| jhesketh | tristanC: Hey. Was just wondering what the status was with the web routes and dashboard work? | 00:28 |
|---|---|---|
| jhesketh | I'd like to help out and if you'd like I can try rebasing and fixing the current merge conflicts | 00:28 |
| tristanC | corvus: not sure i'll have time to write the test for https://review.openstack.org/577357 this week. | 00:29 |
| tristanC | jhesketh: Hey, we need to agree on the api schema first. In particular, all projects and jobs object needs to be list. | 00:31 |
| tristanC | Then fix the web-pages, they are currently designed for the initial implementation which only look for the first job variant and the first project pipeline configuration | 00:32 |
| tristanC | jhesketh: but i'm now conflicted, it seems like vuejs would be a much better framework to write the interface component :-) | 00:33 |
| jhesketh | tristanC: is there a place the schema is being discussed? | 00:33 |
| tristanC | jhesketh: i got most of the feedback on https://review.openstack.org/#/c/550978/ | 00:34 |
| tristanC | and https://review.openstack.org/#/c/550979 . Those are the most important to begin with | 00:34 |
| jhesketh | cool, thanks | 00:36 |
| tristanC | there is also the multi-tenant view that is broken, and i'm waiting for suggestion from http://lists.zuul-ci.org/pipermail/zuul-discuss/2018-July/000464.html | 00:40 |
| tristanC | but again, vuejs would make those things easier to implement, so it may be worthy to start a prototype and compare it with angular | 00:40 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: Add instructions on building static web components https://review.openstack.org/581256 | 01:13 |
| openstackgerrit | Merged openstack-infra/zuul master: Fix incorrect API URL for project key https://review.openstack.org/580034 | 01:37 |
| mordred | jhesketh, tristanC: I'm not really here, but I'm gonna try to get the multi-tenant view thing fixed tomorrow | 01:59 |
| *** rlandy|bbl has quit IRC | 02:30 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: timer: skip projects not using the pipeline https://review.openstack.org/581553 | 02:55 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: Add instructions on building static web components https://review.openstack.org/581256 | 02:55 |
| tristanC | pabelanger: tobiash: here is a fix to reduce the load on periodic pipelines: https://review.openstack.org/581553 | 02:55 |
| pabelanger | tristanC: great! Will look more in the morning | 03:03 |
| *** sshnaidm|bbl has quit IRC | 03:16 | |
| tobiash | tristanC: cool, will look later. The periodic pipeline is indeed a problem with many repos | 03:25 |
| tobiash | We had a user who created a 1min periodic pipeline that made zuul continuously throwing merge jobs at the mergers | 03:27 |
| *** sshnaidm|bbl has joined #zuul | 03:29 | |
| tristanC | tobiash: indeed, rdo is using a 4h periodic pipeline that enqueues more than 4000 events... it causes a lot of stress on the scheduler and stall the service for no reason | 04:12 |
| tobiash | corvus: responded on 580967, what do you think? | 04:13 |
| tobiash | tristanC: added a comment | 04:21 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul-jobs master: upload-logs: generate a script to download logs https://review.openstack.org/581204 | 04:32 |
| openstackgerrit | Ian Wienand proposed openstack-infra/zuul-jobs master: upload-logs: generate a script to download logs https://review.openstack.org/581204 | 04:50 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: timer: skip projects not using the pipeline https://review.openstack.org/581553 | 05:15 |
| tristanC | tobiash: thanks, added a test and release note | 05:15 |
| tobiash | tristanC: thanks | 05:26 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: wip: add event filter debug https://review.openstack.org/580547 | 05:27 |
| *** hwoarang has quit IRC | 05:31 | |
| *** hwoarang has joined #zuul | 05:32 | |
| *** CrayZee has joined #zuul | 05:33 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Report config errors when removing cross-repo jobs https://review.openstack.org/579997 | 05:53 |
