| *** jamesmcarthur has joined #zuul | 00:17 | |
| *** jamesmcarthur has quit IRC | 00:26 | |
| *** hamalq has quit IRC | 00:36 | |
| *** rfolco has quit IRC | 00:50 | |
| *** jamesmcarthur has joined #zuul | 01:02 | |
| *** wuchunyang has joined #zuul | 01:03 | |
| *** jamesmcarthur has quit IRC | 01:04 | |
| *** Goneri has quit IRC | 01:15 | |
| *** jamesmcarthur has joined #zuul | 01:17 | |
| *** wuchunyang has quit IRC | 01:21 | |
| *** jamesmcarthur has quit IRC | 01:22 | |
| *** jamesmcarthur has joined #zuul | 01:26 | |
| *** jamesmcarthur has quit IRC | 01:39 | |
| *** jamesmcarthur has joined #zuul | 01:47 | |
| *** swest has quit IRC | 01:58 | |
| *** jamesmcarthur has quit IRC | 02:01 | |
| *** jamesmcarthur has joined #zuul | 02:02 | |
| *** jamesmcarthur has quit IRC | 02:06 | |
| *** swest has joined #zuul | 02:12 | |
| *** tdasilva_ has joined #zuul | 02:24 | |
| *** tdasilva has quit IRC | 02:25 | |
| *** tdasilva_ is now known as tdasilva | 02:25 | |
| *** jamesmcarthur has joined #zuul | 02:30 | |
| *** vblando has quit IRC | 02:36 | |
| *** mnasiadka has quit IRC | 02:36 | |
| *** samccann has quit IRC | 02:36 | |
| *** Open10K8S has quit IRC | 02:37 | |
| *** Open10K8S has joined #zuul | 02:37 | |
| *** samccann has joined #zuul | 02:38 | |
| *** mnasiadka has joined #zuul | 02:38 | |
| *** ChrisShort has quit IRC | 02:38 | |
| *** vblando has joined #zuul | 02:40 | |
| *** jamesmcarthur has quit IRC | 02:45 | |
| *** ChrisShort has joined #zuul | 02:46 | |
| *** jamesmcarthur has joined #zuul | 02:46 | |
| *** bhavikdbavishi has joined #zuul | 02:49 | |
| *** jamesmcarthur has quit IRC | 02:50 | |
| *** bhavikdbavishi1 has joined #zuul | 02:54 | |
| *** bhavikdbavishi has quit IRC | 02:55 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 02:55 | |
| *** jamesmcarthur has joined #zuul | 02:56 | |
| *** wuchunyang has joined #zuul | 03:00 | |
| *** jamesmcarthur has quit IRC | 03:39 | |
| *** ysandeep|away is now known as ysandeep | 04:01 | |
| *** wuchunyang has quit IRC | 04:04 | |
| *** evrardjp has quit IRC | 04:33 | |
| *** evrardjp has joined #zuul | 04:33 | |
| *** vishalmanchanda has joined #zuul | 04:41 | |
| *** bhavikdbavishi has quit IRC | 05:02 | |
| *** dmellado has quit IRC | 05:12 | |
| *** sgw has quit IRC | 05:22 | |
| *** bhavikdbavishi has joined #zuul | 05:23 | |
| *** bhavikdbavishi has quit IRC | 05:34 | |
| *** bhavikdbavishi has joined #zuul | 05:40 | |
| *** ysandeep is now known as ysandeep|brb | 05:41 | |
| *** marios has joined #zuul | 05:54 | |
| *** ysandeep|brb is now known as ysandeep | 06:11 | |
| *** y2kenny has quit IRC | 06:28 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 06:48 |
|---|---|---|
| *** jcapitao has joined #zuul | 07:13 | |
| *** bhagyashris|pto is now known as bhagyashris | 07:16 | |
| *** bhavikdbavishi has quit IRC | 07:18 | |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect merger to Zookeeper https://review.opendev.org/716221 | 07:19 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect executor to Zookeeper https://review.opendev.org/716262 | 07:19 |
| openstackgerrit | Jan Kubovy proposed zuul/zuul master: Connect fingergw to Zookeeper https://review.opendev.org/716875 | 07:19 |
| *** hashar has joined #zuul | 07:24 | |
| *** yolanda has joined #zuul | 07:26 | |
| *** sshnaidm|afk is now known as sshnaidm|ruck | 07:28 | |
| *** tosky has joined #zuul | 07:28 | |
| *** bhagyashris is now known as bhagyashris|lunc | 07:29 | |
| *** iurygregory has quit IRC | 07:42 | |
| *** jpena|off is now known as jpena | 07:50 | |
| *** wuchunyang has joined #zuul | 07:54 | |
| *** bhavikdbavishi has joined #zuul | 07:55 | |
| *** iurygregory has joined #zuul | 08:01 | |
| *** nils has joined #zuul | 08:07 | |
| *** bhagyashris|lunc is now known as bhagyashris | 08:37 | |
| *** masterpe has quit IRC | 08:38 | |
| *** masterpe has joined #zuul | 08:51 | |
| *** bhavikdbavishi has quit IRC | 08:53 | |
| *** bhavikdbavishi has joined #zuul | 08:53 | |
| zbr | does zuul have a limit on job description length? if using long, even multiline ones, would create UI problems? | 08:55 |
