| SpamapS | jeblair: sorry to miss zuul meeting today. I had several stacked things that kept me from even being able to be on my phone :P | 01:06 |
|---|---|---|
| *** saneax is now known as saneax-_-|AFK | 02:52 | |
| *** jamielennox is now known as jamielennox|away | 03:08 | |
| *** jamielennox|away is now known as jamielennox | 03:22 | |
| *** rmoe has quit IRC | 03:54 | |
| *** rmoe_ has joined #zuul | 03:54 | |
| *** timrchavez has joined #zuul | 04:53 | |
| *** timrchavez has quit IRC | 04:54 | |
| *** timrchavez_ has joined #zuul | 04:56 | |
| *** timrchavez_ is now known as timrc_ | 05:01 | |
| timrc_ | /trigger set setdict conditions "${nick} == ${info:irc_nick,${server}} && ${channel} =~ [.-](fr|it|es)$ && ${aspell.dict.irc.${server}.${channel}} ==" | 05:04 |
| timrc_ | ooops | 05:05 |
| jlk | timrc_: what are you even trying to do there? | 05:21 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Don't getChange on source not triggering a change https://review.openstack.org/406129 | 05:22 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Do not process unknown project https://review.openstack.org/412751 | 05:22 |
| timrc_ | jlk: Well it was a copy-pasta error. Its part of a block of weechat instructions that set the dictionary based on channel name. | 05:23 |
| timrc_ | But that is not um important to me :) | 05:23 |
| *** saneax-_-|AFK is now known as saneax | 05:31 | |
| tristanC | jhesketh: Greeting sir, would you happen to remember the reasoning behind http://git.openstack.org/cgit/openstack-infra/zuul/tree/zuul/scheduler.py#n1022 ? For a third-party CI connected to review.openstack.org, it makes zuul fetch *all* the changes, even the one for (unknown) projects where no jobs are attached to. | 05:31 |
| tristanC | jhesketh: well I've proposed to revert to the old behavior here https://review.openstack.org/412751, and I confirm this works even when a change depends-on an unknown project | 05:37 |
| jhesketh | tristanC: I can't entirely remember... but from the comment I think it is if you have a change from a non-foreign project later on reference (ie depend on) this foreign change | 05:54 |
| *** willthames has quit IRC | 06:49 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Do not process unknown project https://review.openstack.org/412751 | 06:49 |
| *** jamielennox is now known as jamielennox|away | 07:28 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Do not process unknown project https://review.openstack.org/412751 | 07:29 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul: Don't getChange on source not triggering a change https://review.openstack.org/406129 | 07:29 |
| *** abregman has joined #zuul | 07:41 | |
| *** abregman has quit IRC | 08:13 | |
| *** hashar has joined #zuul | 09:19 | |
| *** bhavik1 has joined #zuul | 09:25 | |
| *** abregman has joined #zuul | 09:31 | |
| *** bhavik1 has quit IRC | 10:56 | |
| *** saneax is now known as saneax-_-|AFK | 12:24 | |
| *** saneax-_-|AFK is now known as saneax | 13:22 | |
| *** abregman has quit IRC | 13:26 | |
| *** abregman has joined #zuul | 13:26 | |
| *** saneax is now known as saneax-_-|AFK | 13:34 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool: Mitigate config disagreement issues https://review.openstack.org/413125 | 14:49 |
| *** harlowja has joined #zuul | 15:49 | |
| *** abregman has quit IRC | 16:24 | |
| *** harlowja has quit IRC | 16:42 | |
| jeblair | Shrews: did you need to remove the run method and split out reloadconfig? or is that just a code organization/style change? | 16:59 |
| Shrews | jeblair: i *thought* i was going to reuse it, but turns out i didn't, and i liked how i could use it to remove the _run() method | 17:00 |
| Shrews | so i left it | 17:00 |
| jeblair | ok | 17:02 |
| Shrews | jeblair: but it's also there if we find other places where we should do a reload | 17:02 |
| *** harlowja has joined #zuul | 17:53 | |
| *** hashar has quit IRC | 18:14 | |
| *** bhavik1 has joined #zuul | 18:37 | |
| jhesketh | Morning | 18:58 |
| mordred | morning jhesketh ! | 18:59 |
| *** harlowja has quit IRC | 19:35 | |
| rcarrillocruz | clarkb: https://review.openstack.org/#/c/403732/11/playbooks/roles/copy_mirror_config/tasks/main.yaml , i wonder if we should add constants in devstack_gate_vars for 'jenkins' and 'tempest' users, the same way we have for BASE. i.e. something like: "CI_USER: jenkins" | 19:56 |
| clarkb | well in that case its 3 users that are not jenkins | 19:57 |
| clarkb | rcarrillocruz: do you mean for the source? | 19:58 |
| rcarrillocruz | we expect to find in this particular case /home/jenkins/.pydistutils. I wonder if we should have a general constant for this, that is CI_USER: jenkins, and any mention (in this playbook or wherever) we just reference {{ CI_USER }} . In case we move over from this legacy name to 'zuul' for example | 20:01 |
| rcarrillocruz | i.e. unhardcode that and just use a variable, like BASE: /opt/stack | 20:02 |
| clarkb | ya it can't hurt | 20:02 |
| *** bhavik1 has quit IRC | 20:04 | |
| pabelanger | rcarrillocruz: why don't we just run this code as the jenkins user or use becomes: jenkins | 20:08 |
| pabelanger | over hardcoding a var | 20:08 |
| clarkb | pabelanger: because you ahve to be root to do the other things | 20:09 |