| tobiash | corvus: rebased and fixed due to the merge of the contamination fix ^ | 05:53 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Fix zuul startup with inexisting project template and gate https://review.openstack.org/579859 | 05:54 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Tolerate missing project https://review.openstack.org/579872 | 05:55 |
| *** ianw is now known as ianw_pto | 06:00 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: timer: skip projects not using the pipeline https://review.openstack.org/581553 | 06:12 |
| *** jiapei has joined #zuul | 06:26 | |
| openstackgerrit | Merged openstack-infra/zuul master: Add zuul.child_jobs in ansible inventory file https://review.openstack.org/578181 | 06:34 |
| jiapei | Hello zuulers, I'm here again here to ask for help. I've setup zuul v3, but if fails to pull Ironic or other projects from gerrit. It complains "Host key verification failed". Some people told me that it may caused by permissions, but I've checked that the permission is right. Could anyone help me and have a look of it? The log is here http://paste.openstack.org/show/725525/, you can also login the VPS to investigate. Thank you | 06:36 |
| jiapei | in advance | 06:36 |
| openstackgerrit | Merged openstack-infra/zuul master: Support zuul.child_jobs via zuul_return https://review.openstack.org/578230 | 06:36 |
| *** gtema has joined #zuul | 07:06 | |
| tristanC | jiapei: could you try: "sudo -u zuul ssh review.openstack.org -p 29418" and accept the server key? | 07:11 |
| tristanC | "Host key verification failed." may be because zuul doesn't have the server key in its .ssh/known_host file | 07:12 |
| *** hashar has joined #zuul | 07:13 | |
| *** hashar is now known as hasharAway | 07:13 | |
| jiapei | tristanC: I su to zuul, and try "ssh -i ./.ssh/nodepool_rsa.pub review.openstack.org -p 29418", but it saied "Permission denied (publickey)." | 07:38 |
| tobiash | jiapei: that will try to use zuul as user for review.openstack.org which is probably wrong | 07:44 |
| jiapei | tobiash: yes, and then I use "ssh -i ./.ssh/nodepool_rsa -p 29418 lenovo_lxca_ci@review.openstack.org", and it says "welcome to Gerrit", but zuul executor still is error | 07:49 |
| tobiash | jiapei: what's your zuul.conf? | 07:49 |
| jiapei | tobiash: it's here http://paste.openstack.org/show/725531/ | 07:50 |
| *** fbo|off is now known as fbo | 07:53 | |
| tobiash | jiapei: what's the home dir of the zuul user? | 07:54 |
| jiapei | tobiash: /var/lib/zuul | 07:54 |
| jiapei | tobiash: I just tried "git clone ssh://lenovo_lxca_ci@review.openstack.org:29418/openstack/ironic.git", it says "Permission denied (publickey)." too, seems git can't find nodepool_rsa | 07:55 |
| tobiash | jiapei: you say the executor has the error so the scheduler can connect? | 07:56 |
| jiapei | tobiash: no, scheduler also reports error | 07:56 |
| jiapei | it reports "AttributeError: 'MergeJob' object has no attribute 'updated'" | 07:57 |
| tobiash | jiapei: can you send a complete scheduler and executor startup log? | 07:57 |
| tobiash | jiapei: that exception happens if the merger/executor fails to clone the repo | 07:58 |
| jiapei | sure, let me paste it | 07:58 |
| tobiash | jiapei: so the scheduler probably connected successfully to gerrit | 07:58 |
| jiapei | tobiash: I paste the full log here http://paste.openstack.org/show/725533/ | 08:01 |
| jiapei | tobiash: so scheduler can connect, but executor has the error, then scheduler throws exception? | 08:03 |
| tobiash | jiapei: yes | 08:04 |
| tobiash | jiapei: the executor error clearly states that it cannot verify the host keys of gerrit. This means that it's not correct in the .ssh.known_hosts | 08:05 |
| tobiash | jiapei: what's the content of .ssh/known_hosts? | 08:05 |
| tobiash | or do you have multiple restarts in that log? | 08:06 |
| tobiash | later I see 'Permission denied (publickey). which means it uses the wrong private key | 08:06 |
| jiapei | tobiash: it's here http://paste.openstack.org/show/725534/ | 08:07 |