| *** bhavikdbavishi1 has joined #zuul | 08:56 | |
| *** bhavikdbavishi has quit IRC | 08:58 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 08:58 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 09:04 |
| *** sshnaidm|ruck has quit IRC | 09:14 | |
| *** sshnaidm has joined #zuul | 09:23 | |
| *** sshnaidm has quit IRC | 09:42 | |
| *** sshnaidm has joined #zuul | 09:55 | |
| *** wuchunyang has quit IRC | 10:01 | |
| *** jpena has quit IRC | 10:07 | |
| *** mugsie has quit IRC | 10:07 | |
| *** mugsie has joined #zuul | 10:10 | |
| *** marios has quit IRC | 10:40 | |
| *** wuchunyang has joined #zuul | 10:46 | |
| *** hashar has quit IRC | 10:47 | |
| *** ysandeep is now known as ysandeep|afk | 10:53 | |
| *** jcapitao is now known as jcapitao_lunch | 11:03 | |
| *** wuchunyang has quit IRC | 11:18 | |
| *** marios has joined #zuul | 11:27 | |
| *** bhagyashris is now known as bhagyashris|brb | 11:32 | |
| *** jpena has joined #zuul | 11:32 | |
| tobiash | zbr: afaik there is no real limit on that | 11:37 |
| tobiash | many jobs use multiline descriptions | 11:38 |
| *** ysandeep|afk is now known as ysandeep | 11:39 | |
| *** bhavikdbavishi has quit IRC | 11:46 | |
| *** jpena is now known as jpena|lunch | 11:46 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 11:55 |
| *** bhavikdbavishi has joined #zuul | 11:57 | |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: GitHub Reporter: Fix User Email in Merge Commit Message https://review.opendev.org/738590 | 12:02 |
| *** rfolco has joined #zuul | 12:02 | |
| *** mordred has quit IRC | 12:04 | |
| *** bschanzel has joined #zuul | 12:06 | |
| *** mordred has joined #zuul | 12:06 | |
| *** jamesmcarthur has joined #zuul | 12:07 | |
| *** rlandy has joined #zuul | 12:08 | |
| *** bhagyashris|brb is now known as bhagyashris | 12:10 | |
| *** jcapitao_lunch is now known as jcapitao | 12:17 | |
| *** hashar has joined #zuul | 12:22 | |
| zbr | tobiash: thanks. i asked as I seen in some places in UI the description is appended after job name. Good to know we can properly document them. | 12:24 |
| openstackgerrit | Simon Westphahl proposed zuul/nodepool master: Ignore unparsable/empty image upload ZNode data https://review.opendev.org/738013 | 12:26 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Support per branch change queues https://review.opendev.org/718531 | 12:27 |
| openstackgerrit | Simon Westphahl proposed zuul/nodepool master: Ignore unparsable/empty image upload ZNode data https://review.opendev.org/738013 | 12:31 |
| *** jpena|lunch is now known as jpena | 12:34 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Move queue from pipeline to project https://review.opendev.org/720182 | 12:36 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add optional support for circular dependencies https://review.opendev.org/685354 | 12:36 |
| *** bhavikdbavishi has quit IRC | 12:38 | |
| *** jamesmcarthur has quit IRC | 12:39 | |
| *** wuchunyang has joined #zuul | 12:41 | |
| *** wuchunyang has quit IRC | 12:46 | |
| *** ysandeep is now known as ysandeep|afk | 12:47 | |
| *** rlandy is now known as rlandy|training | 12:55 | |
| *** ysandeep|afk is now known as ysandeep | 12:55 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: WIP: Make repo state buildset global https://review.opendev.org/738603 | 12:56 |
| *** sgw has joined #zuul | 13:00 | |
| openstackgerrit | Merged zuul/zuul-jobs master: upload-git-mirror: use retries to avoid races https://review.opendev.org/738187 | 13:14 |
| *** bschanzel has quit IRC | 13:20 | |
| zbr | had anyone considered moving test_result_table under change metadata (first column) instead of the narrow middle column with code-review? | 13:21 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Refactor github auth handling into its own class https://review.opendev.org/710034 | 13:22 |
