| corvus | mordred, clarkb, fungi: https://review.openstack.org/641508 is green again | 00:02 |
|---|---|---|
| mordred | corvus: you're green again | 00:03 |
| corvus | it's not easy | 00:03 |
| tristanC | SpamapS: we are using openshift labels with https://review.openstack.org/#/c/570669/ | 00:04 |
| tristanC | mugsie: i don't mind you taking over the azure driver, here is a useful piece of documentation: https://review.openstack.org/639595 and a test environment: https://review.openstack.org/639632 | 00:05 |
| tristanC | corvus: speaking of bugs, would be nice to have a new zuul release with the semaphore fixes and the github protected branch fix | 00:07 |
| corvus | tristanC: yeah, i'd like to land 641508 and make sure it works first, so we don't release a known broken new feature | 00:08 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Allow soft job dependencies https://review.openstack.org/641439 | 00:12 |
| corvus | tristanC: you may find that interesting ^ | 00:12 |
| tristanC | that sounds useful indeed | 00:14 |
| *** pabelanger has joined #zuul | 00:30 | |
| pabelanger | tristanC: yah, https://review.openstack.org/590697/ is our most recent issue. thankfully fixed by tobiash already | 00:39 |
| tristanC | pabelanger: yes, i'm tempted to backport all those patches on 3.5.0 as our users keep on reporting issues that are already fixed in master | 00:42 |
| pabelanger | I'm hoping for a release this week | 00:43 |
| pabelanger | but yah, been burned a few times on a few issues already fixed in master | 00:44 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix duplicate and reversed artifacts https://review.openstack.org/641508 | 00:49 |
| *** daniel2 has quit IRC | 00:49 | |
| *** daniel2 has joined #zuul | 00:50 | |
| *** daniel2 has quit IRC | 00:50 | |
| *** daniel2 has joined #zuul | 00:51 | |
| SpamapS | clarkb: you know why you can't just use the k8s module? | 01:10 |
| SpamapS | k8s – Manage Kubernetes (K8s) objects | 01:10 |
| SpamapS | New in version 2.6. | 01:10 |
| clarkb | SpamapS: oh huh | 01:11 |
| SpamapS | Could vendor it. | 01:12 |
| SpamapS | but I'd rather not :-P | 01:12 |
| mordred | or just land tobiash's multi-ansible work | 01:12 |
| clarkb | tobiash's work would make it possible | 01:12 |
| clarkb | ya | 01:12 |
| SpamapS | Light speed to Endor! | 01:12 |
| SpamapS | Pretty sure we need to decide if those constitute a 4.0 or not. :) | 01:13 |
| mordred | SpamapS: you're a 4.0 | 01:13 |
| * SpamapS proudly bears the 4.0 label | 01:13 | |
| clarkb | depends on whether or not the implementation requires the operator to do anything to upgrade imo | 01:14 |
| mordred | this is an excellent point | 01:14 |
| clarkb | I haven't followed closely enough to know if we are requiring a post pip install type task to preseed ansibles | 01:14 |
| SpamapS | IIRC we can make the default install Ansible at executor startup. | 01:14 |
| SpamapS | However, that places a new requirement on executors: that they be able to install stuff. | 01:15 |
| clarkb | ya if we can make it transparent then I think I'd still with 3x. If it isn't transparent in the default case then 4.0 | 01:15 |
| mordred | yeah | 01:15 |
| mordred | I feel like there was another thing we were talking about that was a potential 4.0 too | 01:15 |
| mordred | maybe I'm thinking of require-rdbms | 01:16 |
| SpamapS | Also we should think about what it means for 3.x users, because with a 4.0, we're saying "slow down, this is going to get bumpy", and thus we should expect some folks to stay on 3.x a while. | 01:16 |
| mordred | but I think there was at least one more thing that was slightly more obscure but was also important | 01:16 |
| mordred | SpamapS: well - I don't think we're intending to say slow down this is going to get bumpy with the 4.x | 01:16 |