| tobiash | jiapei: you could just use id_rsa instead id_nodepool then it's found automatically by ssh | 08:07 |
| jiapei | I guess executor can't find ./.ssh/nodepool_rsa | 08:07 |
| tobiash | jiapei: because nodepool_rsa is not a default keyfile name ssh looks for | 08:07 |
| jiapei | well, let me copy it to id_rsa then | 08:07 |
| jiapei | tobiash: it changes to a new error "git.exc.GitCommandError: Cmd('git') failed due to: exit code(-13)" , http://paste.openstack.org/show/725535/ | 08:12 |
| tobiash | jiapei: that doesn't tell me anything, now you might be able to try the manual clone again | 08:14 |
| jiapei | tobiash: I can maunual clone the code, but the log puzzle me too, there is stdout , but n stderr | 08:15 |
| jiapei | tobiash: could you try "ssh centos@46.101.221.200" with password centos75 and have a look of it? | 08:17 |
| tobiash | jiapei: sorry, but I have firewall restrictions that block outbound ssh | 08:20 |
| *** pbrobinson has quit IRC | 08:20 | |
| *** pbrobinson has joined #zuul | 08:22 | |
| jiapei | tobiash: well, the log is tricky that it doesn't show more things, do you have any suggestions on how to debug it? | 08:23 |
| tobiash | jiapei: maybe this helps: https://devopshero.wordpress.com/2013/06/19/git-trick-git_trace-1/ | 08:24 |
| tobiash | jiapei: maybe you might also have some weird selinux problems so you could try to disable this (temporarily!) | 08:25 |
| jiapei | I'll try it then | 08:26 |
| *** electrofelix has joined #zuul | 08:33 | |
| *** fbo has quit IRC | 08:33 | |
| *** ianychoi_ has joined #zuul | 09:38 | |
| *** ianychoi has quit IRC | 09:41 | |
| *** sshnaidm|bbl is now known as sshnaidm|rover | 09:45 | |
| *** pawelzny_ has joined #zuul | 09:47 | |
| *** pawelzny has quit IRC | 09:47 | |
| *** sambetts|afk is now known as sambetts | 10:00 | |
| *** ianychoi_ has quit IRC | 10:05 | |
| *** pawelzny_ has quit IRC | 10:06 | |
| *** spsurya_ has joined #zuul | 10:08 | |
| *** pawelzny has joined #zuul | 10:09 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Update README and add TESTING similar to Zuul repo https://review.openstack.org/581348 | 10:14 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Update README and add TESTING similar to Zuul repo https://review.openstack.org/581348 | 10:16 |
| *** ianychoi has joined #zuul | 10:30 | |
| *** adam_g has quit IRC | 10:49 | |
| *** adam_g has joined #zuul | 10:52 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: [WIP] Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 11:07 |
| *** EvilienM is now known as EmilienM | 11:36 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 11:42 |
| odyssey4me | Folks, I need one last bit of help to finalise https://review.openstack.org/581329 - I've added a note in https://review.openstack.org/#/c/581329/7/nodepool/tests/test_launcher.py for what's needed. Once that part's done, I'd appreciate any reviews. | 11:45 |
| *** jiapei has quit IRC | 12:06 | |
| *** rlandy has joined #zuul | 12:30 | |
| tobiash | odyssey4me: do you really want to make this an error? | 12:35 |
| tobiash | odyssey4me: if you own the cloud it may be perfectly fine hammering the cloud when hitting the quota | 12:35 |
| tobiash | odyssey4me: with 536930 it should not lead to node_failures | 12:36 |
| odyssey4me | tobiash: I don't particularly mind - could just add a note to the docs to be very careful | 12:38 |
| odyssey4me | I'm open to suggestion - just thought it'd be nice to have some sort of validation to prevent someone doing something dumb. | 12:38 |
| tobiash | odyssey4me: I'd vote for a note in the docs as this might be a valid use case for someone, but I'd also like to hear Shrews opinion on that | 12:39 |
| tobiash | fwiw hammering the cloud was nodepools default behavior in v2 (on any error) ;) | 12:42 |
| odyssey4me | tobiash: yeah, I suppose one can throttle that a bit by setting the 'rate' for the provider higher | 12:46 |
| odyssey4me | tobiash: is there a way to assert a config load failure in the tests? I'm not seeing another example of that | 12:49 |
| Shrews | i don't think we want that | 12:49 |