| tobiash | clarkb: I've addressed one of your comments there and responded on the other ^ | 13:23 |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: [TEST DNM] Help Debugging Slow kubectl Connections https://review.opendev.org/738619 | 13:35 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Remove some unused variables https://review.opendev.org/738620 | 13:37 |
| zbr | clarkb: is there any way to run ansible nested in zuul but to still display tasks using the same UI? | 13:47 |
| zbr | maybe not now, but if this is doable in the future | 13:47 |
| clarkb | zbr: I don't know | 13:48 |
| zbr | asking this because we need full freedom regarding what we run: collections, modules, plugins, stuff that zuul does not allow for security reasons | 13:48 |
| AJaeger | zuul-jobs-maint, https://review.opendev.org/737352 updates a README, please review | 13:53 |
| openstackgerrit | Benjamin Schanzel proposed zuul/zuul master: [TEST DNM] Help Debugging Slow kubectl Connections https://review.opendev.org/738619 | 13:56 |
| *** Goneri has joined #zuul | 13:56 | |
| corvus | zbr: it's not possible now, but we have thought about it and i think it would be possible in the future; it's mostly a matter of adding the zuul json callback to the nested ansible. my guess is putting that callback in a collection is going to be the first step. | 14:10 |
| zbr | corvus: thanks. that was my guess but needed confirmation. | 14:11 |
| zbr | i guess I do not need to explain more why i see it useful | 14:11 |
| corvus | nope i think it's a great idea :) | 14:12 |
| *** ysandeep is now known as ysandeep|away | 14:17 | |
| openstackgerrit | Matthieu Huin proposed zuul/zuul master: List builds, buildsets results in constants https://review.opendev.org/738632 | 14:19 |
| AJaeger | zuul-jobs-maint, a small change for review, please: https://review.opendev.org/733419 | 14:24 |
| mhu | Hey there, can someone explain to me again the difference between CANCELED and ABORTED as build results? | 14:26 |
| mhu | nevermind, found it in my history | 14:27 |
| mhu | mhu: canceled (note the spelling) is by request of the scheduler, aborted is unexpected error on executo | 14:27 |
| mhu | with that said, I need help with descriptions of the various possible build results: https://review.opendev.org/#/c/738632/1/doc/source/reference/jobs.rst | 14:35 |
| openstackgerrit | Merged zuul/zuul-jobs master: prepare-workspace: Add Role Variable in README.rst https://review.opendev.org/737352 | 14:45 |
| *** sgw1 has quit IRC | 14:50 | |
| *** bschanzel has joined #zuul | 14:51 | |
| bschanzel | Hi there, some weeks ago I proposed https://review.opendev.org/#/c/734580/, which fixes the `Reviewed-by` clauses in GitHub merge commit messages. Unfortunately, it contained a bug where the email addresses of reviewers were always "None". Here's a fix https://review.opendev.org/#/c/738590/. The downside of this fix is that we're doing additional | 14:54 |
| bschanzel | GitHub API requests per reviewer and PR on GitHub merges (unless a user is cached). | 14:54 |
| *** sgw1 has joined #zuul | 14:55 | |
| openstackgerrit | Merged zuul/zuul-jobs master: Return upload_results in upload-logs-swift role https://review.opendev.org/733564 | 15:01 |
| tobiash | bschanzel: we added a pretty decent user caching a few months ago (https://review.opendev.org/710985) so I don't think that's becoming a problem | 15:03 |
| bschanzel | tobiash: that's what I meant with "unless a user is cached" ;-) | 15:04 |
| *** bschanzel has quit IRC | 15:08 | |
| tristanC | it seems like the latest centos-8 image in opendev's zuul is having some sort of a random network issue: jobs are being retried a lot and when it happens three times in a row, the log_url is a finger... any idea what can we do to debug and understand what is going on in such situation? | 15:15 |
| clarkb | tristanC: yes, we've already spent a bit of time debugging this... | 15:16 |
| clarkb | fomr the opendev side we've confirmed there is no kernel panic | 15:16 |
| tristanC | we also suspect a kernel bug, would it be possible to grab the console log before node deletion, e.g. adding a post-console-log toggle in nodepool configuration? | 15:16 |
