*** sshnaidm has quit IRC | 00:04 | |
*** sshnaidm has joined #zuul | 00:05 | |
*** zenkuro has quit IRC | 00:40 | |
*** Goneri has quit IRC | 01:32 | |
*** sshnaidm has quit IRC | 02:19 | |
*** sshnaidm has joined #zuul | 02:25 | |
*** bhavikdbavishi has joined #zuul | 03:03 | |
*** bhavikdbavishi1 has joined #zuul | 03:54 | |
*** bhavikdbavishi has quit IRC | 03:55 | |
*** bhavikdbavishi1 is now known as bhavikdbavishi | 03:55 | |
*** vishalmanchanda has joined #zuul | 04:22 | |
*** bhavikdbavishi has quit IRC | 04:26 | |
*** bhavikdbavishi has joined #zuul | 04:26 | |
*** wuchunyang has joined #zuul | 04:52 | |
*** wuchunyang has quit IRC | 04:57 | |
*** wuchunyang has joined #zuul | 04:58 | |
*** evrardjp has quit IRC | 05:33 | |
*** evrardjp has joined #zuul | 05:33 | |
*** jfoufas1 has joined #zuul | 05:58 | |
*** bhavikdbavishi1 has joined #zuul | 05:58 | |
*** bhavikdbavishi has quit IRC | 06:00 | |
*** bhavikdbavishi1 is now known as bhavikdbavishi | 06:00 | |
*** jfoufas1 has quit IRC | 06:12 | |
*** saneax has joined #zuul | 06:42 | |
*** jfoufas1 has joined #zuul | 06:45 | |
openstackgerrit | Ian Wienand proposed zuul/nodepool master: rename imagesdir/elementdir with underscore https://review.opendev.org/c/zuul/nodepool/+/764279 | 06:52 |
---|---|---|
openstackgerrit | Ian Wienand proposed zuul/nodepool master: Add images-dir-required-free option https://review.opendev.org/c/zuul/nodepool/+/764280 | 06:52 |
*** bhavikdbavishi has quit IRC | 06:55 | |
*** bhavikdbavishi has joined #zuul | 07:21 | |
*** mach1na has joined #zuul | 07:22 | |
*** rpittau|afk is now known as rpittau | 07:26 | |
*** reiterative has quit IRC | 07:26 | |
*** reiterative has joined #zuul | 07:27 | |
openstackgerrit | Tobias Henkel proposed zuul/nodepool master: Drop python 3.5 support https://review.opendev.org/c/zuul/nodepool/+/764282 | 07:43 |
tobiash | zuul-maint: this should unbreak the nodepool gate ^ | 07:43 |
*** hashar has joined #zuul | 07:47 | |
*** jcapitao has joined #zuul | 08:05 | |
tobiash | ianw: I've commented on https://review.opendev.org/c/zuul/nodepool/+/764280, what do you think? | 08:10 |
*** saneax has quit IRC | 08:14 | |
*** Phoenikzz has joined #zuul | 08:16 | |
*** saneax has joined #zuul | 08:19 | |
*** jpena|off is now known as jpena | 08:24 | |
*** tosky has joined #zuul | 08:37 | |
*** bhavikdbavishi has quit IRC | 08:47 | |
*** bhavikdbavishi has joined #zuul | 08:51 | |
*** mach1na has quit IRC | 08:56 | |
openstackgerrit | Tobias Henkel proposed zuul/nodepool master: Drop python 3.5 support https://review.opendev.org/c/zuul/nodepool/+/764282 | 09:21 |
*** bhavikdbavishi has quit IRC | 09:21 | |
*** bhavikdbavishi has joined #zuul | 09:23 | |
zbr | apparently tox-py36 tries to run on xenian and fails due to bindep... https://zuul.opendev.org/t/zuul/builds?job_name=tox-py36&project=zuul/nodepool | 09:24 |
zbr | when we do not add jobs to periodic pipelines they can rot, last success was back in march | 09:25 |
*** bhavikdbavishi has quit IRC | 09:28 | |
tobiash | yeah, updated that (forgot to remove the nodeset override) | 10:00 |
*** mach1na has joined #zuul | 10:02 | |
*** wuchunyang has quit IRC | 10:11 | |
*** wuchunyang has joined #zuul | 10:12 | |
*** wuchunyang has quit IRC | 10:16 | |
zbr | tobiash: maybe you can help with https://review.opendev.org/c/zuul/zuul/+/763333 ? | 10:22 |
tobiash | zbr: are the existing comments addressed? | 10:24 |
zbr | yep. i home i did not miss any. | 10:25 |
zbr | it went bit different than the one i proposed on zuul-discuss. | 10:25 |
tobiash | k, I'll try to get to this today | 10:35 |