| tobiash | odyssey4me: the only example I found is by directly invoking the validator, but that just does schema validation | 12:49 |
| Shrews | setting no limits should be a valid thing to do | 12:50 |
| odyssey4me | Shrews: heh, in that case I don't need the extra test :p | 12:50 |
| odyssey4me | ok, I can pull that out | 12:50 |
| tobiash | odyssey4me: problem solved :) | 12:50 |
| odyssey4me | just another thing - using 'ignore_quota: false' is effectively what every other test does (it's the default value), so is there a point in re-testing that | 12:51 |
| odyssey4me | I'm thinking just testing 'true' will do. | 12:51 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 12:54 |
| Shrews | odyssey4me: you still want that. the 1st part validates that out of quota will fail as you expect (not every test does that) | 12:54 |
| Shrews | odyssey4me: i'm now confused on the intent of this flag. you want to ignore the cloud reported quota, but *still* compare max-* values against cloud reported quota | 12:57 |
| odyssey4me | Shrews: no, not against the cloud quota - but against the request coming in | 12:58 |
| Shrews | odyssey4me: how can i compare max-cores against the request coming in? | 12:58 |
| odyssey4me | so we have max-servers = 10 in the pool, when a request for a new node comes in it must validate against the pool quota, but ignore the cloud quota | 12:59 |
| odyssey4me | that's already being done in the current quota checks | 12:59 |
| odyssey4me | it does cloud first, then pool | 12:59 |
| Shrews | odyssey4me: nodepool has no idea of how many cores/ram it is using w/o asking the cloud | 12:59 |
| odyssey4me | here we just skip the cloud part | 12:59 |
| *** hasharAway is now known as hashar | 13:01 | |
| odyssey4me | it does know how many its using, because it calculates that both on start and whenever a request comes in | 13:01 |
| *** openstack has joined #zuul | 13:06 | |
| *** ChanServ sets mode: +o openstack | 13:06 | |
| Shrews | ok, i think i follow | 13:08 |
| Shrews | maybe this would better named ignore_provider_quota | 13:09 |
| odyssey4me | yeah, I was wondering if the name was clear enough | 13:09 |
| odyssey4me | ignore_provider_quota, or ignore_Cloud_quota | 13:09 |
| odyssey4me | I think I like 'provider' better | 13:10 |
| Shrews | left comments | 13:13 |
| mordred | gtema: crap. I was just going over openstacksdk patches and realized that https://review.openstack.org/#/c/580475 hasn't landed yet (which we need for your shade-openstacksdk patch) | 13:14 |
| mordred | gtema: but also, dtantsur made a good point in review there about making the parameter name task_manager instead of manager to be internally consistent | 13:14 |
| gtema | mordred: ok, thanks | 13:16 |
| gtema | so you mean renaming manager to task_manager? | 13:16 |
| mordred | gtema: I'm gonna get the sdk patch landed as soon as it passes tests - and then get another release cut | 13:17 |
| gtema | ok, great | 13:17 |
| mordred | gtema: and yeah - the argument name to the constructor will be task_manager | 13:17 |
| gtema | ok, I will then wait for the release and update the change | 13:17 |
| mordred | Shrews: if you're bored, if we get https://review.openstack.org/#/c/580475 in we can hopefully get those nodepool/sdk patches landed | 13:17 |
| mordred | gtema: ++ | 13:17 |
| *** spsurya_ has quit IRC | 13:18 | |
| odyssey4me | man, I thought I was going crazy because my tests weren't being executed... then I realised that stestr is outputting the top 10 slowest and not all tests | 13:20 |
| tobiash | did anyone already tried this: https://sweetness.hmmz.org/2018-07-10-mitogen-released.html ? | 13:21 |
| tobiash | if that works it could dramatically reduce our per task overhead in every job | 13:21 |
| *** elyezer has quit IRC | 13:22 | |
| mordred | tobiash: I haven't played with it yet - but it does seem worth at least investigating | 13:24 |
| tobiash | mordred: one of the major pain points I hear from users is that it takes quite long until the build really starts (aka run playbook) while the actual work of the pre playbooks is not that much but many tasks that each costs a second or two | 13:26 |
