| jeblair | Shrews: wfm (another approach might be to built a set while iterating, then check that the sets match, but i don't think that's necessary). | 00:01 |
|---|---|---|
| Shrews | yeah, didn't think that was necessary | 00:02 |
| *** jamielennox_ has joined #zuul | 00:02 | |
| *** saneax-_-|AFK has joined #zuul | 00:04 | |
| *** jamielennox_ is now known as jamielennox | 00:06 | |
| SpamapS | oy | 01:43 |
| SpamapS | I've gotten nowhere on this | 01:43 |
| SpamapS | 2 hours in and I basically have spent 8 minutes actually debugging :-P | 01:43 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 02:11 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 02:15 |
| Shrews | If 649 passes, I'm spending tomorrow drinking | 02:15 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 02:18 |
| *** Cibo_ has joined #zuul | 02:32 | |
| *** Zara_ has joined #zuul | 03:06 | |
| *** adam_g_ has joined #zuul | 03:06 | |
| *** gundalow_ has joined #zuul | 03:06 | |
| *** jkt_ has joined #zuul | 03:06 | |
| *** nibz has joined #zuul | 03:06 | |
| *** cinerama` has joined #zuul | 03:07 | |
| *** Cibo_ has quit IRC | 03:07 | |
| *** Zara has quit IRC | 03:07 | |
| *** adam_g has quit IRC | 03:07 | |
| *** Cibo has quit IRC | 03:07 | |
| *** cinerama has quit IRC | 03:07 | |
| *** nibalizer has quit IRC | 03:07 | |
| *** gundalow has quit IRC | 03:07 | |
| *** jkt has quit IRC | 03:07 | |
| *** adam_g_ is now known as adam_g | 03:07 | |
| *** Cibo has joined #zuul | 03:08 | |
| *** saneax-_-|AFK is now known as saneax | 03:20 | |
| *** saneax is now known as saneax-_-|AFK | 03:22 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 03:40 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 03:45 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 03:46 |
| Shrews | jeblair: I think we need some more smarts built around the min-ready code path. If an image never becomes available for a provider, nodepoold will just keep generating requests for nodes of that type. We can see that here: http://logs.openstack.org/49/431649/8/check/gate-dsvm-nodepool/c4b26fa/logs/screen-nodepool.txt.gz | 04:01 |
| Shrews | jeblair: I'm too tired to come up with a solution for that, so hopefully you have some thoughts. | 04:02 |
| Shrews | jeblair: Also, I'm not clear on why those images aren't available in the gate-dsvm-nodepool job there. But, again... tired | 04:03 |
| * Shrews enters sleep mode | 04:03 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix missing mutex release when aborting builds https://review.openstack.org/432211 | 06:31 |
| *** saneax-_-|AFK is now known as saneax | 06:39 | |
| *** isaacb has joined #zuul | 07:25 | |
| *** isaacb has quit IRC | 07:51 | |
| *** saneax is now known as saneax-_-|AFK | 07:58 | |
| *** isaacb has joined #zuul | 08:00 | |
| *** Cibo_ has joined #zuul | 08:37 | |
| *** Cibo_ has quit IRC | 08:42 | |
| *** saneax-_-|AFK is now known as saneax | 08:43 | |
| *** Cibo_ has joined #zuul | 08:51 | |
| *** gundalow_ is now known as gundalow | 08:59 | |
| *** hashar has joined #zuul | 09:08 | |
| *** bhavik1 has joined #zuul | 09:16 | |
| *** Zara_ is now known as Zara | 09:51 | |
| *** rzetikx has joined #zuul | 09:56 | |
| *** rzetikx has quit IRC | 09:57 | |
| *** isaacb has quit IRC | 10:14 | |
| *** isaacb has joined #zuul | 10:15 | |
| *** isaacb has quit IRC | 10:25 | |
| *** bhavik1 has quit IRC | 10:33 | |
| *** isaacb has joined #zuul | 11:22 | |
| *** jkt_ is now known as jkt | 11:32 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul feature/zuulv3: Merge branch 'master' into feature/zuulv3 https://review.openstack.org/434232 | 11:35 |
| *** Cibo_ has quit IRC | 11:43 | |
| *** isaacb has quit IRC | 12:11 | |
| *** hashar has quit IRC | 12:25 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul feature/zuulv3: Merge branch 'master' into feature/zuulv3 https://review.openstack.org/434232 | 12:34 |
| *** hashar has joined #zuul | 12:58 | |
| *** isaacb has joined #zuul | 13:02 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Remove TODO comment that seems to be done https://review.openstack.org/433961 | 13:03 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Re-enable test_footer_message https://review.openstack.org/430486 | 13:04 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Drop test_node_label https://review.openstack.org/430473 | 13:04 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Re-enable test_client_enqueue_ref test https://review.openstack.org/393887 | 13:05 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 13:10 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 13:10 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 13:13 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 13:13 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 13:32 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 13:39 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 13:56 |