*** ashbullock has joined #zuul | 10:43 | |
*** bhavikdbavishi has joined #zuul | 11:10 | |
*** Eighth_Doctor has quit IRC | 11:34 | |
*** mordred has quit IRC | 11:34 | |
*** masterpe has quit IRC | 11:35 | |
*** zenkuro has joined #zuul | 11:43 | |
*** mordred has joined #zuul | 11:43 | |
*** bhavikdbavishi has quit IRC | 11:54 | |
*** bhavikdbavishi has joined #zuul | 11:55 | |
*** jcapitao is now known as jcapitao_lunch | 11:58 | |
*** masterpe has joined #zuul | 12:00 | |
*** Eighth_Doctor has joined #zuul | 12:00 | |
*** mach1na has quit IRC | 12:14 | |
*** bhavikdbavishi1 has joined #zuul | 12:31 | |
*** bhavikdbavishi has quit IRC | 12:33 | |
*** bhavikdbavishi1 is now known as bhavikdbavishi | 12:33 | |
*** ianychoi_ has joined #zuul | 12:35 | |
*** ianychoi has quit IRC | 12:36 | |
*** rfolco has joined #zuul | 12:37 | |
*** zenkuro has quit IRC | 12:43 | |
*** zenkuro has joined #zuul | 12:43 | |
*** jcapitao_lunch is now known as jcapitao | 12:44 | |
*** mach1na has joined #zuul | 12:57 | |
openstackgerrit | Felix Edel proposed zuul/zuul master: Switch to using zookeeper instead of gearman for jobs https://review.opendev.org/c/zuul/zuul/+/762506 | 13:09 |
openstackgerrit | Felix Edel proposed zuul/zuul master: WIP Switch to ZooKeeper backed result event queues https://review.opendev.org/c/zuul/zuul/+/764344 | 13:09 |
*** rfolco is now known as rfolco|ruck | 13:09 | |
*** ikhan has joined #zuul | 13:35 | |
*** bhavikdbavishi has quit IRC | 13:37 | |
*** Phoenikzz has quit IRC | 13:49 | |
*** zenkuro has quit IRC | 14:09 | |
*** zenkuro has joined #zuul | 14:09 | |
*** Goneri has joined #zuul | 14:13 | |
*** ikhan has quit IRC | 14:42 | |
*** frickler is now known as frickler_pto | 15:02 | |
*** ikhan has joined #zuul | 15:02 | |
*** iurygregory has quit IRC | 15:02 | |
*** ikhan has quit IRC | 15:07 | |
avass | it should be alright to add artifact information to the mqtt reporter right? | 15:13 |
avass | we technically only want metadata from artifacts produced by jobs as well :) | 15:15 |
fungi | presumably okay, i have a feeling the mqtt reporter is already leaking privileged information if it's not acl'd similarly to tenant restrictions on data availability anyway | 15:16 |
fungi | it's hard for me to reason through since the zuul i help run is 100% public so i tend to have blind spots as to what sorts of access control other deployment models might require | 15:17 |
mhu | IIRC ACL could be set broker-side for MQTT | 15:19 |
fungi | yep, mosquitto can acl specific topics/prefixes/patterns | 15:20 |
fungi | and i think we include tenant early in the topic string, so that wouldn't be hard to match on | 15:21 |
*** saneax has quit IRC | 15:41 | |
*** ashbullock has quit IRC | 15:44 | |
*** sduthil has quit IRC | 15:48 | |
*** hashar has quit IRC | 15:48 | |
*** sduthil has joined #zuul | 15:52 | |
avass | fungi: the mqtt already reports job_name, log_url, result etc so I don't think having an artifact link or the metadata provided to zuul should hurt | 16:02 |
avass | and yeah the tenant is the first part of the topic | 16:03 |
avass | oh no or at least it is for the examples | 16:04 |
avass | though maybe it would be good if it was possible to configure what should be reported by the mqtt driver | 16:07 |
*** jfoufas1 has quit IRC | 16:38 | |
*** mach1na has quit IRC | 16:58 | |
*** jcapitao has quit IRC | 17:00 | |
avass | does anyone have an idea why a github driver wouldn't work. I think I've set it up correctly and the api is receiving the webhook notifications but there doesn't seem to be anything happening in the scheduler | 17:15 |
avass | it's also able to checkout repos and load pipelines/jobs etc | 17:15 |
fungi | i think it needs branch protection enabled or it skips reading configuration in those branches | 17:16 |
avass | hmm maybe it's that | 17:16 |
fungi | i will readily admit to knowing little about the github driver aside from what's in our docs | 17:16 |
*** tosky has quit IRC | 17:17 | |
avass | exclude-unprotected-branches seem to default to false but it might be a good idea to enable it anyway | 17:19 |
