*** jesusaur has quit IRC | 00:02 | |
*** threestrands has joined #zuul | 00:22 | |
*** jesusaur has joined #zuul | 00:23 | |
* SpamapS watching zuul deploy his almost-gone-live-RSN django app with glee | 00:33 | |
* SpamapS has a pile of kubernetes and AWS roles to upstream | 00:33 | |
ianw | tox minversion = 3.1 ... hrm that's not in fedora29 yet | 00:51 |
---|---|---|
ianw | should make it soon though ... https://bugzilla.redhat.com/show_bug.cgi?id=1652657 | 00:52 |
openstack | bugzilla.redhat.com bug 1652657 in python-tox "upgrade python-tox" [Unspecified,On_qa] - Assigned to mhroncok | 00:52 |
ianw | interesting ... "tox -e docs" doesn't seem to work with it though | 00:52 |
ianw | oh, haha helps if you run it in the right repo | 00:53 |
openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Add gearman stats reference https://review.openstack.org/620192 | 01:04 |
ianw | corvus: ^ re our prior discussion | 01:04 |
*** threestrands has quit IRC | 01:28 | |
*** threestrands has joined #zuul | 01:32 | |
*** bhavikdbavishi has joined #zuul | 02:55 | |
tobiash | ianw: cool | 05:13 |
ianw | tobiash: i've since realised the nodepool stats are a bit wonky ... change coming :) | 05:15 |
tobiash | :) | 05:16 |
*** chandan_kumar has joined #zuul | 05:26 | |
*** chandan_kumar is now known as chkumar|ruck | 05:28 | |
openstackgerrit | Merged openstack-infra/zuul master: Add support for Gerrit v2.16's change URL schema https://review.openstack.org/619533 | 05:45 |
openstackgerrit | Merged openstack-infra/nodepool master: Update node during lockNode https://review.openstack.org/616450 | 05:47 |
openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Add allowed-labels tenant setting https://review.openstack.org/617740 | 06:00 |
*** bjackman has joined #zuul | 06:05 | |
*** bhavikdbavishi has quit IRC | 06:16 | |
*** bhavikdbavishi has joined #zuul | 06:35 | |
openstackgerrit | Tobias Henkel proposed openstack-infra/nodepool master: Fix test race in test_hold_expiration_higher_than_default https://review.openstack.org/620222 | 06:44 |
tobiash | Shrews, corvus: that should fix a test race of a test case that failed several times in the stack ^ | 06:48 |
*** pwhalen has quit IRC | 06:51 | |
*** pcaruana has joined #zuul | 07:24 | |
*** threestrands has quit IRC | 07:44 | |
openstackgerrit | Merged openstack-infra/nodepool master: Add extra safety belt when reusing a node https://review.openstack.org/616465 | 07:49 |
*** themroc has joined #zuul | 08:18 | |
*** gtema has joined #zuul | 08:39 | |
*** jpena|off is now known as jpena | 08:51 | |
*** hashar has joined #zuul | 08:56 | |
*** nilashishc has joined #zuul | 10:12 | |
openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Add gearman stats reference https://review.openstack.org/620192 | 10:19 |
openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: Remove STATE_PENDING https://review.openstack.org/620284 | 10:19 |
openstackgerrit | Ian Wienand proposed openstack-infra/zuul master: [wip] Rework zuul nodepool stats reporting https://review.openstack.org/620285 | 10:19 |
*** bhavikdbavishi has quit IRC | 10:22 | |
*** bjackman_ has joined #zuul | 10:31 | |
*** bjackman has quit IRC | 10:34 | |
openstackgerrit | Sorin Sbarnea proposed openstack-infra/zuul master: Bump ansible from 2.5 to 2.6 https://review.openstack.org/620291 | 10:56 |
*** electrofelix has joined #zuul | 11:17 | |
*** bhavikdbavishi has joined #zuul | 11:19 | |
*** dkehn has quit IRC | 11:29 | |
*** chkumar|ruck is now known as chkumar|away | 11:32 | |
*** rfolco is now known as rfolco_doctor | 11:45 | |
*** bhavikdbavishi has quit IRC | 11:54 | |
*** jpena is now known as jpena|lunch | 12:36 | |
*** arxcruz is now known as arxcruz|brb | 13:00 | |
openstackgerrit | BenoƮt Bayszczak proposed openstack-infra/zuul master: add fetch_vault_secrets Ansible module https://review.openstack.org/620311 | 13:05 |
