14:00:47 <bstinson> #startmeeting CBS/Infra
14:00:47 <centbot> Meeting started Mon Jul 30 14:00:47 2018 UTC.  The chair is bstinson. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:47 <centbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:00:52 <bstinson> #topic Agenda
14:01:00 <bstinson> #info Topic: Status Updates
14:01:08 <bstinson> #info Topic: Account Requests
14:01:11 <bstinson> #info Topic: CI
14:01:16 <bstinson> #info Topic: Open Floor
14:01:20 <bstinson> #topic Status Updates
14:01:31 <alphacc> o/
14:01:33 <bstinson> hi all, do we have anything to report this week?
14:01:46 <mjturek> o/
14:02:05 <alphacc> I fixed the el7.centos bug in all el7.centos targets.
14:02:18 <bstinson> alphacc++
14:02:28 <alphacc> After 7.5 change to default disttag
14:04:20 <bstinson> i've started on the deps for building some various ansible streams in the configmanagment SIG. i should have something ready for posting to centos-devel this week
14:06:11 <bstinson> anything else we should mention for status updates?
14:06:42 <mjturek> I'm wondering about the status of two packages,  okay to bring up now?
14:06:44 <Arrfab> bstinson: probably the qemu-img-ev bug that lpancescu mentioned
14:07:07 <Arrfab> so waiting on a candidate/fix for it, to then be able to be back on tracks wrt vagrant images
14:07:57 <bstinson> #link https://lists.centos.org/pipermail/centos-devel/2018-July/016843.html
14:08:09 <bstinson> i don't think anything new has happened since ^
14:08:28 <bstinson> but we can check in on this next week
14:08:40 <bstinson> mjturek: sure thing, how can we help?
14:08:52 <mjturek> https://cbs.centos.org/koji/taskinfo?taskID=495307 http://cbs.centos.org/koji/buildinfo?buildID=23201 I've mentioned these in channel before bstinson wondering if there's any way to help get them published to altarch now that they use contentdir
14:09:04 <mjturek> I think you mentioned a contact
14:09:27 <bstinson> mjturek: yes! we just need someone from the storage SIG to tell us they're ready
14:10:03 <mjturek> cool! I can reach out, anyone you suggest? or just wait until storage-sig meeting
14:12:17 <bstinson> mjturek: you can wait for the next storage SIG meeting if you like. but gfidente, ndevos, and maybe ktdreyer could be good contacts for that.
14:12:39 <mjturek> perfect, I'll ask if they can weigh in, thanks bstinson
14:12:53 <bstinson> if you can provide feedback on the new centos-release files (i.e. test a few upgrades and such) that would help tremendously
14:13:36 <mjturek> I think  we could help there yeah
14:15:01 <bstinson> awesome
14:15:07 <bstinson> other things for status updates today?
14:15:34 <rakkaus_> hello, any update on this https://bugs.centos.org/view.php?id=14981 ?
14:15:55 <rakkaus_> bstinson: you wanted to start work on this 2w ago, is there any issues with this?
14:16:06 <bstinson> rakkaus_: we ran into some issues with that plugin
14:16:45 <bstinson> i'm hoping we can resolve them shortly. i need to schedule a short maintenance window for jenkins on wednesday night (eastern time)
14:16:47 <rakkaus_> bstinson: can you please comment ticket with current state and what are the issues
14:16:52 <bstinson> but we'll discuss that in the CI section
14:17:15 <bstinson> sure
14:18:47 <bstinson> #topic Account Requests
14:19:10 <bstinson> #info CI: 2 credentials reset requests processed today. No new accounts
14:19:16 <bstinson> kbsingh: anything to report from ACO?
14:21:01 <Arrfab> bstinson: that looks quiet those days, except one guy asking to be involved in all groups, but unknown by all SIGs chairs ...
14:21:31 <bstinson> cool
14:21:39 <bstinson> #topic CI
14:22:19 <bstinson> As I mentioned earlier, I need to schedule a short downtime for Jenkins and Duffy on Wednesday evening (8PM Eastern)
14:22:28 <bstinson> email to the ci-users list will be coming shortly
14:22:55 <bstinson> we need to patch a couple of plugins, and will take the opportunity to do some maintenance on the machine that hosts Duffy
14:23:22 <bstinson> assuming all goes well, we can attempt the hashicorp vault plugin again
14:24:35 <bstinson> anyone else here to talk about CI?
14:28:57 <bstinson> #topic Open Floor
14:29:03 <rbowen> If the SIG wishes to have any status report in the August newsletter, please try to get me a paragraph or two by Wednesday. Or I can dig through meeting transcripts and try to determine what's important. That doesn't always end well. :D Thanks.
14:29:05 <bstinson> what else for today?
14:30:00 <bstinson> thanks rbowen
14:32:43 <bstinson> if we don't have anything else, we'll close in 2 minutes
14:34:00 <lpancescu> wrt the vagrant images, sbonazzo just told me that he just rebuilds the qemu-kvm-rhev packages from red hat
14:34:37 <bstinson> lpancescu: let's see if we can work on a bugzilla
14:35:01 <lpancescu> i see corrupted images for libvirt both with stock qemu-img from c7 and the -ev version, i guess it's a show stopper for future vagrant releases - unless we're willing to drop libvirt
14:36:21 <TrevorH> with the stock one? I thought it only did it with the newer one
14:36:26 <lpancescu> bstinson, would it be possible to build the latest upstream qemu-img on fedora infra (perhaps COPR) and try that? we had qemu-img-ev actually work for us reliably until last month, not sure what happened
14:36:52 <bstinson> hmmm, we could do a scratch build in CBS
14:37:25 <bstinson> lpancescu: let's think about that though. do we have a short-cut to reproduce this? besides building a whole set of vagrant images?
14:37:45 <lpancescu> TrevorH, it happens with both - at least ext4 corruption with c6 for stock and c7 corruption for -ev (not sure if c7 qemu-img also corrupts the c7 xfs image)
14:38:58 <lpancescu> bstinson, we would need to keep the original image from the installation, the one that gets converted to different formats. then we could even convert the image on fedora, and replace it in the generated images
14:39:01 <bstinson> we booted the c7 image (built with the stock qemu-img) and xfs_check didn't complain
14:39:58 <lpancescu> c6 is almost dead, if c7 works and we can drop hyper-v, then we can also keep using the stock qemu-img...
14:40:58 <lpancescu> we don't know yet if the vmware images boot or not. the hyper-v images have 40GB disk images
14:40:59 <bstinson> lpancescu: if we want to try some builds of the latest qemu-img from fedora on some CI machines, or for local testing i don't mind
14:41:06 <bstinson> i'd hesitate to do that on the builders though
14:42:10 <bstinson> lpancescu: we can take this off-meeting and maybe try some things tomorrow/wednesday, if you have time?
14:42:46 <lpancescu> bstinson, sure, i have vacation this week. which time?
14:43:40 <bstinson> is tomorrow good around this time? we can chat here in case others want to follow along at home :)
14:44:03 <lpancescu> bstinson, sounds good. thanks!
14:44:09 <bstinson> awesome
14:44:32 <bstinson> i'll try to poke some things this evening to prepare
14:44:49 <lpancescu> great, thanks a lot for your help
14:45:20 <bstinson> anything else for open floor today?
14:47:40 <bstinson> going once
14:48:22 <bstinson> twice
14:49:18 <bstinson> thanks all!
14:49:21 <mjturek> thanks bstinson!
14:49:22 <bstinson> #endmeeting