| clarkb | (like write into root's homedir) | 20:09 |
| rcarrillocruz | yeah, and we have other users like tempest or stack | 20:10 |
| pabelanger | you can use becomes per task | 20:10 |
| pabelanger | oh | 20:10 |
| pabelanger | I see | 20:10 |
| SpamapS | so.. I'm trying to re-enable a test in webapp.. and I see something weird in the fixtures for adding events | 20:10 |
| SpamapS | when I add one change (1,1) it has Code-Review set to 'OK' | 20:11 |
| SpamapS | bu the next change, on a different repo, has it set to 'NEED' | 20:11 |
| clarkb | SpamapS: are they two different projects? (possibly the fixture has something different for vote ranges on different projects) | 20:11 |
| SpamapS | yes | 20:13 |
| SpamapS | There's some very hard to understand logic which seems to govern it in the fixture's 'getSubmitRecords' method. | 20:14 |
| clarkb | basically gerrit lets you configure the ranges and the default "function" on the range is at least one max vote in every column and no min votes | 20:14 |
| clarkb | I wouldn't be surprised if zuul's testing was faking that and usingdifferent ranges for completeness | 20:14 |
| clarkb | yup cat_min and cat_max are checked against | 20:15 |
| clarkb | but they don't appear to be variable across projects https://git.openstack.org/cgit/openstack-infra/zuul/tree/tests/base.py?h=feature/zuulv3#n106 | 20:16 |
| SpamapS | Yeah still not finding where it's choosing NEED vs. OK | 20:16 |
| clarkb | https://git.openstack.org/cgit/openstack-infra/zuul/tree/tests/base.py?h=feature/zuulv3#n332 ? | 20:17 |
| clarkb | basically if you have a cat_min vote then its reject (this wins same as gerrit), if you don't have that and do have cat max then you are OK and ready to merge, else you NEED | 20:17 |
| SpamapS | but the votes are identical, I'm not even setting a code review vote | 20:18 |
| clarkb | there are no addApproval() calls? | 20:18 |
| SpamapS | both have an addApproval | 20:19 |
| SpamapS | the only difference is project name | 20:19 |
| SpamapS | it's tests/test_webapp.py, test_webapp_find_change | 20:19 |
| clarkb | maybe something is overriding self.categories on one of them? | 20:19 |
| clarkb | SpamapS: and which category is set to NEED? | 20:21 |
| clarkb | should be something like label: approved, status: NEED | 20:22 |
| SpamapS | categories is the same each time | 20:23 |
| SpamapS | clarkb: code-review | 20:23 |
| clarkb | SpamapS: look at the addAprovals for the two changes | 20:24 |
| clarkb | (its actually only runnign addApproval against one change) | 20:24 |
| clarkb | for code-review | 20:24 |
| SpamapS | herpaderp | 20:25 |
| SpamapS | ty for 2nd pair of eyes | 20:25 |
| SpamapS | that was it | 20:25 |
| SpamapS | clarkb: you get a cookie | 20:25 |
| clarkb | omnomnomnom | 20:26 |
| SpamapS | 20:29 | |
| openstackgerrit | Clint 'SpamapS' Byrum proposed openstack-infra/zuul: Re-enable test_webapp_find_change https://review.openstack.org/413226 | 20:29 |
| SpamapS | clarkb: ^ the result | 20:29 |
| clarkb | SpamapS: is that test actually comparing the label data too? or was it just something you noticed? | 20:31 |
| clarkb | (because I don't see where its asserting the label data for the changes) | 20:31 |
| SpamapS | clarkb: because the change wasn't approved, it was not enqueued | 20:32 |
| SpamapS | and thus not in the status | 20:32 |
| clarkb | aha | 20:32 |
| SpamapS | 39 skips left | 20:33 |
| SpamapS | we're enabling tests at a rate of about 2 per day | 20:33 |
| SpamapS | (enabling or deleting ;) | 20:33 |
| SpamapS | oh and now with that change I just did, only test_scheduler has skips | 20:34 |
| mordred | SpamapS: woot! | 20:43 |
| *** jamielennox|away is now known as jamielennox | 20:44 | |
| *** hashar has joined #zuul | 21:13 | |
| jeblair | yay! | 21:33 |
| jeblair | mordred, Shrews: i updated https://storyboard.openstack.org/#!/story/2000768 and assigned things to the 3 of us | 21:34 |
| jeblair | jamielennox: did you want to look at https://review.openstack.org/408849 ? | 21:35 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Add Zookeeper to tests https://review.openstack.org/413241 | 21:52 |
| jeblair | pabelanger: fyi https://review.openstack.org/413226 continues earlier work you did | 22:00 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Add Zookeeper to tests https://review.openstack.org/413241 | 22:02 |
| openstackgerrit | Merged openstack-infra/zuul: Re-enable test_webapp_find_change https://review.openstack.org/413226 | 22:02 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: Add Zookeeper to tests https://review.openstack.org/413241 | 22:14 |
| jamielennox | jeblair: i do want to go through it, but mostly to understand what's going on. if you are wanting to push it through you certainly don't need to wait for me | 22:31 |
| pabelanger | jeblair: great! | 22:32 |
| pabelanger | jeblair: Ah, and my bug too | 22:33 |
| jeblair | jamielennox: thanks; not super-urgent but i won't block on you then | 22:33 |
| *** willthames has joined #zuul | 23:01 | |
| *** hashar has quit IRC | 23:29 | |
| jamielennox | jeblair: i posted a comment on the review and it's mostly me trying to wrap my head around the way things are related now | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!