*** themroc has quit IRC | 13:10 | |
*** themroc has joined #zuul | 13:12 | |
*** rlandy has joined #zuul | 13:28 | |
*** jpena|lunch is now known as jpena | 13:38 | |
*** jhesketh_ has joined #zuul | 13:44 | |
*** dkehn has joined #zuul | 13:46 | |
*** jhesketh has quit IRC | 13:50 | |
*** bjackman_ has quit IRC | 14:20 | |
*** bjackman_ has joined #zuul | 14:22 | |
Shrews | pabelanger: you ask a good question re: metadata. I left a comment that hopefully clarifies things. And we should probably clarify in the docs the difference in instance-properties and the new field. | 14:24 |
Shrews | tobiash: why not just use cached=False there? | 14:27 |
tobiash | Shrews: oh yes, that will work too | 14:30 |
openstackgerrit | Antoine Musso proposed openstack-infra/zuul master: gerrit: use change: when querying changes https://review.openstack.org/620328 | 14:30 |
pabelanger | Shrews: ack, will look after this meeting | 14:35 |
*** mordred has joined #zuul | 14:36 | |
Shrews | tobiash: when you change that, could you take a quick scan for other tests that might be affected? | 14:40 |
Shrews | i have a feeling there's probably more than the one | 14:41 |
tobiash | Shrews: yes, but will do it tomorrow | 14:41 |
Shrews | ++ | 14:41 |
tobiash | I'm not really here today | 14:41 |
*** nilashishc has quit IRC | 14:41 | |
Shrews | lies! | 14:42 |
* mordred is also not really here today | 14:44 | |
*** rfolco_doctor is now known as rfolco | 14:53 | |
*** ianychoi_ is now known as ianychoi | 15:06 | |
*** zigo has joined #zuul | 15:20 | |
ttx | tristanC: hey, was wondering about status for the Azure nodepool driver. Kata containers got an Azure account to test Kata/Azure compatibility, and they are OK with us using it to manually test the Azure nodepool driver as needed. Did you plan to further work on that, or should I just take over ? | 15:22 |
*** bhavikdbavishi has joined #zuul | 15:31 | |
*** bjackman_ has quit IRC | 15:49 | |
mordred | ttx: I believe tristanC would be fine with you taking over work on it | 16:19 |
*** gtema has quit IRC | 16:21 | |
ttx | noted :) | 16:23 |
corvus | ttx, mordred: i'd love to have a nodepool functional test job for azure like we added for k8s. we can add the azure creds as a secret to the nodepool repo for that. | 16:24 |
corvus | i'm happy to help with that when we're ready | 16:24 |
ttx | corvus: yes that's the long-term goal... I thought I should first do the heavy lifting in making it work using manual tests | 16:25 |
clarkb | corvus: the k8s test uses a locak k8s though right? | 16:25 |
corvus | ttx: yeah, it's not going to be easy to iterate on the functional test job | 16:25 |
*** j^2 has joined #zuul | 16:25 | |
corvus | clarkb: yes, so this will be structurally different | 16:25 |
corvus | clarkb: the internal test framework can be similar, but the test setup will involve new playbooks and zuul secrets to use the real external cloud. also a new "restricted check" pipeline. | 16:26 |
corvus | (runs check jobs after code-review +2, and can use secrets) | 16:26 |
ttx | I still need to wrap my head around the concept of using nodepool to test nodepool inception | 16:31 |
ttx | One thing at a time. Like all my development work it will likely go slowly :) | 16:32 |
*** arxcruz|brb is now known as arxcruz | 16:54 | |
*** themroc has quit IRC | 16:57 | |
*** sshnaidm|afk is now known as sshnaidm | 17:08 | |
*** j^2 has quit IRC | 17:29 | |
clarkb | pabelanger: corvus Shrews trying to catch up on the executor zone changes. Is the zuul side change ready now? any reason to not merge that yet? and for the nodepool side I left some thoughts re metadata being generic | 17:48 |
SpamapS | corvus: I could probably donate a wee bit of EC2 quota for something similar for the EC2 driver. | 17:49 |
*** hashar has quit IRC | 17:49 | |
pabelanger | clarkb: yah, I'll need to also update the zuul side. | 17:51 |