| Shrews | sorry for the noise. trying to track down a lock issue | 13:57 |
| *** isaacb has quit IRC | 14:05 | |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Set Node image_id and launcher attributes https://review.openstack.org/433242 | 14:05 |
| *** isaacb has joined #zuul | 14:06 | |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Add generator API method for node iteration https://review.openstack.org/433252 | 14:08 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Disconnect from ZooKeeper at shutdown https://review.openstack.org/433919 | 14:08 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix missing mutex release when aborting builds https://review.openstack.org/432211 | 14:13 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Remove subnodes from nodepool https://review.openstack.org/432403 | 14:13 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 14:24 |
| *** jianghuaw has joined #zuul | 14:33 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul feature/zuulv3: Add reporter for Federated Message Bus (fedmsg) https://review.openstack.org/426861 | 14:43 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 14:52 |
| *** isaacb has quit IRC | 14:59 | |
| *** isaacb has joined #zuul | 15:10 | |
| jeblair | Shrews: how about we have the launcher peek at zk and only submit a min-ready request if there's a ready image for any provider? that wouldn't catch all the cases, but it would catch the most common one. | 15:23 |
| jeblair | Shrews, pabelanger: and i think those imagase which aren't available in the devstack job are the ones that are paused, so we aren't building them. | 15:24 |
| Shrews | jeblair: that was the first thought i had. but yeah, doesn't catch the other cases (quota exceeded, provider problems) | 15:25 |
| Shrews | jeblair: another thought i had was add a failure code/reason to the NodeRequest and add some sort of backoff algorithm based on the fail code | 15:26 |
| Shrews | but trying to track down this node lock issue so i haven't been able to think too much about either of those solutions | 15:26 |
| jeblair | Shrews: yeah, or we could consider moving the min-ready supply into the providerworker itself (if i have quota, and i have an image ready, and there is a label that's under min-ready, build one) | 15:27 |
| *** saneax is now known as saneax-_-|AFK | 15:28 | |
| jeblair | Shrews: (iow, bypass the request process) | 15:28 |
| *** yolanda has quit IRC | 15:34 | |
| *** yolanda has joined #zuul | 15:35 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: WIP: Re-enable devstack test job https://review.openstack.org/431649 | 15:44 |
| *** abregman has joined #zuul | 15:56 | |
| *** abregman has quit IRC | 15:56 | |
| *** abregman has joined #zuul | 15:58 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Implement node cleanup https://review.openstack.org/433736 | 16:00 |
| *** isaacb has quit IRC | 16:09 | |
| *** morgan_ is now known as morgan | 16:10 | |
| Shrews | jeblair: ok, think i found it. It's a timing issue in check_devstack_plugin.sh. The node is marked as READY, but isn't unlocked until shortly after when the request is marked FULFILLED. The 'delete' command is trying to lock the node and set the DELETE state during the period it is still locked (which is very short). I'm adding a 5 second wait time to the lock attempt in the 'delete' command. | 16:14 |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Re-enable devstack test job https://review.openstack.org/431649 | 16:15 |
| Shrews | s/wait time/lock timeout/ | 16:16 |
| jeblair | Shrews: i wonder if we should add a 'locked' field to the node list command, and have 'waitfornode' also wait for the node to be unlocked? | 16:18 |
| Shrews | jeblair: hah. i did that to debug this. but i didn't keep it because it would be slightly racey. we'd have to store the field as None just before we actually unlock it. | 16:21 |
| Shrews | jeblair: oh, my method added a field to Node. maybe you're just suggesting trying the lock in the list command itself | 16:22 |
| jeblair | Shrews: oh, i wasn't necessarily suggesting a new field, but rather a check to see if a lock node exists. that's almost equally racy of course. | 16:22 |
| jeblair | ya | 16:22 |
| Shrews | yes | 16:22 |
| jeblair | but you know, it's as accurate as we can get, and it's probably useful info for an operator anyway | 16:22 |
| Shrews | i'll work something up | 16:23 |
| Shrews | but i think the delete lock timeout is still worthy. it should mostly be instantaneous for normal use | 16:23 |
| jeblair | yeah | 16:23 |
| *** cinerama` is now known as cinerama | 16:23 | |
