarkb | dhellmann: we can and thats likely the simple fix. I just wanted to make sure release team saw that bumping up lower bound was problematic | 16:03 |
| clarkb | we really need to have ranges that overlap when making these transitions | 16:03 |
| dhellmann | yeah | 16:03 |
| yfried | clarkb: so, what's the ETA on a working shade? | 16:13 |
| clarkb | yfried: I'd have to defer to mordred on that. I don't know what the current delta is between HEAD and previous release | 16:14 |
| yfried | mordred: ? | 16:15 |
| clarkb | you can propose the change to the requirements repo as a start | 16:16 |
| clarkb | er releases repo | 16:16 |
| yfried | clarkb: sorry? | 16:37 |
| yfried | clarkb: what is this? | 16:37 |
| clarkb | yfried: well I thought that shade was managed by the release team but I'm not seeing it in the releases repo | 16:38 |
| clarkb | yfried: much of openstack releases are managed by the openstack/releases repo. You push metadata there then automated jobs created tags and do release things | 16:38 |
| clarkb | but doesn't look like shade is there so mordred would ideally make the tag and push it (on whatever commit is most appropriate) | 16:39 |
| *** mgagne_ is now known as mgagne | 18:05 | |
| *** Matias_ has quit IRC | 20:32 | |
| Shrews | yup, mordred is the only one that has the shade release powerz | 20:50 |
| openstackgerrit | Morgan Fainberg proposed openstack-infra/shade master: change test_endpoints to use requests mock https://review.openstack.org/444541 | 21:21 |
| *** gouthamr has quit IRC | 21:25 | |
| notmorgan | mordred: sigh... | 21:55 |
| notmorgan | mordred: so yeah endpoints are about the sloppiest junk in keystone | 21:55 |
| notmorgan | mordred: looks like v2 is region, v3 is region *and* region_id and that looks like it's a typo and should have only been region_id | 21:56 |
| notmorgan | also the api-spec says region_id. | 21:56 |
| *** cdent has quit IRC | 21:57 | |
| openstackgerrit | Morgan Fainberg proposed openstack-infra/shade master: change test_endpoints to use requests mock https://review.openstack.org/444541 | 22:01 |
| *** gouthamr has joined #openstack-shade | 22:18 | |
| openstackgerrit | Morgan Fainberg proposed openstack-infra/shade master: Convert simple role crud to requests mock https://review.openstack.org/445673 | 22:32 |
| *** ChanServ changes topic to "#openstack-shade" | 22:42 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!