| clarkb | mordred: state in zk was the traditional 4.0 | 01:16 |
| SpamapS | mordred: thats what I take from any major version bump. | 01:17 |
| mordred | clarkb: yah - but I thnk there were a couple of minor-ish things like this that were still upgrade actions for the operator that brought up a discussion of tagging a 4.0 | 01:17 |
| SpamapS | Could be I'm misguided. | 01:17 |
| mordred | SpamapS: well - it usually is true - but in this case I think we're simply trying to signal "action is needed" rather than "major changes" | 01:17 |
| clarkb | which is the actual semver meaning | 01:18 |
| clarkb | "you have to do something to cross this boundary" it doesn't say anything about quality or bumpyness | 01:18 |
| mordred | yeah. but I agree, peopel frequently use it to mean the more aggressive thing | 01:18 |
| mordred | we need an additional number ... one to indicate the "it's getting bumpy up in here" - one to indicate "action needed but otherwise the system is largely the same" - and one to indicate "changes happened, but no action needed" | 01:19 |
| SpamapS | We're saying the same thing. :) | 01:20 |
| SpamapS | I consider "you have to do something" bumpy. | 01:20 |
| SpamapS | "This may be broken" is more like "We lost an engine and we're heading to the nearest airport." ;-) | 01:21 |
| clarkb | if you are linus you bump versions arbitrarily to troll spamaps | 01:21 |
| SpamapS | mordred: I don't think things that are broken should be turned on. I like the practice of marking new features experimental. Kubernetes and Linux have both done this well. | 01:21 |
| SpamapS | clarkb: yeah, Linux has *not* done semver. :) | 01:21 |
| SpamapS | But anyway, point being, multi-ansible deserves careful consideragion. | 01:22 |
| clarkb | I dont think anyone is suggesting we release broken code | 01:22 |
| SpamapS | consideragion: n. - A thing that makes you think, and be angry, at the same time. | 01:22 |
| mordred | SpamapS: ++ | 01:22 |
| SpamapS | Now, for the time being... even if k8s did exist in 2.5... wouldn't the untrusted protections disallow it? | 01:23 |
| clarkb | openstack runs from head and does major sanoty checking | 01:23 |
| mordred | SpamapS: both to consideragion, and that other bits | 01:23 |
| SpamapS | For some reason I thought untrusted mode only allowed known good things with known good paths. | 01:23 |
| clarkb | SpamapS: no I dont think so | 01:23 |
| mordred | nope. it's a blacklist | 01:23 |
| SpamapS | ah ok | 01:23 |
| clarkb | since that module is just http requests | 01:23 |
| mordred | which is one of the reasons we were talking about dropping it | 01:23 |
| clarkb | there is no reason to prevent it | 01:23 |
| mordred | yeah | 01:23 |
| SpamapS | So I *could* vendor 2.6's k8s module into the 2.5 tree and it might work. | 01:24 |
| mordred | yeah | 01:24 |
| mordred | you could try just putting it in to a role | 01:24 |
| SpamapS | Harumph. | 01:24 |
| SpamapS | I thought untrusted mode also prevented you from writing your own modules. | 01:24 |
| mordred | oh gosh no | 01:25 |
| SpamapS | Since then you could just vendor a blacklisted one. | 01:25 |
| mordred | you can completely do your own modules | 01:25 |
| mordred | modules run remotely | 01:25 |
| mordred | you can't write your own action plugins | 01:25 |
| SpamapS | But I am running this on the executor. | 01:25 |
| SpamapS | That's the whole point. :) | 01:25 |
| mordred | hrm | 01:25 |
| * SpamapS has forgotten how the protections work | 01:26 | |
| SpamapS | I've already, honestly, oriented my mind toward "those are going away anyway..." :-P | 01:26 |
| * SpamapS beats back his biases | 01:26 | |
| SpamapS | Nope, can't just slurp k8s module.. it's go a ton of tie-in to ansible.module_utils | 01:29 |