| mordred | tobiash: yeah - that's a fair complaint | 13:27 |
| tobiash | mordred: it seems to be implemented as a strategy so yay that we didn't patch/fork the strategy plugins | 13:28 |
| mordred | tobiash: ++ | 13:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 13:37 |
| odyssey4me | Shrews ^ updated, thanks for the feedback | 13:37 |
| mordred | tobiash: ooh! https://mitogen.readthedocs.io/en/stable/#module-forwarder | 13:44 |
| mordred | tobiash: that's super cool (although I don't know how much it helps us specifically for zuul) | 13:44 |
| mordred | but it seems like a glorious answer to the typical ansible module bootstrap issue of remote nodes not having the right python libraries installed | 13:44 |
| tobiash | wow that's crazy | 13:45 |
| rcarrillocruz | you folks should chat with jimi|ansible | 13:49 |
| rcarrillocruz | he's been chatting with dw about mitogen | 13:49 |
| rcarrillocruz | about how to maybe getting it into ansible at some point, not sure how far that is | 13:49 |
| Shrews | odyssey4me: you missed the release note i requested | 13:49 |
| odyssey4me | Shrews: oops, will add one quickly | 13:50 |
| *** elyezer has joined #zuul | 13:50 | |
| Shrews | tobiash: since you're the quota expert, could you review odyssey4me's change when you get a chance? Especially its interaction with your aborted node change | 13:52 |
| tobiash | Shrews: it's on my review list for this evening | 13:53 |
| Shrews | tobiash: thx! | 13:53 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: Add command socket handler for full reconfiguration https://review.openstack.org/581754 | 13:55 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 13:57 |
| mordred | jimi|ansible: apparently we think the mitogen stuff is potentially interesting :) | 13:58 |
| tobiash | mordred: shit the mitogen docs say py26, py27 and py36 are supported... | 14:06 |
| mordred | tobiash: but no py35? | 14:06 |
| tobiash | mordred: so apparently no py35 | 14:06 |
| mordred | well - I'd like to shift to py36 anyway ;) | 14:07 |
| tobiash | I'm already on py36 ;) | 14:07 |
| tobiash | so maybe I can try this out | 14:07 |
| tobiash | if this works we maybe could add that as an optional technology preview with some exclamation marks in the docs? | 14:08 |
| clarkb | the difference if supporting 27 and 36 for 35 should be nil right? | 14:09 |
| clarkb | I don't know that you can write py36 code that is compatible with 27 that won't run on 35 | 14:09 |
| clarkb | I guess you could have split code bases | 14:10 |
| tobiash | I have no idea, maybe they just don't support that official because it's untested on py35 | 14:10 |
| mordred | clarkb: yeah - I'm guessing it's just about being untested | 14:10 |
| mordred | I agre - the chances you can write code that works with 36 and 27 and not 35 ... I'd be very impressed | 14:10 |
| mordred | tobiash: I'll definitely be interested to hear about your experience if you try it - because it actually significantly reduces execution overhead / time but still works properly- that could add up to some pretty large savings for us in openstack | 14:12 |
| mordred | my biggest concern would be possibility for behavioral changes - but it seems like that's a thing the author also wants to avoid | 14:13 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: Add a dequeue command to zuul client https://review.openstack.org/95035 | 14:13 |
| tobiash | mordred: I cannot promise that I can do that right now though | 14:13 |
| mordred | tobiash: oh - certainly not :) | 14:14 |
| mordred | tobiash: just saying -if you did, I'd love to hear the results | 14:14 |
| mordred | this: https://mitogen.readthedocs.io/en/stable/#logging-forwarder <-- this I've wanted in ansible for ages - and is related to how I wanted to rework log streaming | 14:15 |
| tobiash | mordred: I'm pretty sure won't keep that secret when I get a chance to try it :) | 14:15 |
| mordred | :) | 14:15 |
| tobiash | mordred: I'll leave the logging forwarder to you as the logging expert ;) | 14:16 |
