15:34:59 <bookwar> #startmeeting Fedora CI SIG
15:34:59 <zodbot> Meeting started Wed Nov  6 15:34:59 2019 UTC.
15:34:59 <zodbot> This meeting is logged and archived in a public location.
15:34:59 <zodbot> The chair is bookwar. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:34:59 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:34:59 <zodbot> The meeting name has been set to 'fedora_ci_sig'
15:35:10 <bookwar> #topic Init
15:35:14 <bookwar> .hello2
15:35:15 <zodbot> bookwar: bookwar 'Aleksandra Fedorova' <alpha@bookwar.info>
15:35:16 <tflink> .hello2
15:35:18 <zodbot> tflink: tflink 'Tim Flink' <tflink@redhat.com>
15:36:04 <fbo> .hello2
15:36:07 <zodbot> fbo: fbo 'Fabien Boucher' <fboucher@redhat.com>
15:36:10 <bgoncalv> .hello2
15:36:11 <zodbot> bgoncalv: bgoncalv 'Bruno Goncalves' <bgoncalv@redhat.com>
15:36:21 <msrb> .hello2
15:36:21 <mvadkert> .hello2
15:36:22 <zodbot> msrb: msrb 'Michal Srb' <msrb@redhat.com>
15:36:25 <zodbot> mvadkert: mvadkert 'None' <mvadkert@redhat.com>
15:36:27 <bookwar> dcantrell: will yo join?
15:36:28 <mvadkert> woot
15:36:39 <dcantrell> I'm here
15:36:46 <bookwar> cool, let's get started
15:36:51 <dcantrell> .hello2
15:36:52 <zodbot> dcantrell: Sorry, but you don't exist
15:36:57 <dcantrell> alright, whatever
15:37:03 <bookwar> #topic Agenda doc
15:37:30 <bookwar> i was thinking, wouldn't it be nice to have a place where you can put topics before the meeting
15:37:56 <bookwar> so it is not just me who tries to bring topics but everyone can contribute
15:38:13 <tflink> dcantrell: .hello2 only works when your fas id matches your irc nick, aren't yours different?
15:38:42 <mvadkert> bookwar: sounds reasonable
15:38:49 <dcantrell> tflink: yes, they are.  my FAS account matches my RH account which was created in a time period where account names were restricted to 8 characters because of oracle
15:38:53 <bookwar> i don't know the better place for it, so i created the doc in the gnome etherpad: https://etherpad.gnome.org/p/fedora-ci#
15:39:07 <dcantrell> tflink: wanting to keep life simple, I just used that username everywhere
15:39:12 <bookwar> if you like the idea, then just start using it, any other ideas are welcome too
15:40:08 <bookwar> for today it is not filled with details, so i am just going to go through the topics randomly anyway, but for next time, let's try it
15:40:14 <mvadkert> bookwar: maybe taiga has also something, but etherpad is nice
15:40:29 <bookwar> any questions on this part?
15:40:49 <bookwar> moving on
15:40:51 <bookwar> #topic Zuul update
15:41:03 <bookwar> fbo: your turn )
15:41:29 <fbo> thanks
15:42:03 <fbo> so not so much, I have been working on adding a functional test on a package that use the zuul workflow
15:42:19 <fbo> That can be used later as a showcase
15:42:27 <fbo> let me find the link
15:42:59 <fbo> https://src.fedoraproject.org/fork/fbo/rpms/nodepool/c/57204ace65a794c122409173b60d3db320afb780
15:43:23 <fbo> So the entry point is standard like for Standart test interface of Fedora
15:43:58 <fbo> but it shows that we can use both on the Zuul CI (test with STI or w/o)
15:44:34 <bookwar> so it doesn't use the roles, only the STI interface aka tests.yaml
15:45:19 <fbo> yes it does not use the roles
15:45:40 <bookwar> anything else?
15:46:00 <mvadkert> fbo: yeah, those are basically then ansible tests ... maybe it would be nice to expose results.yml
15:46:16 <mvadkert> fbo: so it generates results in the expected format, although in fedora nothing yet consumes them
15:46:44 <fbo> mvadkert: yes we can easily add an ARA report on top
15:47:09 <mvadkert> fbo: that is also an solution :)
15:47:13 <fbo> we already do that in another CI contect on top of Zuul
15:47:21 <bookwar> fbo: it is different i think, mvadkert was mentioning the https://docs.fedoraproject.org/en-US/ci/standard-test-interface/#_results_format
15:47:36 <bookwar> it is a part of STI spec
15:48:25 <bookwar> let's move to next thing
15:48:26 <fbo> well keep in mind this is not the STI
15:48:41 <fbo> Also from my side
15:49:00 <fbo> I don't know how to move forward with https://pagure.io/fedora-infrastructure/issue/8320
15:49:19 <fbo> Kevin found that's maybe a DNS issue on src.f.o
15:49:29 <bookwar> pingou: ^ any comments?
15:50:06 <fbo> but it make the system based on Zuul mostly un-usable as it work half the time
15:50:46 <bookwar> #info https://pagure.io/fedora-infrastructure/issue/8320 still has a heavy impact on reliability of Zuul
15:51:08 <mvadkert> yeah, we should escalate that, we do not monitor much issues like this in Fedora CI pipeline bgoncalv ?
15:51:09 <fbo> so it makes me hesitate talking more about the Zuul worklow ...
15:51:23 <bookwar> fbo:i'll mention it in the meeting notes again, but we need pingou and others to comeent on that
15:51:30 <bookwar> let's move on
15:51:34 <fbo> bookwar: thanks
15:51:50 <bookwar> #topic Jenkins plugin upgrade
15:51:58 <bookwar> jbair: want to comment?