| pabelanger | jeblair: re: 433235 where should I be looking to query all nodes? | 16:29 |
| jeblair | pabelanger: there's a getNodes method on the zk object, and Shrews recently added an iterator (i'd recommend using that) | 16:30 |
| jeblair | Shrews: and maybe you want to take a look at my comment on 433235 before pabelanger proceeds? | 16:31 |
| pabelanger | jeblair: thanks for the pointer | 16:31 |
| pabelanger | sure, I can wait | 16:32 |
| Shrews | jeblair: pabelanger: yeah, that sounds like a good idea there | 16:48 |
| Shrews | pabelanger: zk.nodeIterator() is what you want | 16:48 |
| *** abregman has quit IRC | 16:49 | |
| *** Cibo_ has joined #zuul | 16:59 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/nodepool feature/zuulv3: Add lock state to node listing https://review.openstack.org/434397 | 17:05 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool feature/zuulv3: Re-enable test_diskimage_build_only test https://review.openstack.org/433265 | 17:14 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool feature/zuulv3: Re-enable test_dib_upload_fail test https://review.openstack.org/433235 | 17:14 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool feature/zuulv3: Re-enable test_image_upload_fail test https://review.openstack.org/433270 | 17:14 |
| openstackgerrit | Paul Belanger proposed openstack-infra/nodepool feature/zuulv3: Re-enable working test_builder.py tests https://review.openstack.org/433262 | 17:14 |
| pabelanger | jeblair: Shrews: okay, I've updated 433235. | 17:14 |
| pabelanger | for loop based on http://stackoverflow.com/questions/3345785/getting-number-of-elements-in-an-iterator-in-python | 17:14 |
| mordred | jeblair: fwiw, I'm tracking down shade's gate being broken by what I think is a new nova microversion. I do not believe this will impact my ability to get logging done - but just FYI as to why there are so many shade patches flying atm | 17:19 |
| Shrews | mordred: oh joy | 17:23 |
| mordred | yah. tell me about it | 17:23 |
| mordred | such things will be made better once we're microversion aware - but that's not going to in the next couple of weeks | 17:23 |
| jeblair | mordred: ack. also: *ackthpth*. | 17:23 |
| mordred | jeblair: I wholeheartedly agree | 17:24 |
| clarkb | mordred: if you aren't microversion aware wouldn't you just use the base 2.0 api and not worry about changes? | 17:25 |
| mordred | clarkb: nope. not being microversion aware and using novaclient means we get latest | 17:26 |
| clarkb | mordred: couldnt you just hardcode the 2.0 version with novaclient though | 17:26 |
| clarkb | I guess thats what I am suggesting, if not microversion aware then safeest option is base option | 17:26 |
| mordred | possibly - but since we've been using 'latest' so far, I'd need to verify that we didn't break things | 17:26 |
| mordred | by 'downgrading' in some places - it's probably fine - and probably a good idea | 17:27 |
| *** Cibo_ has quit IRC | 17:47 | |
| *** hashar has quit IRC | 17:52 | |
| *** saneax-_-|AFK is now known as saneax | 17:57 | |
| *** jamielennox is now known as jamielennox|away | 18:13 | |
| *** saneax is now known as saneax-_-|AFK | 18:48 | |
| SpamapS | jeblair: FYI, had a great discussion with some networking folks about a feature they'd like to see in Zuul: https://storyboard.openstack.org/#!/story/2000869 | 18:54 |
| mordred | SpamapS: neat. is that mostly so that they can start responding to issues as they roll in? | 18:58 |
| SpamapS | mordred: yes, some of the jobs take 5 hours, some 9 hours. | 18:58 |
| mordred | nod | 18:59 |
| SpamapS | I did wonder if they could have post-1, post-2, post-3, and just let the comments from each pipeline feed into the others. | 18:59 |
| clarkb | that info would be good to add to the story (use cases are often really handy when designing for features) | 18:59 |
| mordred | SpamapS: it's _possible_ (although I'll let jeblair tell me I'm wrong) that that could be considered a case of fail-fast - we've talked before about being able to notice and do something more quickly if some of the jobs associated with somethign fail | 18:59 |
| clarkb | like maybe you only want to know if something fails but otherwise aggregate successes | 19:01 |
| mordred | where $do_somthing might be vague/hand-wavey/configurable ... but the "let the dev know that the job failed pep8 before the entire devstack job is done running" - or "reparent to the nearest non-failing change if we notice that" or whatnot | 19:01 |
| mordred | clarkb: yah | 19:01 |
| clarkb | (which is personally what I'd want in that use case) | 19:01 |
| SpamapS | clarkb: good point, added | 19:01 |
| mordred | but there are a few different potential aspects to "set of jobs take a while, action/reporting before the total set is done is desirable" for many values of each of those things | 19:01 |