| mordred | uhoh. we're all in trouble | 14:16 |
| jimi|ansible | mordred: yeah we're hoping to get some kind of tech preview on that going | 14:26 |
| *** spsurya_ has joined #zuul | 14:37 | |
| *** acozine1 has joined #zuul | 14:54 | |
| *** acozine1 has left #zuul | 15:02 | |
| *** acozine1 has joined #zuul | 15:02 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: Add a dequeue command to zuul client https://review.openstack.org/95035 | 15:29 |
| *** fbo has joined #zuul | 15:31 | |
| mordred | tristanC, tobiash: I believe I have found the appropriate angular constructs to deal with our routing fun | 15:34 |
| mordred | tristanC, tobiash: https://angular.io/guide/router#lazy-loading-route-configuration | 15:34 |
| mordred | so I will make a module (that being a module can have the ZuulService injected to it) - that will query the api/info endpoint for the information and the provide the correct routing table to the main router | 15:36 |
| tobiash | mordred: cool :) | 15:38 |
| *** nchakrab has joined #zuul | 15:52 | |
| *** dmsimard has quit IRC | 16:04 | |
| tobiash | odyssey4me: I added a couple of comments to 581329 | 16:05 |
| *** dmsimard has joined #zuul | 16:05 | |
| tobiash | odyssey4me: some are nits but I think I found a serious issue in there | 16:05 |
| odyssey4me | orly? thanks - taking a look now | 16:06 |
| tobiash | odyssey4me: (but easy to fix) | 16:06 |
| pabelanger | I hate to ask again, but where are we standing on the next release of zuul? IIUC we had an issue with broken configuration being able to merge, which might be preventing a release | 16:15 |
| odyssey4me | ok, let me add the test cases first to validate because it seemed to work for me as expected, so I'm not sure if the concern in hasProviderQuota is right, unless I'm misunderstanding it | 16:16 |
| odyssey4me | oh, I think I see it now, nvm | 16:17 |
| tobiash | pabelanger: we have some things to sort out first | 16:21 |
| tobiash | pabelanger: the web routing | 16:21 |
| tobiash | pabelanger: and I think we should include https://review.openstack.org/577357 | 16:24 |
| tobiash | as this breaks working with unprotected branches (and thus also every github zuul-from scratch user as zfs doesn't make use branch protection) | 16:25 |
| pabelanger | tobiash: what are your thoughts about maybe tracking the blockers some place for n+1 of zuul? I feel bad always ping to see the release criteria. we also want to make sure what ever hash we think is good to schedule restarts in openstack too | 16:27 |
| tobiash | pabelanger: and we need https://review.openstack.org/579997 (which needs a non-me and non-corvus vote) | 16:27 |
| pabelanger | this is mostly to help point downstream people to something to help show what needs to be done | 16:27 |
| pabelanger | ack | 16:27 |
| openstackgerrit | Fabien Boucher proposed openstack-infra/zuul master: zuul-web: key endpoint: return 404 when tenant or project not found https://review.openstack.org/581793 | 16:29 |
| tobiash | pabelanger: maybe we could make use of storyboard for this? corvus what do you think? | 16:29 |
| tobiash | pabelanger: and it's really hard to track the current release blockers in everyones head, so yes I also think we need to track that | 16:30 |
| tobiash | at least it's hard to track these in my head, I cannot speak for the others ;) | 16:31 |
| pabelanger | agree, as much as I don't like storyboard, I did find it a useful to look at what we needed to get a 3.0 release of zuul done. | 16:31 |
| *** nchakrab has quit IRC | 16:39 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 16:46 |
| tobiash | odyssey4me: I've responded | 17:00 |
| tobiash | odyssey4me: sorry, I had wrong thoughts about the test case and added further info there | 17:00 |
| odyssey4me | tobiash: no worries, thanks for taking the time! | 17:02 |
| odyssey4me | ah, I see - that makes more sense now | 17:03 |
| odyssey4me | let me add that test case too :) | 17:03 |
| tobiash | :) | 17:03 |