pabelanger | I should be able to do that later today | 17:51 |
clarkb | pabelanger: oh right zuul side will need to look for that info in a different location? | 17:52 |
SpamapS | btw.. I'm working on putting this behind an oauth proxy, but for now, you can watch our zuul work.. https://zuul.gdmny.co/ | 17:52 |
pabelanger | clarkb: yup | 17:53 |
corvus | SpamapS: re ec2: thanks! | 17:55 |
corvus | SpamapS: also there's been a new zuul release with a much improved dashboard :) | 17:57 |
corvus | SpamapS: which is my way of saying "neat, thanks for the link!" | 17:58 |
corvus | clarkb: once i realized you were saying "key-prime" in your nodepool comment (and it was not a comment about yaml quoting) it made sense and i agree :) | 18:00 |
clarkb | corvus: oh sorry, I shouldn't realize that that was maybe ambiguous yaml | 18:00 |
corvus | well, it's not, but my brain is used to parsing yaml so that's where i went first | 18:01 |
corvus | brains are weird | 18:01 |
SpamapS | corvus: yeah, I have to choose between rebasing my patches or waiting until they land to upgrade. ;) | 18:06 |
SpamapS | oh .. hm. maybe my patches landed | 18:07 |
SpamapS | Zuul: Change has been successfully merged by Zuul (2018-11-26 15:09:23-0800) | 18:08 |
SpamapS | w000t | 18:08 |
*** bhavikdbavishi has quit IRC | 18:08 | |
*** jpena is now known as jpena|off | 18:09 | |
openstackgerrit | Clint 'SpamapS' Byrum proposed openstack-infra/zuul-jobs master: Add a markdownlint job and role https://review.openstack.org/607691 | 18:20 |
*** rlandy is now known as rlandy|brb | 18:28 | |
tobiash | Shrews: I think tomorrow I'll run tox in an endless loop and try to fix any test race there. I thought I found most of them but that doesn't seem to be the case :/ | 18:37 |
Shrews | tobiash: something does seem off :/ | 18:46 |
fungi | SpamapS: neat! your zuul doesn't look terribly busy at the moment? | 18:47 |
*** j^2 has joined #zuul | 18:57 | |
Shrews | pabelanger: Given the current meeting results, if it would help you, I could manage the nodepool side of the new metadata node attribute if you'd like. Just let me know. | 18:58 |
pabelanger | Shrews: Sure, I don't mind if you want to work on that | 18:59 |
Shrews | pabelanger: k. i'll handle it | 18:59 |
pabelanger | thank you! | 18:59 |
Shrews | np | 19:00 |
*** rlandy|brb is now known as rlandy | 19:09 | |
*** electrofelix has quit IRC | 19:25 | |
openstackgerrit | Merged openstack-infra/nodepool master: Update devstack test to Fedora 28 https://review.openstack.org/614375 | 19:44 |
*** lennyb_ has joined #zuul | 20:17 | |
*** lennyb has quit IRC | 20:20 | |
*** pcaruana has quit IRC | 21:46 | |
SpamapS | fungi: yeah I only have 5 nodes of capacity to work with at the moment and the EC2 driver does NODE_FAILURE if we try to allocate nodes too fast.. working on that. :-P | 21:48 |
fungi | sounds like a treat ;) | 22:04 |
clarkb | corvus: with currently deployed openstack zuul and the error reporting that we now have, are there still cases we need to worry about a project globally hosing the zuul config? | 22:18 |
clarkb | corvus: like maybe redefining the tempest job? and creating an ambiguous config? | 22:19 |
AJaeger_ | clarkb: we had one broken setup where a repo was branched - branched from an older git commit and then we had a list of broken jobs | 22:19 |
AJaeger_ | But not a global hosing zuul config. | 22:20 |
clarkb | AJaeger_: ya I'm wondering how broad the potential reach of such a change or merge could be | 22:20 |
AJaeger_ | clarkb: but that's in general the only situation that I am aware. | 22:20 |
clarkb | put in other words, is gating still required to protect zuul from its config | 22:20 |
dmsimard | commits are gated | 22:21 |
dmsimard | branches are not | 22:21 |