*** mach1na has joined #zuul | 17:23 | |
*** rfolco|ruck has quit IRC | 17:40 | |
*** wuchunyang has joined #zuul | 17:45 | |
*** wuchunyang has quit IRC | 17:50 | |
avass | alright I fixed that but now I'm getting a traceback in the graphql: http://paste.openstack.org/show/800472/ | 17:55 |
tobiash | I think it should be enabled by default | 17:55 |
avass | not sure if that's a config error on my part or something else | 17:55 |
tobiash | avass: your github app might lack the checks api access | 17:56 |
avass | let me check | 17:56 |
avass | it has read&write for checks | 17:56 |
avass | though, I don't have any checks so could that be None and the code just doesn't check for that? | 17:57 |
tobiash | and for status? | 17:58 |
tobiash | The whole graphql call fails if a permission is wrong | 17:59 |
avass | commit status? read&write | 17:59 |
avass | just saw the logs is saying 401 so yeah | 17:59 |
tobiash | Repo administration read? (Needed to query branch protection) | 18:00 |
avass | > Read access to administration and metadata | 18:01 |
avass | > Read and write access to checks, code, commit statuses, issues, and pull requests | 18:01 |
avass | I got that | 18:01 |
avass | pretty sure I double checked the permissions that were documented but I guess I could have missed something | 18:02 |
tobiash | I also have read for org members, but that should not be needed for this call | 18:02 |
tobiash | what about pull requests read/write? | 18:03 |
tobiash | there is issues and pull requests separately | 18:04 |
tobiash | ah you mentioned that | 18:04 |
tobiash | did you change the permissions of the github app? in that case you need to accept them in all installations | 18:04 |
avass | don't think so, guess I could re-installing it | 18:05 |
avass | oh, uninstalling removes it when it's on a personal account | 18:06 |
tobiash | and I guess we should add better error handling into the graphql calls and catch and log errors there | 18:06 |
avass | and I can't see an option of accepting any new permissions | 18:07 |
tobiash | every installation owner gets an email when the permission change | 18:07 |
avass | so reinstalling it didn't work and I don't think I've gotten and email | 18:08 |
tobiash | did you restart zuul after re-installing? | 18:09 |
tobiash | if not it might have cached the old installation id | 18:09 |
avass | that didn't seem to help either :/ | 18:11 |
tobiash | still 401? | 18:11 |
avass | yep | 18:12 |
avass | but only on the graphql | 18:12 |
tobiash | that is weird | 18:12 |
tobiash | but you're using app auth right? | 18:12 |
avass | yeah | 18:12 |
avass | oh.. | 18:13 |
avass | Think I found ot | 18:13 |
avass | hmm no | 18:15 |
avass | found it | 18:16 |
tobiash | which administration did you enable? | 18:16 |
avass | the key file was misspelled | 18:16 |
tobiash | repo or org? | 18:16 |
avass | that could probably be handled better since I couldn't see any error related to that | 18:16 |
tobiash | oh :D | 18:16 |
tobiash | how did the other requests work? | 18:17 |
avass | I guess they don't need any authentication | 18:17 |
avass | that was teh only graphql one | 18:17 |
tobiash | could be | 18:17 |
avass | now I'm getting this instead: "Failed to create check run vassast/check: 422 Validation Failed" | 18:17 |
avass | heh :) | 18:18 |
tobiash | are you using file comments? | 18:18 |
avass | I'm still trying to get this set up so no :) | 18:19 |
tobiash | k, if you will, more than 50 will fail without https://review.opendev.org/c/zuul/zuul/+/762869 | 18:19 |
avass | oh | 18:21 |
avass | the executor seem to fail since I haven't set up any ssh keys | 18:22 |
avass | I guess I'll set up a dummy one since I'm only using k8s :) | 18:22 |
openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add debugging information to 422 https://review.opendev.org/c/zuul/zuul/+/764378 | 18:23 |
tobiash | you can use that to further analyze the 422 ^ | 18:23 |
tobiash | I added this to our deployment lately to debug occasional 422 we see when reporting check runs | 18:24 |