| tobiash | yeah that quota stuff is non-trivial. It took me quite some time to get that into nodepool initially. | 17:05 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul master: github: prevent AttributeError when missing branch protection https://review.openstack.org/577357 | 17:08 |
| tobiash | tristanC, corvus: I think this is an important bugfix we should include in the next release so I added test coverate to that ^ | 17:09 |
| tobiash | pabelanger: ^ | 17:09 |
| *** sshnaidm|rover is now known as sshnaidm|bbl | 17:12 | |
| tobiash | s/coverate/coverage | 17:18 |
| *** sambetts is now known as sambetts|afk | 17:29 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: Add a dequeue command to zuul client https://review.openstack.org/95035 | 17:30 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 17:31 |
| odyssey4me | tobiash ^ extra test added, and seems to pass now :) | 17:31 |
| *** electrofelix has quit IRC | 17:35 | |
| tobiash | odyssey4me: one last comment in the release note | 17:37 |
| tobiash | odyssey4me: code lgtm now :) | 17:37 |
| odyssey4me | haha, my editor keeps messing with me | 17:37 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack-infra/nodepool master: Add ability to ignore provider quota for a pool https://review.openstack.org/581329 | 17:37 |
| odyssey4me | tobiash thanks again for all your help :) time for me to conclude my day, chat tomorrow again! | 17:39 |
| tobiash | odyssey4me: you have my +2 now and have a nice evening | 17:40 |
| *** sshnaidm|bbl is now known as sshnaidm|rover | 17:45 | |
| openstackgerrit | Fabien Boucher proposed openstack-infra/zuul master: zuul-web: jobs list endpoint: return 404 when tenant not found https://review.openstack.org/581810 | 18:09 |
| *** gtema has quit IRC | 18:13 | |
| *** sshnaidm|rover is now known as sshnaidm|off | 18:19 | |
| corvus | pabelanger: i'd be happy for you to track release blockers and will be happy to provide you that information | 18:26 |
| corvus | tobiash, tristanC: i never saw an answer to my question on 577357; do you know the answer? | 18:26 |
| tobiash | corvus: github returns 404 if no branch protection is enabled on a branch | 18:27 |
| tobiash | corvus: so without this fix gating will be impossible on a branch without enabled branch protection | 18:28 |
| corvus | tobiash: got it, thanks. lgtm. | 18:28 |
| tobiash | corvus: that method is used by the the required status check when entering the gate | 18:28 |
| pabelanger | corvus: ack, I'll work on setting up boartty again this afternoon | 18:40 |
| pabelanger | for zuul, would it be sane to have a pipeline, that is supercedent for both post and periodic job? In an effort to have that is always running, if that make sense | 18:51 |
| pabelanger | the idea is the job is an artifact publisher, but run really long | 18:52 |
| corvus | pabelanger: worth a try :) | 18:52 |
| pabelanger | yah, that is what I am thinking too | 18:52 |
| pabelanger | periodic timer could run ever x hours, and if post already enqueued it, should noop | 18:53 |
| pabelanger | and prevent jobs from backing up | 18:53 |
| pabelanger | corvus: I'll hold off until https://review.openstack.org/581553/ lands, to make mergers happy | 18:54 |
| *** harlowja has joined #zuul | 19:00 | |
| *** CrayZee has quit IRC | 19:12 | |
| Shrews | mordred: looking at pbrx code keeps making me think about dockstack | 19:39 |
| corvus | tobiash, tristanC: jhesketh suggested a test in https://review.openstack.org/577357 think that's possible? | 20:30 |
| corvus | clarkb: did you want to take another look at https://review.openstack.org/579997 or should i +3? | 20:31 |
| tobiash | corvus: PS2 adds a test | 20:31 |
| corvus | tobiash: ah, there's an existing test that hits that condition? | 20:32 |
| tobiash | corvus: I changed the fake github to behave like the real github and that broke around 13 test cases without the fix | 20:32 |
| corvus | tobiash: perfect! thanks :) | 20:32 |
| corvus | tobiash: i think you should re-add your +2 and merge it :) | 20:33 |
| tobiash | corvus: actually it failed exactly 14 tests :) | 20:33 |
| *** acozine1 has quit IRC | 20:33 | |
