15:36:07 <bookwar> #startmeeting Fedora CI SIG
15:36:07 <zodbot> Meeting started Wed Mar 25 15:36:07 2020 UTC.
15:36:07 <zodbot> This meeting is logged and archived in a public location.
15:36:07 <zodbot> The chair is bookwar. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:36:07 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:36:07 <zodbot> The meeting name has been set to 'fedora_ci_sig'
15:36:09 <jimbair> I figured it was another run-on meeting :P
15:36:19 <bookwar> #topic Hello
15:36:22 <bookwar> .hello2
15:36:23 <zodbot> bookwar: bookwar 'Aleksandra Fedorova' <alpha@bookwar.info>
15:36:23 <jimbair> .hello2
15:36:25 <zodbot> jimbair: jimbair 'Jim Bair' <jamesdbair@gmail.com>
15:36:27 <tflink> .hello2
15:36:27 <msrb> .hello2
15:36:28 <zodbot> tflink: tflink 'Tim Flink' <tflink@redhat.com>
15:36:31 <zodbot> msrb: msrb 'Michal Srb' <msrb@redhat.com>
15:36:47 <bookwar> #link https://etherpad.gnome.org/p/fedora-ci
15:37:53 <bookwar> #topic TMT pipelines
15:37:59 <msrb> yes
15:38:00 <bookwar> msrb: can you give an update?
15:38:02 <bookwar> :)
15:38:07 <msrb> so
15:38:25 <msrb> we have a prototype for rpmdeplint
15:38:37 <msrb> I am pushing what I have to https://github.com/fedora-ci
15:39:08 <msrb> we have one repository which contains rpmdeplint bits: https://github.com/fedora-ci/rpmdeplint-image
15:39:24 <msrb> having it in a separate repository means that it can be developed and tested independently, without any pipeline boilerplate, like messaging, triggering, sendind emails, etc.
15:39:47 <msrb> so that's just "containerized rpmdeplint test"
15:40:07 <bookwar> and i can run it to test my package locally?
15:40:07 <msrb> the final output of this repository is a container image, pushed to quay.io (my namespace for now, official fedora ns later)
15:40:20 <msrb> bookwar, yes
15:40:28 <bookwar> msrb: cool
15:40:39 <msrb> then there is a different repo that just references the container image built from the first repository.
15:40:52 <msrb> https://github.com/fedora-ci/rawhide-rpmdeplint-pipeline
15:41:06 <msrb> it represents the "rpmdeplint" CI pipeline. i.e. it has all the Jenkinsfile boilerplate in it.
15:41:24 <msrb> the Jenkinsfile in the repository shouldn't be that difficult to read/edit as I am pushing bits that can be reused by other pipelines to https://github.com/fedora-ci/jenkins-pipeline-library
15:41:53 <msrb> and we should be able to reuse it (the Jenkinsfile) for other similar generic tests/pipelines
15:41:59 <bookwar> one note: we do need a README file in those repos :)
15:42:06 <msrb> two main problems: there are no docs, at all
15:42:12 <msrb> bookwar, yes, exactly :)
15:42:17 <msrb> and it is still deployed in RH internal Jenkins
15:42:24 <msrb> I'd like to focus on those this+next week
15:42:44 <msrb> of course, another missing piece is testing farm
15:42:44 <bookwar> i have a question on having rawhide in the name
15:42:53 <msrb> shoot
15:43:09 <bookwar> when f33 branches do you expect to fork the repo?
15:43:18 <bookwar> to run tests for f33?
15:43:52 <msrb> bookwar, yes, that's the idea right now. that's why I'd like to keep those Jenkinsfiles as simple as possible
15:43:58 <bookwar> i was thinking that we will have rpmdeplint-pipeline and rawhide and f33 as branches there
15:44:18 <msrb> bookwar, also an option
15:44:23 <bookwar> i do like the idea of not having if's in groovy code, but i was wondering if branches maybe an easier way
15:44:27 <msrb> that could work too
15:44:43 <bookwar> depends on jenkins plugin i guess
15:44:47 <bookwar> how it creates jobs
15:44:56 <jimbair> if we do branches, then we need a README in master than explains this :)
15:44:58 <msrb> it handles branches, no problem
15:45:09 <bookwar> ok, other questions?
15:45:10 <jimbair> it also means we'll have to maintain changes across branches
15:45:11 <msrb> jimbair, yes :)
15:45:29 <jimbair> but master could either be unused or rawhide possibly? not sure.
15:45:32 <jimbair> just thinking out loud
15:45:44 <msrb> jimbair, there shouldn't be any heavy lifting going on in those jenkinsfiles
15:45:49 <mvadkert> msrb: hearing it! :)
15:46:11 <msrb> mvadkert, :P
15:46:13 <mvadkert> msrb: soonish I will unblock you
15:46:21 <bookwar> once we branch it pipelines config goes into maintenance mode the same as fedora itself. So you kind of focus on development in rawhide pipelines
15:46:31 <mvadkert> msrb: I expect early next week, we are a bit swamped
15:46:33 <msrb> mvadkert, I am not blocked, take your time :)
15:46:37 <mvadkert> msrb: review will land tmrw
15:46:42 <mvadkert> msrb: for first api draft
15:46:49 <msrb> mvadkert, thanks ;)
15:46:52 <mvadkert> msrb: just need to put it on "paper"
15:46:58 <mvadkert> msrb: no, you are blocked :D
15:46:58 <bookwar> #info rpmdeplint pipeline is available, focus on the doc
15:47:24 <mvadkert> msrb: oh, I though we are on other channel :(
15:47:34 <mvadkert> .hello2
15:47:34 <zodbot> mvadkert: mvadkert 'None' <mvadkert@redhat.com>
15:47:46 <bookwar> #info TFT backend for the tmt pipeline is in preparation, more news next time
15:47:55 <bookwar> moving to next topic?