| tristanC | SpamapS: would you mind giving your opinion on https://review.openstack.org/590092 . This let you spawn pod in a trusted parent pre-run, and let untrusted job use it directly from the run inventory | 01:30 |
| tristanC | example usage L277 of https://review.openstack.org/#/c/570669/8/playbooks/openshift/pre.yaml | 01:31 |
| pabelanger | mordred: I also kinda remember db required was 4.0 being a thing | 01:49 |
| fungi | it was. we accidentally made the rdbms required with one of the dashboard patches we had to promptly revert | 01:53 |
| fungi | and agreed (i think?) that reintroducing it would require a major semver bump | 01:55 |
| *** jamesmcarthur has joined #zuul | 02:01 | |
| *** jamesmcarthur has quit IRC | 02:08 | |
| *** jamesmcarthur has joined #zuul | 02:08 | |
| *** jamesmcarthur has quit IRC | 02:17 | |
| *** jamesmcarthur has joined #zuul | 02:24 | |
| SpamapS | tristanC: I don't need pods, I need namespaces. | 02:36 |
| SpamapS | The entire point of the job I'm trying to write is "can this be applied correctly?" | 02:36 |
| tristanC | SpamapS: ok, then what about a trusted role that would kubectl apply any file present in {{ project.src_dir }}/recipe? | 02:38 |
| *** jamesmcarthur has quit IRC | 02:38 | |
| tristanC | SpamapS: nodepool already give you a namespace | 02:40 |
| SpamapS | tristanC: yes, nodepool gave me a namespace, and now I need a way to use it in an untrusted playbook. | 02:40 |
| SpamapS | I can't command: kubectl ... because of the protections. | 02:40 |
| SpamapS | I am going to try a pod actually, as I can just spin up a pod with kubectl in it, copy my .kube/config, and command: there. | 02:41 |
| SpamapS | but.. EOD | 02:41 |
| SpamapS | tristanC: either way, I actually have that review in my queue and started reading it a few hours ago :) | 02:42 |
| SpamapS | ENOTIME | 02:42 |
| *** bhavikdbavishi has joined #zuul | 02:45 | |
| tristanC | SpamapS: thanks, this have been waiting for almost 10 months ;) | 02:48 |
| *** daniel2 has quit IRC | 03:51 | |
| *** rlandy|bbl has quit IRC | 04:26 | |
| *** strigazi has quit IRC | 04:26 | |
| *** kklimonda_ has quit IRC | 04:26 | |
| *** rlandy|bbl has joined #zuul | 04:53 | |
| *** strigazi has joined #zuul | 04:53 | |
| *** kklimonda_ has joined #zuul | 04:53 | |
| *** jesusaur has quit IRC | 05:15 | |
| *** jesusaur has joined #zuul | 05:20 | |
| *** bjackman has joined #zuul | 05:32 | |
| bjackman | Could anyone make a guess why zuul_stream.CallbackModule._read_log doesn't seem to be working for my "command" or "shell" ansible tasks? I just get a spam of the "Waiting on logger" messages and then the "Log Stream did not terminate" | 05:40 |
| bjackman | Could it be a recent change to Ansible? I'm not sure when the problem appeared. | 05:42 |
| *** jhesketh has quit IRC | 05:53 | |
| *** jhesketh has joined #zuul | 05:54 | |
| *** bjackman has quit IRC | 06:10 | |
| *** zbr has joined #zuul | 06:14 | |
| *** zbr|ssbarnea has quit IRC | 06:15 | |
| *** saneax has joined #zuul | 06:21 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: WIP: Support ansible 2.7 https://review.openstack.org/631932 | 06:21 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Switch default ansible version to 2.7 https://review.openstack.org/637424 | 06:21 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: DNM: zuul-stream-functional debugging https://review.openstack.org/640648 | 06:21 |
| mrhillsman | does zuul-web support ssl | 06:23 |
| tobiash | mrhillsman: not directly, just with a reverse proxy that does ssl termination | 06:25 |
| *** bjackman has joined #zuul | 06:28 | |
| *** rlandy|bbl has quit IRC | 06:40 | |
| *** quiquell|off is now known as quiquell | 07:08 | |
| bjackman | When I nmap the test node while a task is running, I don't see the log stream port open; so maybe the console module hasn't started properly? | 07:23 |