| SpamapS | Seems like the simplest thing to do is just allow reporters for jobs | 19:02 |
| mordred | SpamapS: possibly so. I bring up the other thing to raise the possibility that this is actually related to a few other features that have been desired for a while, so something that takes them all into account might be warranted | 19:10 |
| mordred | or they may be completely unrelated | 19:10 |
| jeblair | mordred, SpamapS: yeah, i think we can make pipeline reporting a bit more granular -- note that the current design is mostly based on traditional gerrit because a partial report is expensive for a human to parse in gerrit (each partial report is another message). we mask some of that in openstack, but it doesn't necessarily make it less so. however, there's now a test results plugin for gerrit which we hope to start using once we upgrade. ... | 19:29 |
| jeblair | ... that greatly reduces the (mostly human) costs associated with partial reports. | 19:29 |
| jeblair | i'm not sure where on the spectrum github falls, but that's something to keep in mind too | 19:29 |
| SpamapS | jeblair: github has test statuses... and those can be [1runner]:[Nstatuses] | 19:30 |
| SpamapS | jeblair: in this case, _they want the partial result to bother a human_ | 19:31 |
| SpamapS | because it's the longer part of their pipeline and thus they need to fail fast and get fixes into the pipeline as fast as possible, so it's worth the noise. | 19:32 |
| jeblair | SpamapS: yep | 19:32 |
| SpamapS | They may also want successes to trigger humans | 19:32 |
| jeblair | i think that will not be too hard to accomplish without major changes | 19:33 |
| jeblair | i'd describe this as 'partial reports' (jobs keep running and issue a final report) vs 'fail fast' (all jobs aborted when first fails and there's a single report) | 19:36 |
| SpamapS | yeah, two different things | 19:40 |
| *** hashar has joined #zuul | 19:44 | |
| jeblair | SpamapS: cool, i left notes on potential implementation in the story | 19:48 |
| SpamapS | sweet | 19:48 |
| SpamapS | Sitting here with some people who bounced off Zuul a year ago (From OpenNFV and OpenDaylight), so if you have questions for them, let me know | 19:49 |
| mordred | SpamapS: I think we mostly want to let them know we value them as humans and that their concerns are all things that are front and center in the v3 work - so we hope they don't stay bounced off | 19:52 |
| SpamapS | That's basically what I said: We weren't really a community for not-OpenStack people yet, but we are now, and we want to help. | 19:53 |
| mordred | ++ | 20:03 |
| mordred | SpamapS: also tell them I'd love to buy them all drinks next time we're colocated :) | 20:03 |
| *** jamielennox|away is now known as jamielennox | 20:36 | |
| openstackgerrit | Antoine Musso proposed openstack-infra/zuul master: Expose commitMessage as a Change attribute https://review.openstack.org/222791 | 21:23 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Run zuul v3 launcher https://review.openstack.org/433347 | 21:44 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Inherit playbooks and modify job variance https://review.openstack.org/430483 | 21:47 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Split merger and launcher git roots https://review.openstack.org/430456 | 21:47 |
| jhesketh | Morning | 21:55 |
| jeblair | jhesketh: good morning! can you take a look at my response on 433964? | 22:02 |
| jhesketh | sure | 22:02 |
| jhesketh | jeblair: yep, makes sense | 22:03 |
| jeblair | cool | 22:03 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Re-enable test_queue_names https://review.openstack.org/429883 | 22:03 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove misleading log statements https://review.openstack.org/433970 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Squelch ref-replication gerrit warnings https://review.openstack.org/433959 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add Git driver https://review.openstack.org/433942 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove gearman settled check https://review.openstack.org/433943 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add (minimal) support for topic-changed event https://review.openstack.org/433966 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Rename zuul-server zuul-scheduler https://review.openstack.org/433980 | 22:05 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Correct getGitwebUrl https://review.openstack.org/433964 | 22:05 |
| jeblair | the git driver change collided with the split repo paths change, so that's a fix and rebase | 22:05 |