| tobiash | corvus: can do, I just didn't at first because the test change was from me | 20:33 |
| corvus | tobiash: yeah. i think it's sufficiently reviewed | 20:34 |
| clarkb | corvus: I think you can go for it. not in a spot for code review | 20:34 |
| tobiash | done | 20:34 |
| tobiash | corvus: did you see my response on 580967? | 20:36 |
| corvus | clarkb: ack. i think i got your comments from earlier reviews. | 20:36 |
| *** spsurya_ has quit IRC | 20:36 | |
| tobiash | corvus: we can change it to your suggestion easily but I think we should at least think about that eventual possibility of a race | 20:37 |
| tobiash | I don't have a precise example of a possible race in mind but had the feeling to better be safe than sorry | 20:39 |
| tobiash | but that might also be not a problem | 20:40 |
| corvus | tobiash: when the reconfiguration events are merged, we still keep track of which things to clear, so i don't think that should be a problem. the current code only adds a tenant reconfiguration event from within the main scheduler loop, so it can't race right now. perhaps something in the future (maybe a tenant reconfig event submitted by rpc) could add the event from outside of that loop, however, i think | 20:41 |
| corvus | the queue has locking to avoid a race there. | 20:41 |
| tobiash | corvus: ok, then we will simplify it | 20:42 |
| corvus | it's a good question; i assumed the thought was something like that, but i worry that it may have the opposite effect if we split the logic in two places. at least the other way, there's one place to look at and think about this :) | 20:42 |
| corvus | (responded in review for the record) | 20:44 |
| tobiash | corvus: thanks | 20:45 |
| tobiash | corvus: btw, that change saved us 150 tenant reconfigurations in one day :) | 20:47 |
| tobiash | so much less user visible service stalls | 20:47 |
| corvus | tobiash: ++ we should be able to make this all really efficient :) | 20:48 |
| tobiash | corvus: the next thing we will work on is branch caching in the github driver and the reconfiguration without cleaning the cache | 20:49 |
| tobiash | that will also improve things | 20:49 |
| openstackgerrit | Merged openstack-infra/zuul master: github: prevent AttributeError when missing branch protection https://review.openstack.org/577357 | 20:55 |
| openstackgerrit | Merged openstack-infra/zuul master: Report config errors when removing cross-repo jobs https://review.openstack.org/579997 | 20:57 |
| *** hashar has quit IRC | 21:04 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Fix broken templates for projects in dependent pipelines https://review.openstack.org/581888 | 21:21 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Report to gerrit over HTTP https://review.openstack.org/577027 | 21:51 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Support line comments in Gerrit https://review.openstack.org/577035 | 21:51 |
| *** harlowja has quit IRC | 23:05 | |
| *** rlandy has quit IRC | 23:10 | |
| corvus | zuul's first line comment! https://review-dev.openstack.org/#/c/107988/16/.zuul.yaml | 23:14 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Report to gerrit over HTTP https://review.openstack.org/577027 | 23:20 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Support line comments in Gerrit https://review.openstack.org/577035 | 23:20 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Report config errors as line comments https://review.openstack.org/581908 | 23:20 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Fix line numbers in job repr https://review.openstack.org/581909 | 23:20 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Fix line numbers in job repr https://review.openstack.org/581909 | 23:21 |
| logan- | fwiw I just found that gerrit connection baseurl=https://review.openstack.org/ does not work, while baseurl=https://review.openstack.org does. i wonder if an rstrip should be added somewhere for user friendliness | 23:43 |
| mordred | corvus: that's awesome! (line comments) | 23:44 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!