| *** pcaruana has joined #zuul | 07:25 | |
| bjackman | (BTW, I was obviously mistaken about the common factor on tasks that are failing to stream the logs - it isn't "shell" and "command" tasks, it's all tasks that actually run on the test node) | 07:26 |
| *** daniel2 has joined #zuul | 07:33 | |
| *** daniel2 has quit IRC | 07:34 | |
| *** quiquell is now known as quiquell|brb | 07:40 | |
| tobiash | bjackman: did you start the zuul_stream module in the base job? | 07:43 |
| bjackman | Ah yes was just about to say! Been starting at zuul-jobs | 07:43 |
| bjackman | I am using prepare-workspace-git, I used to use prepare-workspace | 07:44 |
| bjackman | The latter runs the zuul_console task but the former does not! | 07:44 |
| bjackman | Never realised it had to be started explicitly | 07:44 |
| bjackman | tobiash, do you think I should add a zuul_console task to prepare-workspace git? | 07:45 |
| tobiash | no, to the playbook run by the base job | 07:45 |
| bjackman | But if you use prepare-workspace, you get zuul_console automatically: prepare-workspace-git, I used to use prepare-workspace | 07:46 |
| bjackman | *But if you use prepare-workspace, you get zuul_console automatically: https://github.com/openstack-infra/zuul-jobs/blob/master/roles/prepare-workspace/tasks/main.yaml | 07:46 |
| bjackman | Would be nice to have prepare-workspace-git do the same | 07:46 |
| bjackman | I guess it doesn't really logically belong in that role.. | 07:47 |
| *** daniel2 has joined #zuul | 07:57 | |
| *** daniel2 has quit IRC | 08:00 | |
| *** saneax has quit IRC | 08:00 | |
| *** daniel2 has joined #zuul | 08:06 | |
| *** quiquell|brb is now known as quiquell | 08:07 | |
| *** daniel2 has quit IRC | 08:07 | |
| *** gtema has joined #zuul | 08:07 | |
| *** daniel2 has joined #zuul | 08:08 | |
| *** daniel2 has quit IRC | 08:12 | |
| *** daniel3 has joined #zuul | 08:12 | |
| *** jpena|off is now known as jpena | 08:14 | |
| *** sshnaidm has quit IRC | 08:51 | |
| *** sshnaidm has joined #zuul | 08:52 | |
| *** panda|ruck|off is now known as panda|ruck | 08:52 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: web: add jobs list filter https://review.openstack.org/633652 | 08:58 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: scheduler: add job's tags to the rpc job_list method https://review.openstack.org/633653 | 08:59 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: web: add tags to jobs list https://review.openstack.org/633654 | 08:59 |
| *** saneax has joined #zuul | 09:00 | |
| tobiash | bjackman: yeah, I don't think it should be there... | 09:02 |
| *** hashar has joined #zuul | 09:07 | |
| *** ianw is now known as ianw_pto | 09:52 | |
| *** zbr|ssbarnea has joined #zuul | 10:16 | |
| *** zbr has quit IRC | 10:18 | |
| *** bhavikdbavishi has quit IRC | 10:30 | |
| *** bjackman has quit IRC | 10:42 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: web: add tenant and project scoped, JWT-protected actions https://review.openstack.org/576907 | 10:50 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Allow operator to generate auth tokens through the CLI https://review.openstack.org/636197 | 10:52 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Zuul CLI: allow access via REST https://review.openstack.org/636315 | 10:52 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Add Authorization Rules configuration https://review.openstack.org/639855 | 10:52 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Web: plug the authorization engine https://review.openstack.org/640884 | 10:52 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: [WIP] Zuul Web: add /api/user/actions endpoint https://review.openstack.org/641099 | 10:53 |
| *** odyssey4me has quit IRC | 10:53 | |