| jeblair | i'm going to self re-approve that since it's not a substantial change | 22:06 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add Git driver https://review.openstack.org/433942 | 22:30 |
| *** hashar has quit IRC | 22:32 | |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Remove gearman settled check https://review.openstack.org/433943 | 22:32 |
| jeblair | Shrews: i left two comments on 433736 | 22:43 |
| pabelanger | 64 bytes from nl01.openstack.org (23.253.92.28): icmp_seq=1 ttl=61 time=0.721 ms | 22:56 |
| pabelanger | waiting for puppet to run now | 22:57 |
| jeblair | pabelanger, mordred, jhesketh: i kind of want to disable the behind_dequeue test until we have a chance to dig into it | 23:09 |
| jeblair | i think it's going to burn too much of our time this week and next | 23:09 |
| pabelanger | no objections here | 23:10 |
| jhesketh | that sounds sensible to me | 23:10 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Remove misleading log statements https://review.openstack.org/433970 | 23:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Squelch ref-replication gerrit warnings https://review.openstack.org/433959 | 23:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Add (minimal) support for topic-changed event https://review.openstack.org/433966 | 23:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Rename zuul-server zuul-scheduler https://review.openstack.org/433980 | 23:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Correct getGitwebUrl https://review.openstack.org/433964 | 23:11 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Temporarily disable test_dependent_behind_dequeue https://review.openstack.org/434553 | 23:11 |
| jeblair | pabelanger, jhesketh: okay, that's the stack rebased on 434553; if you could +3 that one i'd appreciate it :) | 23:11 |
| pabelanger | Feb 15 23:11:51 nl01 puppet-user[20255]: Finished catalog run in 5.62 seconds | 23:12 |
| pabelanger | cool, puppet running on nl01.o.o now | 23:12 |
| mordred | jeblair: ++ | 23:13 |
| jhesketh | done | 23:13 |
| pabelanger | done also | 23:14 |
| Shrews | jeblair: ack on ProviderManager comment, though the main NodePool thread does not have managers. They're down in the ProviderWorker threads. I'm going to have to rework moving them to the main thread and providing access for the child threads. | 23:20 |
| pabelanger | jhesketh: Shrews: mind a review on https://review.openstack.org/#/c/433235/ | 23:21 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Temporarily disable test_dependent_behind_dequeue https://review.openstack.org/434553 | 23:23 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Squelch ref-replication gerrit warnings https://review.openstack.org/433959 | 23:24 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Correct getGitwebUrl https://review.openstack.org/433964 | 23:25 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add (minimal) support for topic-changed event https://review.openstack.org/433966 | 23:25 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Re-enable test_dib_upload_fail test https://review.openstack.org/433235 | 23:26 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Remove misleading log statements https://review.openstack.org/433970 | 23:26 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Rename zuul-server zuul-scheduler https://review.openstack.org/433980 | 23:26 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Re-enable working test_builder.py tests https://review.openstack.org/433262 | 23:27 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Re-enable test_diskimage_build_only test https://review.openstack.org/433265 | 23:28 |
| openstackgerrit | Merged openstack-infra/nodepool feature/zuulv3: Re-enable test_image_upload_fail test https://review.openstack.org/433270 | 23:28 |
| pabelanger | merges! | 23:31 |
| jeblair | Shrews: ah, i may have crossed wires with the old code which used Nodepool.getProviderManager | 23:42 |
| jeblair | Shrews: so if you decide you want them on the main nodepool object, you can probably re-use much of that | 23:43 |
| jeblair | Shrews: alternatively, we could move deleting to be something that the providerworker kicks off | 23:43 |
| jeblair | Shrews: (in its main run loop, it could assign handlers, complete handlers, start deleters) | 23:44 |
| jeblair | jhesketh: 432211 is a forward-port of a patch you have +2d on master if you want to +3 it on v3 | 23:46 |
| Shrews | jeblair: no, i don't want each providerworker iterating through all nodes continuously. i'll move the managers up a level | 23:46 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix missing mutex release when aborting builds https://review.openstack.org/384980 | 23:49 |
| jeblair | Shrews: makes sense | 23:49 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!