| clarkb | but the host networking has crashed. | 15:16 |
| clarkb | I've offered to configure zuul holds if I can be told what to hold | 15:16 |
| clarkb | then we can hope that a reboot of the host makes it debuggable | 15:17 |
| clarkb | I've also offered to set up test nodes in our clouds so that people can try to reproduce or they can try to reproduce lcoally using our images | 15:17 |
| clarkb | from the zuul side I'm not sure there is much that can be done. The test node is going away. Maybe we can improve the hold process to hold instances under more specific circumstances so that it is easier tocatch a host in that situation | 15:17 |
| tristanC | clarkb: thank you for all the help, i meant to ask here something that zuul/nodepool/dib could do to fix that kind of issue, not this one in particular | 15:17 |
| clarkb | other than that I don't know | 15:18 |
| clarkb | yes, I wanted to give background to the channel on what sort of debugging had been done | 15:18 |
| clarkb | ultimately if zuul can't reach the host its choices are limited | 15:18 |
| clarkb | not deleting the host may be useful though and doing that reliably would need updates to the hold behavior | 15:18 |
| tristanC | about the finger url, is there a reason why the executor could not ignore the node log collection failure and at least publish the local job-output file? | 15:20 |
| clarkb | there are two specific issues with holds today. The first is we have to wait for retry failure which is three attempts (though we can override that to a single attempt on a per job basis). THe other issue is tripleo has many other failures happening too so being able to filter for "ansibel says the host is unreachable" would be useful | 15:20 |
| tristanC | assuming the log are not published because the node is not reachable | 15:20 |
| fungi | tristanC: also about capturing server console logs in particular, for this case we did and there was nothing of interest, just some unrelated iptables drops and selinux enforcement messages, no errors or kernel panics | 15:21 |
| fungi | but having zuul capture that would not be easy. our executors don't have the api access credentials necessary to get the server consoles | 15:21 |
| clarkb | tristanC: the log streaming is on the remote host not the executor right? | 15:22 |
| clarkb | the information available to the executor is limited iirc. But we could probably publsih something that said network connectivity was lost and the ansible return code? | 15:23 |
| tristanC | fungi: right, i was thinking nodepool would have to get the console log, though it's probably not easy to configure when it should do that. e.g. nodepool may not know the job result of a nodeset it is deleting | 15:24 |
| *** hashar is now known as hasharAway | 15:24 | |
| tristanC | clarkb: iirc the job-output.txt is written by the zuul_stream callback and it is local on the executor build work directory | 15:24 |
| clarkb | tristanC: ah right it relies on the remote port 19885 daemon to stream that hosts data but then it aggregates. Thats how we get the interleaved content for job output | 15:25 |
| clarkb | tristanC: that would all be in the post playbook itself. You'd need to run the remote collection then always run upload for job-output separately | 15:26 |
| clarkb | I think its possible just needs some changes to how post.yaml processes tasks | 15:26 |
| clarkb | (also as a side note it wouldn't help debug the tripleo problem any more than we've already done, the log doesn't have any data other than this script is running and 20 minutes later the network is gone) | 15:27 |
| tristanC | that seems to already be split accross two playbooks https://opendev.org/opendev/base-jobs/src/branch/master/playbooks/base/ post and post-logs | 15:27 |
| clarkb | making it easier to get to that point would be good, but it won't solve this for us aiui | 15:27 |
| tristanC | clarkb: having the logs might help discover if the pre-failure happens after a specific task | 15:28 |
| clarkb | tristanC: yes I'm saying we already have that info and it doesn't help much | 15:28 |