dmsimard | if we manage branches through version control, perhaps it'd be possible but it may also be too much overhead, I don't know. | 22:22 |
clarkb | dmsimard: only if you've configured your zuul to gate every commit | 22:22 |
clarkb | openstack currently has | 22:22 |
clarkb | but in general this isn't required with zuul and I'm trying to understand how badlythings can break if not done | 22:22 |
pabelanger | yah, this is my fear in ansible-network with our zuul | 22:25 |
pabelanger | humans are able to create a branch at any point in time, since they are direct push access | 22:26 |
pabelanger | so far, we haven't run into the issue yet, but I think if it happens in an untrusted project, I think we are somewhat protected in zuul | 22:26 |
dmsimard | let's all move back to svn | 22:26 |
dmsimard | no more branches | 22:26 |
corvus | clarkb: gating will protect zuul in the normal course of events; adding projects or branches (as mentioned) can break parts of the config, but damage is likely to be contained to those projects. | 22:28 |
corvus | clarkb: a snafu with tempest would be bad. but a random project won't be able to break tempest because we carefully ordered the tenant config to give devstack/tempest priority. | 22:28 |
clarkb | corvus: is zuul restartable in the general its a broken project case? | 22:28 |
clarkb | I know it wasn't in the past but I think zuul may have grown guards against that | 22:29 |
corvus | clarkb: zuul is actually *operable* in the case of a broken project | 22:29 |
corvus | clarkb: and yes, (re-)startable too | 22:29 |
clarkb | cool | 22:29 |
openstackgerrit | Merged openstack-infra/zuul master: Improve resource usage with semaphores https://review.openstack.org/619004 | 22:34 |
openstackgerrit | Merged openstack-infra/zuul master: Display ref instead of NA for time triggered items https://review.openstack.org/619252 | 22:34 |
openstackgerrit | Merged openstack-infra/zuul master: Fix manual dequeue of github items https://review.openstack.org/619272 | 22:35 |
openstackgerrit | Merged openstack-infra/zuul master: Retry queries for commits https://review.openstack.org/575137 | 22:35 |
pabelanger | in other news, thanks to the help of logan- and his limestone OSA configuration, I'm now able to CD openstack-ansible via zuul: https://github.com/ansible-network/packet-ci-cloud | 22:39 |
mordred | pabelanger: \o/ | 22:56 |
mordred | pabelanger: that's super cool | 22:56 |
pabelanger | mordred: yah, it's worked well so far | 23:00 |
clarkb | pabelanger: are you basically redeploying the limestone cloud on different hardware? | 23:01 |
clarkb | that is really neat if so | 23:01 |
pabelanger | clarkb: yup, this is a top of packet.net, but almost the same OSA config that limestone is running | 23:01 |
clarkb | nice | 23:01 |
pabelanger | I told john s, once this is up and running happy to help him also do the same for openstack nodepool. He wants to see what else, besides platform9 can be installed on packet.net | 23:03 |
pabelanger | I also want to send an ML post to openstack-discuss and see if any other operators want to help manage the cloud | 23:03 |
pabelanger | also think that would be a cool talk too | 23:03 |
mordred | ++ | 23:04 |
tristanC | ttx: the azure driver has not actually been tested, i wrote it purely based on doc and blog post. i do not plan further work on that, please go ahead and take over. | 23:26 |
*** rlandy has quit IRC | 23:27 | |
pabelanger | mordred: oh, if you have time, do you mind looking at https://review.openstack.org/618634/ and depends-on, I don't think our wheels on pypi have yarn files in them. Or, I just don't fully understand how it works | 23:32 |
pabelanger | that patch at least gets our branch-tarball job and release job working the same way | 23:33 |
*** mattclay has left #zuul | 23:50 | |
*** jhesketh_ is now known as jhesketh | 23:51 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!