| *** odyssey4me has joined #zuul | 10:53 | |
| *** zbr has joined #zuul | 10:56 | |
| *** zbr|ssbarnea has quit IRC | 10:59 | |
| *** bhavikdbavishi has joined #zuul | 11:16 | |
| *** zbr|ssbarnea has joined #zuul | 11:17 | |
| *** zbr has quit IRC | 11:19 | |
| *** hashar is now known as hasharLunch | 11:29 | |
| *** electrofelix has joined #zuul | 11:42 | |
| *** bjackman has joined #zuul | 11:44 | |
| *** zbr has joined #zuul | 11:45 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: Zuul Web: add /api/user/actions endpoint https://review.openstack.org/641099 | 11:46 |
| *** zbr has quit IRC | 11:46 | |
| *** zbr has joined #zuul | 11:47 | |
| *** zbr|ssbarnea has quit IRC | 11:48 | |
| *** rfolco is now known as rfolco|rucker | 12:08 | |
| *** PrinzElvis has joined #zuul | 12:13 | |
| *** hasharLunch is now known as hashar | 12:24 | |
| *** saneax has quit IRC | 12:30 | |
| *** gtema has quit IRC | 12:44 | |
| *** jpena is now known as jpena|lunch | 12:46 | |
| *** saneax has joined #zuul | 12:47 | |
| *** gtema has joined #zuul | 12:54 | |
| *** rfolco|rucker is now known as rfolco|ruck | 12:57 | |
| *** panda|ruck is now known as panda|rover | 12:57 | |
| *** rlandy has joined #zuul | 13:01 | |
| *** jpena|lunch is now known as jpena | 13:50 | |
| tobiash | mordred: are there any known memory issues in openstacksdk when downloading large imaged from openstack? | 13:56 |
| *** quiquell has quit IRC | 14:26 | |
| *** gtema has quit IRC | 14:36 | |
| mordred | tobiash: not KNOWN - but the image download code is way less used/tested than the upload code | 14:45 |
| tobiash | ah, ok | 14:45 |
| tobiash | we need to analyze that further | 14:46 |
| tobiash | will try jemalloc first... | 14:46 |
| mordred | tobiash: (just because the upload code gets hammered by nodepool - so it gets a lot of attention) | 14:46 |
| tobiash | yeah, upload works also with 600gb images ;) | 14:47 |
| mordred | tobiash: the download code is at least attempting to not buffer tons of stuff in memory and to pass file iterators to the underlying requests session | 14:47 |
| * tobiash fears having to debug requests | 14:47 | |
| mordred | tobiash: happy to help dig in if you find issues ... but also, yeah - maybe jemalloc will save the day | 14:48 |
| tobiash | thanks | 14:48 |
| mordred | tobiash: response.iter_content is what we use during the fetch | 14:48 |
| tobiash | in my swift proxy I use response.raw.read in 16k chunks which works quite well so far | 14:51 |
| tobiash | but I have no idea which one is bette from memory point of view | 14:51 |
| tobiash | I'll just wait for the jemalloc result | 14:51 |
| mordred | tobiash: would be happy to switch to response.raw.read if it turns out ot be better - we're doing iter_content in chunk sizes too | 14:53 |
| mordred | so it would be an easy transition | 14:54 |
| tobiash | good to know, but let's see if it's just heap fragmentation | 14:55 |
| corvus | tobiash, dmsimard, pabelanger, mordred: i'd love your thoughts on https://review.openstack.org/641439 | 15:34 |
| tobiash | corvus: that is on the whishlist of my users for quite some time :) | 15:35 |
| tobiash | will look more closely later | 15:35 |
| pabelanger | will review shortly | 15:36 |
| mordred | corvus: yeah - I started reviewing that end of yesterday - then realized I didn't have enough brain powder left | 15:38 |
| corvus | mordred: i assume you dissolved some brain powder in hot water and consumed it this morning? :) | 15:39 |
| mordred | corvus: well, I steeped some brain powder beans in water and strained the result - but yes | 15:39 |
| corvus | essence of brain powder | 15:40 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Remove default zookeeper hosts https://review.openstack.org/635633 | 15:50 |