15:52:03 <bgoncalv> mvadkert: the only thing we monitor is if some PR or build that has test didn't send topics within some time
15:52:36 <jbair> oh hey =)
15:53:07 <jbair> No comments from the emails sent to devel or ci lists yesterday, so I believe we are good to perform the upgrade tomorrow around 9AM Eastern US Time
15:53:20 <bookwar> #info jenkins plugin upgrade scheduled for tomorrow, small jenkins outage is expected 9AM Eastern US time
15:53:29 <bookwar> moving on :)
15:53:44 <bookwar> #topic DevConf CfP
15:53:49 <bookwar> who submitted talks?
15:53:58 <mvadkert> testing farm has submitted a talk about packit integration
15:54:07 <mvadkert> Testing Github PRs on Fedora/Centos with Packit
15:54:36 <bookwar> i also submitted a workshop for 2 hours on all Fedora CI topics, everyone from this group is welcome
15:54:44 <bookwar> fbo: i think you also submitted one?
15:55:15 <bookwar> deadline is today, so you have couple of hours for last minute submission :)
15:55:35 <bookwar> also pingou submitted the talk on fedora gating
15:55:46 <bookwar> so we have quite a presence there
15:56:02 <bookwar> 5 minutes left, next topic :)
15:56:06 <mvadkert> bookwar: yay, I guess we should sync with pingou so we can pull off a nice demo
15:56:16 <bookwar> #topic Annocheck as a gating test
15:56:42 <bookwar> there is a submisson for Fedora change to enable annobin check for all fedora packages
15:56:52 <mvadkert> hmm, should that not be part of rpminspect?
15:56:56 <bookwar> dcantrell, tflink do you have any comments that?
15:57:15 <dcantrell> annocheck is slated for rpminspect as that's functionality that rpmdiff does in the legacy environment
15:57:27 <mvadkert> yeah, makes sense
15:57:36 <dcantrell> to me it makes sense in the scope of rpminspect because that will allow before/after build comparison as well as tracking files moves and subpackage changes
15:57:49 <tflink> it should be do-able eventually if not integrated into rpminspect but the road getting there is unclear to me at this point
15:58:05 <dcantrell> rpminspect already has a lot of ELF checks in place (which I'm working on right now too) so annocheck extending that is a good fit
15:58:20 <fbo> bookwar: yes 2 talks submitted
15:59:25 <bookwar> if i understood correctly, right now rpminspect doesn't cover everything what annocheck covers
15:59:47 <dcantrell> that's correct.  I haven't added the annocheck stuff to rpminspect yet
16:00:08 <bookwar> so we can work on extending rpminspect, but can't we run one after the other in the same pipeline already?
16:00:18 <bookwar> or create a duplicate pipeline with annocheck
16:00:43 <bookwar> just so that we are not blocked on the development of rpminspect with this feature
16:00:58 <tflink> that's certainly a possibility, yes
16:01:26 <tflink> I'm not sure that would line up with where we're trying to go in the long run but as a short term solution, it would work
16:01:40 <bookwar> dcantrell: can you share your plans regarding elf checks for rpminspect on a mailing list?
16:01:59 <dcantrell> bookwar: ok
16:02:01 <bookwar> i think we can get this conversation going, while we still wait for resultsdb-updater deployment
16:02:30 <dcantrell> adding the support to rpminspect is not a big deal, all I have to do is bump it up on the to do list in priority
16:02:37 <dcantrell> the support for annocheck
16:02:52 <bookwar> dcantrell: maybe you can get the developer of annocheck to work on it
16:03:29 <dcantrell> bookwar: I don't think that will really help because it's either I do the work or I explain how to contribute to the library and walk them through doing the same thing
16:03:49 <tflink> bookwar: yeah, I don't think that resultsdb-updater is the biggest hurdle or blocking discussion
16:04:00 <bookwar> dcantrell: ok, up to you
16:04:17 <bookwar> #topic Open floor
16:04:33 <bookwar> we past the time, but in case anyone has anything more to talk about
16:04:40 <dcantrell> bookwar: depending on where the conversation goes on the mailing list.  some people are open to contributing to rpminspect, others don't have time.
16:05:26 <mvadkert> we have enabled centos-stream testing in Packit
16:05:36 <mvadkert> we test also epel-8 against centos-stream
16:06:02 <bookwar> mvadkert: thank you, forgot about it
16:06:09 <mvadkert> I do not have a example around, but we will have some real use case shortly, the idea is to test rhel8 nightly in upstream
16:06:22 <mvadkert> also we added testing centos7
16:06:27 <mvadkert> what is quite nice, but needs more work
16:06:34 <mvadkert> some people are building also epel7
16:06:43 <mvadkert> and want feedback
16:06:57 <bookwar> #info packit runs tests on centos stream and epel now
16:07:04 <mvadkert> also what works now is "/packit test"
16:07:07 <mvadkert> to restart just the test jobs
16:07:11 <mvadkert> via a comment in the PR
16:07:24 <mvadkert> in case of infra issues and update of tests outside of the PR
16:07:27 <mvadkert> it can be handy
16:07:42 <bookwar> cool
16:08:04 <mvadkert> bookwar: any news on AWS?
16:08:27 <bookwar> i am checking the ci objective doc
16:08:32 <bookwar> and going to update it
16:08:37 <bookwar> no other news yet
16:09:12 <bookwar> ok, so let's stop here, and next time i am going to try the agenda doc, for a more organized meeting :)
16:09:23 <bookwar> please add your items
16:09:29 <bookwar> #endmeeting