13:00:43 <andreasn> #startmeeting Cockpit weekly meeting 2016-10-10
13:00:43 <zodbot> Meeting started Mon Oct 10 13:00:43 2016 UTC.  The chair is andreasn. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:43 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:00:43 <zodbot> The meeting name has been set to 'cockpit_weekly_meeting_2016-10-10'
13:00:48 <andreasn> .hello andreasn
13:00:49 <zodbot> andreasn: andreasn 'Andreas Nilsson' <anilsson@redhat.com>
13:00:50 <dperpeet> .hello dperpeet
13:00:52 <zodbot> dperpeet: dperpeet 'None' <dperpeet@redhat.com>
13:00:58 <mvollmer> .hello mvo
13:00:59 <zodbot> mvollmer: mvo 'Marius Vollmer' <marius.vollmer@gmail.com>
13:01:41 <andreasn> #topic agenda
13:02:15 <dperpeet> * RHEL/EPEL testing
13:02:36 <cockpitbot> 4 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-fedora-24/log.html
13:03:08 <dperpeet> * fedora 23
13:03:11 <andreasn> * subscriptions repositories and software installation
13:03:31 <dperpeet> * cockpit on ubuntu
13:04:28 <andreasn> all right, let begin
13:04:35 <andreasn> #topic #RHEL/EPEL testing
13:04:44 <andreasn> eh
13:04:53 <andreasn> #topic RHEL/EPEL testing
13:04:55 <dperpeet> :)
13:05:15 <dperpeet> some of our rhel tests currently rely on packages from epel
13:05:35 <dperpeet> and I want to make sure that our tests will pass even on systems that don't have any epel packages installed
13:05:54 <dperpeet> my plan is for those tests that need epel packages to then be skipped
13:06:12 <dperpeet> but our default path would be to keep creating test images that do use epel
13:06:27 <cockpitbot> 5 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-rhel-7/log.html
13:06:29 <dperpeet> unless there are objections
13:06:47 <dperpeet> should we add a test image without epel?
13:07:01 <dperpeet> I don't think it's necessary
13:09:28 <dperpeet> I guess there are no objections :)
13:09:31 <andreasn> no objections from me, but it's not my area :)
13:09:32 <dperpeet> end of topic
13:09:46 <andreasn> #topic subscriptions repositories and software installation
13:09:55 <andreasn> https://github.com/cockpit-project/cockpit/wiki/Feature:-Subscription-repositories
13:09:59 <andreasn> https://github.com/cockpit-project/cockpit/wiki/Feature:-OS-Package-Installation-and-Updates
13:10:08 <andreasn> added some more info to those
13:10:22 <andreasn> are these spelled out well enought that we can move on to the next stage?
13:11:08 <andreasn> I synced up the stories between the two pages
13:12:22 <cockpitbot> 1 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-centos-7/log.html
13:12:40 <dperpeet> instead of copy/pasting a workflow, why not link the other page?
13:12:58 <dperpeet> e.g. thje "Phillip" Docker workflow on the OS Package Installation page
13:12:59 <andreasn> I put in a link too
13:13:24 <dperpeet> in the notes, right?
13:13:27 <dperpeet> on the subscriptions page
13:13:31 <dperpeet> but nothing in the other direction
13:13:48 <dperpeet> do we want the combined workflow?
13:13:51 <andreasn> done
13:14:10 <dperpeet> do they have to be implemented simultaneously?