| corvus | mnaser: i abandoned https://review.openstack.org/635670 because a similar change merged first. thanks! | 16:03 |
| mnaser | corvus: cool no problem :D | 16:04 |
| *** jamesmcarthur has joined #zuul | 16:08 | |
| *** hashar has quit IRC | 16:16 | |
| openstackgerrit | Merged openstack-infra/zuul master: List changed files for all commits between refs https://review.openstack.org/631797 | 16:18 |
| openstackgerrit | Merged openstack-infra/zuul master: Report the build execute_time via mqtt https://review.openstack.org/633945 | 16:18 |
| mordred | corvus: the jemalloc zuul patch is in merge conflict - I can fix it - but would we rather do https://review.openstack.org/#/c/636017/ instead? | 16:25 |
| corvus | mordred: oh interesting. that works for me. | 16:32 |
| corvus | clarkb: you want to push https://review.openstack.org/636017 over the hump? | 16:33 |
| clarkb | corvus: looking | 16:33 |
| clarkb | fwiw openstack's isntall has seen improvements in swap usage as well http://cacti.openstack.org/cacti/graph.php?action=view&local_graph_id=64005&rra_id=all | 16:37 |
| *** hashar has joined #zuul | 16:40 | |
| *** daniel3 has quit IRC | 17:16 | |
| *** daniel2 has joined #zuul | 17:16 | |
| *** saneax has quit IRC | 17:28 | |
| *** jpena is now known as jpena|away | 17:30 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Run tox remote concurrent https://review.openstack.org/640654 | 17:32 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.6 https://review.openstack.org/631931 | 17:32 |
| *** gtema has joined #zuul | 17:38 | |
| *** sshnaidm is now known as sshnaidm|off | 17:47 | |
| *** gtema has quit IRC | 17:48 | |
| * SpamapS grumbles something about not using swap on servers ;) | 18:01 | |
| *** irclogbot_3 has joined #zuul | 18:03 | |
| *** zbr|ssbarnea has joined #zuul | 18:11 | |
| *** bjackman has quit IRC | 18:13 | |
| tobiash | corvus: posted a comment/question on 641439 | 18:13 |
| *** zbr has quit IRC | 18:13 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Make ansible version configurable https://review.openstack.org/637422 | 18:15 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Install ansible during executor startup if needed https://review.openstack.org/640644 | 18:15 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Run tox remote concurrent https://review.openstack.org/640654 | 18:15 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.6 https://review.openstack.org/631931 | 18:15 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.7 https://review.openstack.org/631932 | 18:16 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Switch default ansible version to 2.7 https://review.openstack.org/637424 | 18:16 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: DNM: zuul-stream-functional debugging https://review.openstack.org/640648 | 18:16 |
| *** pcaruana has quit IRC | 18:46 | |
| jlk | You will OOM and you will LIKE IT | 18:56 |
| SpamapS | I will! | 18:57 |
| *** pcaruana has joined #zuul | 19:05 | |
| *** electrofelix has quit IRC | 19:14 | |
| corvus | tobiash: good idea, i'll work on implementing that. | 19:21 |
| *** bhavikdbavishi has quit IRC | 19:21 | |
| *** jesusaur has quit IRC | 19:22 | |
| tobiash | ++ | 19:24 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.7 https://review.openstack.org/631932 | 19:26 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Switch default ansible version to 2.7 https://review.openstack.org/637424 | 19:26 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: DNM: zuul-stream-functional debugging https://review.openstack.org/640648 | 19:26 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: DNM: Enable sar logging for unit tests https://review.openstack.org/641768 | 19:26 |
| *** jesusaur has joined #zuul | 19:27 | |
| *** jpena|away is now known as jpena|off | 19:32 | |