| clarkb | toci quickstart script execution is the task where it fails | 15:28 |
| *** mgoddard has quit IRC | 15:28 | |
| *** vishalmanchanda has quit IRC | 15:30 | |
| tristanC | darn, alright thank you for all the informations. I guess that specific failure is going to need more investigation, we'll try to reproduce it in rdo zuul | 15:30 |
| tristanC | perhaps having the node console log and trying to rescue the job-output in similar situation might be useful for other retry-limit investigation | 15:31 |
| clarkb | ya I think it could be useful in a more general case | 15:32 |
| clarkb | tristanC: looking at the executor look for running playbooks we run all post playbooks even if unreachable is true | 15:33 |
| clarkb | that makes me wonder if post does curation of the output for post-logs even in the local build dir | 15:33 |
| clarkb | (and that fails because it can't talk to the remote) | 15:33 |
| tristanC | the post-logs.yaml playbook seems to be running to hosts: localhost though | 15:34 |
| *** mgoddard has joined #zuul | 15:36 | |
| clarkb | tristanC: right but maybe that file isn't where post-logs.yaml will copy from if post.yaml failed? | 15:37 |
| clarkb | basically it could be that post-logs.yaml is running and having no work to do? | 15:37 |
| tristanC | yeah i understand, though it should at least zuul_return a log_url | 15:39 |
| clarkb | tristanC: maybe that loop in zuul/executor/server.py is short circuiting because an abort is also happening? | 15:42 |
| tristanC | clarkb: i'm looking at the exact piece of code indeed, but in that case, wouldn't the job result be ABORTED instead of RETRY_LIMIT? | 15:43 |
| clarkb | retry limit is handled in the client based on those results but ya it appears we only get retry limit if we haven't aborted | 15:46 |
| fungi | it's result_unreachable on the executor side | 15:46 |
| clarkb | because we'll always retry the aborted case regardless of attempts | 15:46 |
| fungi | that either gets retried or reinterpreted as retry_limit by the scheduler | 15:46 |
| *** hamalq has joined #zuul | 15:50 | |
| *** sshnaidm has quit IRC | 15:51 | |
| *** hamalq_ has joined #zuul | 15:52 | |
| *** sshnaidm has joined #zuul | 15:53 | |
| *** hamalq has quit IRC | 15:55 | |
| corvus | tristanC, clarkb: i'm trying to digest that conversation -- it seems that the first issue is that we got finger urls instead of logs for retry_limit jobs, yeah? | 15:57 |
| *** hasharAway is now known as hashar | 15:58 | |
| clarkb | corvus: yes, and while we don't expect any logs from the unreachable remote node we do expect a job-output file or two? | 15:58 |
| corvus | clarkb: right | 15:58 |
| corvus | even in the case of an unreachable host, i would expect the post-logs.yaml playbook to run and upload job-output* and return a log url | 15:59 |
| clarkb | I'm wondering now if generate-zuul-manifest may dep on something post.yaml does when it runs in post-logs.yaml | 15:59 |
| clarkb | fungi is trying to pull up some examples in our executor logs so that we can check if we get more than one error | 15:59 |
| corvus | clarkb: do you have a failed job link handy so i can go look? | 15:59 |
| corvus | fungi: ^? | 15:59 |
| corvus | oh i'll switch to #openstack-infra | 15:59 |
| *** jpena is now known as jpena|off | 16:03 | |
| *** marios has quit IRC | 16:04 | |
| *** jpena|off is now known as jpena | 16:06 | |
| corvus | tristanC: i've been digging, and i can't find any retry_limit jobs with no logs. there are certainly some without logs from the remote host, but that's expected if the host is unreachable. but they all at least have the job-output.txt and zuul manifest. | 16:07 |
| tristanC | corvus: the buildset currently at the top of the gate https://zuul.opendev.org/t/openstack/status has one | 16:08 |
| clarkb | ah is it that if they get in the db we're selecting the ones without the issue? | 16:10 |