13:14:25 <andreasn> not nessesary
13:14:42 <andreasn> but they belong very tightly together
13:14:46 <dperpeet> yeah
13:14:58 <dperpeet> but those are details
13:15:03 <andreasn> software installation works on Fedora without subs handling
13:15:13 <dperpeet> I think they're good now
13:15:45 <andreasn> but subs handling without software installation feels a bit half-done
13:15:46 <dperpeet> we might phrase it a bit differently to clarify that the repo is RHEL specific, but that's pretty obvious anyway
13:15:50 <dperpeet> indeed
13:16:38 <andreasn> we should also send them around to the candlepin team
13:16:51 <andreasn> I started sketching a bit on how software installation could work
13:17:10 <dperpeet> let's discuss that outside of this general meeting scope
13:17:15 <dperpeet> I think we're going into details now
13:17:27 <cockpitbot> 1 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-continuous-atomic/log.html
13:17:48 <dperpeet> on fedora we could add fedora-testing
13:17:50 <dperpeet> for example
13:17:58 <dperpeet> to broaden the stories somewhat
13:18:03 <dperpeet> debian experimental
13:18:04 <andreasn> yeah, that's a good idea
13:18:06 <dperpeet> et cetera
13:18:24 <andreasn> ok, lets talk more about it outside the meeting
13:18:26 <dperpeet> not sure how we should scope it
13:18:40 <dperpeet> I guess in general: "extra software sources"
13:18:48 <dperpeet> or additional software sources
13:19:19 <andreasn> something along those lines
13:19:24 <andreasn> ok, next topic?
13:19:37 <dperpeet> sure
13:19:57 <andreasn> #topic Fedora 23
13:20:25 <dperpeet> during our automatic image update our tests stumbled across a regression
13:20:39 <dperpeet> sssd was updated in the last couple of weeks
13:20:42 <dperpeet> on fedora 23
13:21:05 <cockpitbot> 4 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-debian-unstable/log.html
13:21:05 <dperpeet> https://bugzilla.redhat.com/show_bug.cgi?id=1380953
13:21:15 <dperpeet> our tests didn't find this earlier because we weren't creating new images
13:21:21 <dperpeet> due to a test runner issue
13:21:47 <dperpeet> the bug itself isn't the issue - the question is: how much longer will be test fedora-23?
13:21:51 <dperpeet> *will we
13:22:27 <andreasn> when does it hit End of Life?
13:23:38 <dperpeet> ah, I guess a couple of months
13:23:43 <dperpeet> 1 month after release of f25
13:24:13 <andreasn> I think that would be a good outer limit
13:25:04 <dperpeet> ok
13:25:57 <dperpeet> end of topic
13:26:01 <andreasn> cool
13:26:08 <andreasn> #topic Cockpit on Ubuntu
13:26:27 <dperpeet> stef did some work to add ubuntu testing, which I rebased in https://github.com/cockpit-project/cockpit/pull/5094
13:26:47 <dperpeet> I have to skip the storage tests (no storaged on ubuntu), but others are welcome to look at the errors and pitch in
13:26:59 <dperpeet> it'd be great to fix some things instead of just skipping tests
13:27:07 <cockpitbot> 1 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-rhel-atomic/log.html
13:28:47 <andreasn> to add storaged to ubuntu?
13:29:10 <dperpeet> there are a bunch of other failures
13:29:34 <dperpeet> https://fedorapeople.org/groups/cockpit/logs/pull-5094-137f9983-verify-ubuntu-1604/log.html 53 failed tests
13:29:58 <dperpeet> so we still have a way to go before this is green
13:30:20 <andreasn> ah, no kube either
13:30:26 <cockpitbot> 18 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-fedora-25/log.html
13:30:47 <dperpeet> like I said, suggestions or fixes are welcome in the discussion of that pull request
13:30:50 <dperpeet> I will work them in
13:32:18 <andreasn> nice
13:32:51 <cockpitbot> 4 tests failed - http://fedorapeople.org/groups/cockpit/logs/master-7022caef-verify-fedora-atomic/log.html
13:33:46 <andreasn> eot?
13:33:52 <dperpeet> yes
13:34:40 <andreasn> #topic Open Floor
13:35:53 <andreasn> oh, right. I did some more work on kdump https://github.com/cockpit-project/cockpit/wiki/Feature:-kdump-configuration
13:36:19 <andreasn> I'm hoping to meet with fabiand soonish, since this had to be finished pretty soon
13:36:43 <andreasn> but we haven't settled on a meeting time yet
13:39:18 <dperpeet> nice to see that feature page start to take shape :)
13:39:59 <andreasn> the whole thing was a bit loose, so it took me a while to dig through documentation on what would be needed, and hopefully Fabian can help a bit with that too
13:40:13 <andreasn> I guess that's it
13:41:51 <andreasn> #endmeeting