| *** pcaruana has quit IRC | 19:33 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: DNM: Run tox with jemalloc https://review.openstack.org/641776 | 19:44 |
| tobiash | corvus: I'm thinking about the various failures of the quickstart job. One reason is a database migration race: http://logs.openstack.org/71/640871/1/check/zuul-quick-start/477f311/container_logs/web.log | 19:53 |
| tobiash | corvus: wasn't there a change at some time to disable the db migration in zuul-web? | 19:53 |
| *** jesusaur has quit IRC | 19:55 | |
| *** jesusaur has joined #zuul | 20:00 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Fix test race in test_client_dequeue_change_by_ref https://review.openstack.org/640878 | 20:10 |
| *** jamesmcarthur has quit IRC | 20:23 | |
| *** jamesmcarthur_ has joined #zuul | 20:23 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Fix test race in test_container_jobs https://review.openstack.org/641791 | 20:26 |
| tobiash | corvus, mordred, clarkb: yet another test race fix ^ | 20:26 |
| corvus | tobiash: hrm, that sounds familiar | 20:27 |
| corvus | tobiash: i can't find it at the moment | 20:28 |
| tobiash | corvus: you mean the one in test_container_jobs? | 20:29 |
| corvus | tobiash: no sorry, the migration | 20:29 |
| tobiash | ah | 20:30 |
| corvus | tobiash: https://review.openstack.org/610696 | 20:30 |
| tobiash | ah it's abandoned, that's why I didn't find it | 20:30 |
| corvus | tobiash: it looks like i abandoned that because it wasn't the cause of that particular deadlock. but maybe it is the cause of this new one :) | 20:30 |
| tobiash | yes, probably | 20:31 |
| tobiash | maybe it's time to survive it | 20:31 |
| tobiash | corvus: shall I reopen and rebase it? | 20:32 |
| corvus | tobiash: i'll do it | 20:32 |
| tobiash | k | 20:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: SQL: only create tables in scheduler https://review.openstack.org/610696 | 20:36 |
| tobiash | mordred, corvus: is it possible somehow to find out which test case triggered that alarm clock here: http://logs.openstack.org/24/637424/15/check/tox-py36/aa5b001/job-output.txt.gz#_2019-03-07_20_11_10_469213 ? | 20:37 |
| corvus | tobiash: is it this one? http://logs.openstack.org/24/637424/15/check/tox-py36/aa5b001/job-output.txt.gz#_2019-03-07_20_17_44_457266 | 20:38 |
| corvus | clarkb: ^ | 20:38 |
| tobiash | oh that's probably it, thanks | 20:39 |
| tobiash | corvus: commented on 610696 | 20:40 |
| tobiash | corvus: yeah, test_plugins takes isolated locally already around 40s, I'll increase the timeout for this test case | 20:43 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: SQL: only create tables in scheduler https://review.openstack.org/610696 | 20:44 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.6 https://review.openstack.org/631931 | 20:45 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Support ansible 2.7 https://review.openstack.org/631932 | 20:45 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Switch default ansible version to 2.7 https://review.openstack.org/637424 | 20:45 |
| tobiash | corvus: do we want to block 630078 on the removal of the debug prints? | 21:00 |
| tobiash | (buildset route) | 21:01 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix test race in test_client_dequeue_change_by_ref https://review.openstack.org/640878 | 21:02 |
| *** hashar has quit IRC | 21:02 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Increase timeout of test_plugins https://review.openstack.org/641803 | 21:06 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Docker: use the buildset registry if defined https://review.openstack.org/641805 | 21:06 |
| corvus | tobiash: nah, we can clean that up later i think. | 21:07 |