15:48:15 <mvadkert> +1
15:48:25 <bookwar> #topic Zuul infrastructure migration
15:48:42 <bookwar> fbo: not sure if you are here, can you give us a status update?
15:49:25 <bookwar> so there was a move of the infrastructure yesterday
15:49:46 <bookwar> it is done, but we found some leftovers, fbo is fixing them
15:50:07 <bookwar> if you find something - let him now via pagure ticket as usual or here
15:50:29 <bookwar> #topic Tim's Jenkins
15:50:45 <tflink> wow, not even rpminspect :-P
15:50:51 <msrb> :D
15:50:57 <bookwar> how is it going?
15:51:01 <tflink> anyhow, the instance I had running rpminspect has pretty much died
15:51:20 <tflink> still don't understand why it won't work and I've been unable to revive it
15:51:37 <tflink> I've got something mostly working in stg but ran into a few hiccups there
15:52:06 <jimbair> tflink: I'm no expert but I can help look at it with you after my meeting following this one; even if only a second set of eyes
15:52:07 <tflink> once I'm mostly sure that's working, I plan to delete the prod non-working instance and rebuild it, starting the build numbers from about where it left off
15:52:17 <bookwar> so currently there are no rpminspect test results in rawhide?
15:52:22 <tflink> correct
15:52:22 <bookwar> we are waiting for the fix
15:52:36 <tflink> yep. I also need to write up a script to re-schedule all the missing runs
15:52:47 <bookwar> #info rpminspect tests are blocked, fix is in progress
15:53:01 <bookwar> do you need help?
15:53:11 <bookwar> if you do - jimbair is here :)
15:53:23 <tflink> I don't think so. I'm looking to put in the minimal amount of effort here as it won't live long anyways
15:53:36 <bookwar> ok, thanks for the update
15:53:42 <tflink> if it was going to stick around longer, there are things I'd prefer to do to mitigate stuff like this in the future
15:53:48 <bookwar> #topic Jim's Jenkins
15:53:55 <bookwar> you saw that coming )
15:53:56 <jimbair> lol
15:53:58 <jimbair> i did
15:53:58 <mvadkert> lol
15:54:15 <bookwar> jimbair: so how is it going?
15:54:44 <jimbair> I've mostly been held up with strange communishift issues the past two weeks; as you know, we finally got the standard template deploying after some fixes but SSO wasn't working
15:54:56 <jimbair> While doing more testing, I managed to break more things; fixed one yesterday, another remains broken but not a blocker
15:55:34 <jimbair> Reading through Red Hat docs hasn't been super helpful in trying to follow best practice, but work continues - I still need to setup a git repo that we can key from for builds
15:55:52 <jimbair> So work continues; less progress than I had hoped for sadly
15:56:24 <bookwar> so you are trying auth via OpenId now?
15:56:47 <jimbair> My current task is to get a repo setup to get openshift building against, and then start adjusting the config to use openid, yes
15:57:33 <bookwar> ok, so you just need to keep going
15:57:35 <bookwar> good
15:57:42 <jimbair> correct
15:57:45 <jimbair> still plenty to do
15:58:08 <jimbair> hopefully communishift is more stable as a result of my tickets? =)
15:58:19 <bookwar> #info work in progress, some openshift issues in the way, but no blockers
15:58:29 <bookwar> #topic Open floor
15:58:33 <tflink> until it goes down in a couple of weeks, anyways
15:58:58 <bookwar> anything else?
15:59:08 <bookwar> if not we close in a minute
15:59:10 <tflink> jimbair: I assume you're aware of the datacenter migration plans for communishift?
15:59:12 <mvadkert> I have some updates
15:59:19 <jimbair> tflink: yep :) mid April it goes down to migrate
15:59:20 <mvadkert> we have in tmt podman and testcloud provisioners
15:59:31 <bookwar> mvadkert: cool
15:59:37 <bookwar> #topic tmt update
15:59:38 <tflink> jimbair: cool, just making sure :)
15:59:46 <msrb> mvadkert, what is test cloud?
15:59:57 <bookwar> #info podman and testcloud providers are now supported
16:00:01 <mvadkert> msrb: fedora-qa's tool a simple wrapper around libvirt :)
16:00:07 <tflink> testcloud is a tool to launch cloud images on your local machine
16:00:15 <mvadkert> msrb: yeah, and we connected it to tmt
16:00:19 <msrb> mvadkert, tflink thanks ;)
16:00:28 <mvadkert> so it can now easily download nad run arbitrary qcows
16:00:37 <mvadkert> what vagrant could not do
16:00:49 <mvadkert> so it is another provisioning plugin for VMs on localhost
16:00:51 <bookwar> any link for a doc?
16:01:27 <tflink> for tmt or testcloud?
16:01:39 <bookwar> fot testcloud in tmt
16:02:03 <mvadkert> bookwar: I guess I need to update the docs for that, good point
16:02:07 <mvadkert> I have some example here: https://github.com/fedora-ci/generic-tests/blob/master/README.adoc
16:02:11 <bookwar> mvadkert: thanks
16:02:16 <mvadkert> for installability
16:02:32 <mvadkert> how to run it, the idea for any fmf goes in similar directions
16:02:36 <mvadkert> bookwar: filing an issue for docs :)
16:02:45 <bookwar> ok, we are out of time, let's close for today
16:02:45 <mvadkert> bookwar: and also for podman :)
16:02:50 <mvadkert> oki
16:02:53 <bookwar> and thanks everyone for coming
16:02:53 * mvadkert done
16:03:00 <bookwar> #endmeeting