avass | maybe the executor shouldn't fail if the ssh-agent can't load an ssh-key since the connection isn't guaranteed to be ssh | 18:27 |
avass | but in practice most zuuls are probably using that the most | 18:27 |
*** jpena is now known as jpena|off | 18:28 | |
*** saneax has joined #zuul | 18:29 | |
avass | tobiash: thanks! | 18:29 |
avass | does the kubernetes driver handle private registries? | 19:17 |
*** sassyn has joined #zuul | 19:21 | |
sassyn | hi all | 19:21 |
sassyn | Good evening | 19:21 |
avass | hello :) | 19:23 |
sassyn | Quick question: Consider the following: I have 10 nodes in my node pool. There is 10 REPO configured as untrusted Repo + 1 Repo which hold all the zuul/ansible jobs/roles. | 19:23 |
sassyn | PUSH EVENTS comes to Gerrit for repo X, Y, Z, M, O, Z, X, X , M, X, X, Y, X, X, X , X, X ,X, L, O, P etc.. | 19:24 |
sassyn | I want zuul to trigger a gate job for all, but limit the repo X to take up to 3 nodepool | 19:25 |
sassyn | in parallel | 19:25 |
avass | so you want to limit the number of nodes for a repo? | 19:26 |
sassyn | yes | 19:26 |
avass | I don't think there's a direct way of doing that | 19:27 |
sassyn | hmmmm | 19:29 |
avass | you could however use different pipelines for that repo with a lower priority | 19:31 |
sassyn | but I have a general pipeline in my zuul-config trusted one | 19:38 |
sassyn | if I had a local pipeline in the project | 19:38 |
sassyn | will this overwrite the global one? | 19:38 |
avass | you can only define pipelines in a config repo, so no | 19:38 |
sassyn | hmm OK | 19:40 |
*** saneax has quit IRC | 19:44 | |
fungi | if the projects use different queues, the zuul's queuing strategy should prevent x from monopolizing the available resources by trying to enqueue equal numbers of items for the other projects first | 19:50 |
fungi | implemented by https://review.openstack.org/620954 which merged two years ago | 19:52 |
fungi | sassyn: ^ | 19:53 |
*** mach1na has quit IRC | 19:56 | |
sassyn | ping fungi | 20:03 |
sassyn | Shalom | 20:03 |
sassyn | How are you? | 20:03 |
sassyn | hope everything is good on your side | 20:03 |
*** mach1na has joined #zuul | 20:03 | |
sassyn | avass: I think I can use the Semaphore option | 20:04 |
sassyn | but maybe there is a better way | 20:04 |
*** mach1na has quit IRC | 20:08 | |
*** rfolco has joined #zuul | 20:10 | |
sassyn | I guess other option is to set in the nodeset for the job and then assign lets say 3 nodes with this label. then the gate will work up to the limit of the nodepool with this limit | 20:20 |
sassyn | but this is a waste of resources cause it might that this node will be free as there is no commit to repo X | 20:21 |
sassyn | let's see what fungi will say | 20:21 |
fungi | yes, restricting a project to its own nodeset and limiting the number of nodes available for that, or setting a semaphore for the jobs run by that pronect, are probably also viable options. zuul tries to provide fair queuing between projects but also allows projects to use all the available resources if they're the only ones requesting them, so that they don't get forced to wait for their builds to start if | 20:31 |
fungi | there are available resources not requested by any other project | 20:31 |
*** sshnaidm has quit IRC | 20:41 | |
*** rpittau is now known as rpittau|afk | 20:45 | |
*** tosky has joined #zuul | 21:09 | |
*** wuchunyang has joined #zuul | 21:47 | |
*** wuchunyang has quit IRC | 21:52 | |
*** vishalmanchanda has quit IRC | 22:21 | |
*** sassyn has quit IRC | 22:33 | |
*** sshnaidm has joined #zuul | 23:00 | |
*** sshnaidm is now known as sshnaidm|off | 23:01 | |
*** holser_ has joined #zuul | 23:18 | |
*** holser has quit IRC | 23:19 | |
*** rfolco has quit IRC | 23:32 | |
*** rfolco has joined #zuul | 23:32 | |
*** sean-k-mooney has quit IRC | 23:34 | |
*** sean-k-mooney has joined #zuul | 23:36 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!