| tobiash | kk | 21:08 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Allow soft job dependencies https://review.openstack.org/641439 | 21:21 |
| corvus | tobiash: updated with your suggestion, thanks :) | 21:21 |
| tobiash | looking :) | 21:22 |
| tobiash | corvus: lgtm | 21:34 |
| tobiash | wohoo, the multi ansible stack is green https://review.openstack.org/#/q/status:open+project:openstack-infra/zuul+branch:master+topic:multi-ansible | 21:35 |
| tobiash | \o/ | 21:35 |
| corvus | tobiash: yay! i guess it's time to get serious about release planning and reviewing for that | 21:40 |
| tobiash | yes I think so | 21:40 |
| tobiash | regarding release planning, it will be backwards compatible if zuul can access pypi during startup | 21:41 |
| tobiash | by default it will install ansible into the state dir on startup if it's not pre-installed | 21:41 |
| tobiash | that dir is guaranteed to be writable | 21:42 |
| tobiash | ok, well the thing that is not necessarily backwards compatible is the default ansible version ;) | 21:43 |
| tobiash | but that's overridable per zuul.conf, tenant or job | 21:43 |
| tobiash | that will make sure that ansible upgrades are as easy as possible even in large multi-tenant installations | 21:44 |
| mordred | yay! but also - oh my I have to review that now | 21:48 |
| tobiash | yes, that was quite some work | 21:49 |
| tobiash | but most of it is new symlinks | 21:49 |
| tobiash | so looks larger than it is | 21:49 |
| openstackgerrit | Merged openstack-infra/zuul master: web: add buildsets page https://review.openstack.org/630041 | 21:50 |
| openstackgerrit | Merged openstack-infra/zuul master: web: add /{tenant}/buildset/{uuid} route https://review.openstack.org/630078 | 21:50 |
| *** rlandy has quit IRC | 21:53 | |
| tobiash | corvus: if you have time for a discussion tomorrow, I'd have an interesting architecture design question around spreading zuul-executors over multiple openshift clusters | 22:02 |
| *** rlandy has joined #zuul | 22:02 | |
| *** jamesmcarthur_ has quit IRC | 22:03 | |
| corvus | tobiash: ok, we should be able to find some time :) | 22:04 |
| tobiash | thanks :) | 22:04 |
| *** jesusaur has quit IRC | 22:09 | |
| *** jesusaur has joined #zuul | 22:13 | |
| openstackgerrit | Merged openstack-infra/zuul master: Log exception on module failure with empty stdout https://review.openstack.org/640650 | 22:18 |
| openstackgerrit | Merged openstack-infra/zuul master: Allow soft job dependencies https://review.openstack.org/641439 | 22:18 |
| tobiash | \o/ | 22:18 |
| *** jesusaur has quit IRC | 22:21 | |
| *** jesusaur has joined #zuul | 22:31 | |
| openstackgerrit | David Shrewsbury proposed openstack-infra/zuul-preview master: DNM: Port in changes from the c++ version https://review.openstack.org/641089 | 22:37 |
| *** jesusaur has quit IRC | 22:41 | |
| *** jesusaur has joined #zuul | 22:42 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Wait for command processor thread to stop https://review.openstack.org/641854 | 22:44 |
| Shrews | SpamapS: it isn't necessary to include the Cargo.lock file in a rust repo, is it? That seems to be generated from the .toml file at build time so not necessary to include, IMO | 22:44 |
| SpamapS | Shrews: correct. The reason to include it is if you want to make sure your build works. Usually CLI tools and daemons ship a lock, libraries should never ship one. | 22:46 |
| Shrews | SpamapS: ah, i see | 22:47 |
| mordred | Shrews: https://doc.rust-lang.org/cargo/faq.html#why-do-binaries-have-cargolock-in-version-control-but-not-libraries | 23:14 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Docker: use the buildset registry if defined https://review.openstack.org/641805 | 23:14 |
| Shrews | mordred: oh, helpful. thx | 23:14 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!