| tristanC | i also though that the executor would only perform post playbook of successful pre, and if the pre failure happen on the base job, then that would explain the empty log url. But it seems like zuul tries to perform every post, without looking for matching successful pre | 16:10 |
| clarkb | I see it | 16:13 |
| clarkb | getting a link to share | 16:13 |
| clarkb | we're failing in run phase with host unreachable so we hit https://opendev.org/zuul/zuul/src/branch/master/zuul/executor/server.py#L1413-L1416 | 16:13 |
| clarkb | we could put post-logs.yaml in the cleanup phase to address this | 16:14 |
| clarkb | (I'd need to think a bit about what the remifactions of that are though) | 16:14 |
| *** holser has quit IRC | 16:15 | |
| *** holser has joined #zuul | 16:16 | |
| tristanC | clarkb: oh nice, great finding | 16:17 |
| *** sshnaidm is now known as sshnaidm|ruck | 16:18 | |
| tristanC | clarkb: though how can this result in a job with retry_limit result and a finger url? | 16:18 |
| clarkb | we never run post in that case and so log url isnt set? something like that | 16:19 |
| corvus | so the issue is that in this case the finger url isn't updated in the status page | 16:19 |
| corvus | clarkb: post always runs | 16:19 |
| avass | corvus: unless it's dequeued | 16:19 |
| avass | :) | 16:19 |
| corvus | avass: this is not happening in this case | 16:20 |
| tristanC | clarkb: the block you highlight is nested in a `if not pre_failed:` block | 16:20 |
| corvus | clarkb: that's why we couldn't find any in the db where post didn't run. | 16:20 |
| tristanC | oh my bad, could retry_limit happen without pre_failed? | 16:20 |
| corvus | when that change reports, the logs will be available at the usual place, just like the other retry limits | 16:20 |
| clarkb | corvus: I grepped the logs for that specific job on ze03 and it did not run post | 16:21 |
| clarkb | it went from run to cleanup | 16:21 |
| corvus | however, it seems that the issue here is that we expect the logs to have been uploaded to object storage, and we don't put that url in the status page for retry_limits | 16:21 |
| tristanC | corvus: looking at https://zuul.opendev.org/t/openstack/builds?project=openstack%2Ftripleo-heat-templates&result=RETRY_LIMIT , the top few result doesn't have any log_url, e.g. : https://zuul.opendev.org/t/openstack/build/6ec193d4d5c04e569cad7444c7be0b66 | 16:22 |
| corvus | tristanC, clarkb, okay that's more like it, thanks :) | 16:22 |
| clarkb | and my read of the code I linked is we short circuit if an unexpected ansible error occurs in run | 16:23 |
| clarkb | in that casepost does not run | 16:23 |
| corvus | clarkb: the "return None" at the bottom if "if not pre_failed"? | 16:24 |
| fungi | corvus: i posted the early examples of retry_limit jobs with no logs in #openstack-infra | 16:24 |
| fungi | just now | 16:24 |
| fungi | sorry, i had an emergency distraction come up | 16:24 |
| clarkb | corvus: yes | 16:25 |
| corvus | clarkb: yeah, i agree we should be able to change that to do what the pre-runs do | 16:25 |
| corvus | i can do that | 16:25 |
| corvus | clarkb: unless you're already in progress? | 16:26 |
| clarkb | corvus: I'm not. trying to make breakfast | 16:27 |
| corvus | k. on it. | 16:27 |
| tristanC | iiuc, if the node connection is lost during the run phase, then the job is retried without post executions, and on the last attempt the job result will be RETRY_LIMIT? even though the failure happened during the run phase | 16:36 |
| clarkb | yes | 16:37 |
| *** jcapitao has quit IRC | 16:39 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Run post-run playbooks on unreachable https://review.opendev.org/738668 | 16:40 |
| corvus | tristanC: yes -- unreachable nodes ar | 16:41 |
| corvus | grr | 16:41 |
| corvus | tristanC: yes -- we retry jobs with unreachable nodes because we presume some external network issue caused the problem (that assumption is wrong in this case, but it's a conservative assumption) | 16:41 |
| tristanC | clarkb: corvus: that makes sense, thanks a lot | 16:44 |
| corvus | tristanC: thank you :) | 16:46 |
| *** jpena is now known as jpena|off | 16:52 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 17:00 |
| clarkb | corvus: looks like tristanC left a comment on that fix. I don't know what the answer is off the top of my head | 17:06 |
| clarkb | but the fix looks good to me otherwise | 17:06 |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 17:11 |
| corvus | clarkb, tristanC: replied | 17:34 |
| *** hashar is now known as hasharAway | 17:40 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Use a temporary registry with buildx https://review.opendev.org/738517 | 17:51 |
| *** nils has quit IRC | 17:53 | |
| -openstackstatus- NOTICE: Due to a flood of connections from random prefixes, we have temporarily blocked all AS4837 (China Unicom) source addresses from access to the Git service at opendev.org while we investigate further options. | 18:24 | |
| openstackgerrit | Simon Westphahl proposed zuul/zuul master: Ensure refs for recent branches are not GCed https://review.opendev.org/738454 | 18:49 |
| *** fbo has quit IRC | 19:01 | |
| *** hasharAway is now known as hashar | 19:01 | |
| *** sshnaidm|ruck is now known as sshnaidm|bbl | 19:02 | |
| *** ianw_pto is now known as ianw | 19:04 | |
| *** yolanda has quit IRC | 19:05 | |
| *** rlandy|training is now known as rlandy | 19:36 | |
| *** armstrongs has joined #zuul | 19:45 | |
| *** armstrongs has quit IRC | 19:54 | |
| *** hashar is now known as hasharAway | 19:55 | |
| *** noonedeadpunk has quit IRC | 20:03 | |
| *** noonedeadpunk has joined #zuul | 20:03 | |
| *** tobiash has quit IRC | 20:06 | |
| *** tobiash has joined #zuul | 20:07 | |
| *** yolanda has joined #zuul | 20:11 | |
| *** hasharAway has quit IRC | 20:28 | |
| *** hashar has joined #zuul | 20:29 | |
| *** sshnaidm|bbl has quit IRC | 20:34 | |
| *** sshnaidm|bbl has joined #zuul | 20:35 | |
| *** sshnaidm|bbl is now known as sshnaidm|afk | 20:46 | |
| *** stevthedev has quit IRC | 21:14 | |
| *** stevthedev has joined #zuul | 21:14 | |
| *** jbryce has quit IRC | 21:15 | |
| *** _erlon_ has quit IRC | 21:15 | |
| *** jbryce has joined #zuul | 21:15 | |
| *** _erlon_ has joined #zuul | 21:15 | |
| *** tobiash has quit IRC | 22:04 | |
| *** tobiash has joined #zuul | 22:06 | |
| *** hashar has quit IRC | 22:07 | |
| *** rlandy is now known as rlandy|afk | 22:15 | |
| *** rfolco has quit IRC | 22:17 | |
| *** piotrowskim has quit IRC | 22:22 | |
| *** rfolco has joined #zuul | 22:33 | |
| *** tosky has quit IRC | 22:57 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Use a temporary registry with buildx https://review.opendev.org/738517 | 23:15 |
| corvus | roles/write-inventory/tasks/main.yaml:1: [E106] Role name write-inventory does not match ``^[a-z][a-z0-9_]+$`` pattern | 23:16 |
| corvus | is this a new ansible-lint thing? | 23:16 |
| corvus | that's apparently true for collections, but i don't think that applies here | 23:18 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Ignore ansible lint E106 https://review.opendev.org/738716 | 23:19 |
| corvus | zuul-jobs-maint: we may want to merge that asap ^ | 23:20 |
| *** wuchunyang has joined #zuul | 23:20 | |
| clarkb | looking | 23:22 |
| *** wuchunyang has quit IRC | 23:24 | |
| fungi | yeah, this came up before, it was hotly debated in the ansible community apparently, but the implementation plan for collections prevents being able to use hyphens in their names | 23:27 |
| corvus | yep. i won't try to stop the locomotive, just point out that this car isn't hooked up to it :) | 23:28 |
| fungi | i guess ansible-lint only just got around to adding a rule to check for it, and assumes all roles are collections now | 23:28 |
| fungi | and all restaurants are now taco bell | 23:28 |
| openstackgerrit | Merged zuul/zuul-jobs master: Ignore ansible lint E106 https://review.opendev.org/738716 | 23:33 |
| corvus | avass: looks like https://review.opendev.org/738517 should be ready now -- if you have a minute tomorrow to take a look at that, we can decide if we want to squash that into your change | 23:35 |
| *** hamalq_ has quit IRC | 23:38 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!