fedora-meeting-1
LOGS
21:00:15 <pwhalen> #startmeeting Fedora ARM weekly status meeting
21:00:15 <zodbot> Meeting started Wed Feb 20 21:00:15 2013 UTC.  The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:15 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
21:00:15 <pwhalen> #chair pwhalen jonmasters bconoboy ctyler pbrobinson dgilmore
21:00:16 <zodbot> Current chairs: bconoboy ctyler dgilmore jonmasters pbrobinson pwhalen
21:00:20 <jonmasters> .fas jonmasters
21:00:20 <zodbot> jonmasters: jcm 'Jon Masters' <jonathan@jonmasters.org>
21:00:25 <pwhalen> .fas pwhalen
21:00:26 <zodbot> pwhalen: pwhalen 'Paul Whalen' <pwhalen@redhat.com>
21:00:37 <ctyler> .fas chris@tylers.info
21:00:37 <zodbot> ctyler: ctyler 'Chris Tyler' <chris@tylers.info>
21:00:46 <pwhalen> good afternoon everyone
21:01:06 <bconoboy> .fas blc@
21:01:06 <zodbot> bconoboy: blc '' <blc@redhat.com>
21:01:52 <jonmasters> good afternoon
21:01:58 * j_dulaney waves
21:02:00 <pwhalen> #topic 0) Status of ACTION items from our previous meeting
21:02:16 <pwhalen> #info ONGOING - pwhalen, jonmasters, pbrobinson to work on kernel update compatibility coverage
21:02:28 <pwhalen> #info COMPLETE - dgilmore, nirik, and pbrobinson to coordinate koji move
21:02:31 <ahs3> .fas ahs3
21:02:31 <zodbot> ahs3: ahs3 'Al Stone' <ahs3@redhat.com>
21:02:41 * nirik looks up
21:02:43 <pwhalen> #info INPROGRESS - pbrobinson working on eclipse issue
21:03:22 <bconoboy> anything else?
21:03:25 <oatley> .fas oatley
21:03:25 <zodbot> oatley: oatley '' <acoatley-willis@myseneca.ca>
21:03:30 <pwhalen> I think thats all we had from last week, the Secret Decoder doc is also underway, I'll send an email to the list once its been fleshed out a little more
21:03:41 <pwhalen> but those were the action items listed
21:04:16 <ctyler> pwhalen: Can we see/contribute to the WIP on secret decoder?
21:04:39 <pwhalen> ctyler, absolutely, I'll email the list asking for exactly that
21:04:41 <jcapik> .fas jcapik
21:04:41 <zodbot> jcapik: jcapik 'Jaromír Cápík' <jcapik@redhat.com>
21:04:53 <jcapik> hello
21:05:11 <pwhalen> #link last weeks meeting minutes - http://meetbot.fedoraproject.org/fedora-meeting-1/2013-02-13/fedora-meeting-1.2013-02-13-21.00.html
21:05:40 <pwhalen> shall we continue on to item #1?
21:05:45 <dmarlin> .fas dmarlin
21:05:45 <zodbot> dmarlin: dmarlin 'David A. Marlin' <dmarlin@redhat.com>
21:06:13 <pwhalen> #topic 1) Problem Packages
21:06:55 <pwhalen> from Peter's email:
21:06:58 <pwhalen> eclipse - being actively worked on, we think we're close
21:06:58 <pwhalen> java-1.7.0-openjd - being worked on, not sure of latest status
21:06:58 <pwhalen> kernel - ICE in tegra kernel. gcc build incoming to fix
21:06:58 <pwhalen> python - build in progress that should close the bug out
21:07:45 <pwhalen> do we have any others?
21:07:47 <bconoboy> #info eclipse, java-1.7.0openjdk, kernel, python, gcc problems are nearly resolved
21:07:50 <bconoboy> llvm 3.2?
21:08:01 <jonmasters> I'm working on the kernel
21:08:09 <jonmasters> I've a fix for mongodb that is with Peter
21:08:15 <pwhalen> bconoboy, looks built
21:08:17 <jonmasters> I have agreed to look at llvm when I get free time
21:08:23 <pwhalen> http://arm.koji.fedoraproject.org/koji/buildinfo?buildID=116111
21:08:26 <bconoboy> #info jonmasters has provided a fix for mongodb
21:08:28 <jonmasters> it'll build but it needs some help still
21:08:32 <jonmasters> (llvm)
21:08:49 <bconoboy> jonmasters: have you re-evaluated since 3.2 was released?
21:09:11 <jonmasters> bconoboy: I do not believe that they rewrote the triplet handling
21:09:13 <jonmasters> verifying
21:09:21 <bconoboy> #info llvm 3.2 is out, still has issues with hard float abi
21:09:40 <jonmasters> answer in 30 secs
21:09:51 <jonmasters> (or the next one's free)
21:10:00 <bconoboy> Also, grubby
21:10:31 <bconoboy> #link Grubby segfaults when used with --default-kernel https://bugzilla.redhat.com/show_bug.cgi?id=912873
21:11:04 <jonmasters> nope still as braindead as before
21:11:14 <jonmasters> bconoboy: it's not fixed, might happen to work, but it's not fixed
21:11:26 <bconoboy> jonmasters: are you going to look at it or report it upstream?  I think we need upstream support here.
21:11:39 <jonmasters> I agree. I have asked a few folks about upstream development on it
21:11:51 <jonmasters> I'll directly ping upstream but I expect to be told where to go
21:11:51 <bconoboy> any sort of trackable issue # for it?
21:12:02 <jonmasters> as in "it works on Ubuntu, what are you crazies doing different?"
21:12:33 <bconoboy> #action jonmasters to file bug report with llvm upstream community
21:12:36 <jonmasters> sadly not at this stage. I asked Linaro also if they had resources to look at it, but they don't
21:12:36 <bconoboy> good?
21:12:38 <jonmasters> ok
21:13:04 <bconoboy> those are all the packages I am aware of having issues
21:13:06 <jonmasters> upstream is ARM basically, so unless I twist ARMs it means I still get to fix it
21:13:19 <jonmasters> (so we can file issues until we're blue in the face, but it won't change reality :) )
21:13:50 <pwhalen> #topic 2) ARM Koji moved to Phoenix - Remaining changes, any issues encountered?
21:14:01 <jonmasters> biggest feedback from me is the kernel
21:14:16 <jonmasters> we discussed this and kernel upgrades won't be done without my signoff, and that of Peter
21:14:26 <bconoboy> dgilmore fixed the koji web interface login issue yesterday which was the only issue I am aware of
21:14:32 <jonmasters> had PHX gone to 3.7 it would have stopped mock working, as an example
21:14:43 <nirik> yeah, we can be careful of updates.
21:15:06 <jonmasters> nirik: you expertly have the ball, I'll provide some technical assistance, it'll be awesome
21:15:14 <nirik> :)
21:15:21 <pwhalen> what about koji-builder? will the hub also be updated to keep them in sync?
21:15:21 <bconoboy> anybody else seeing any issues with new koji system?  Seems zippy to me
21:15:27 <jonmasters> seems awesome
21:15:30 <pwhalen> seems great
21:15:33 <jonmasters> pwhalen: you had mentioned some buildroot issues?
21:15:34 <nirik> I think possibly trying to do them at the same time as we do primary... then we have scheduled outage windows, etc
21:15:48 * jonmasters has only seen errors so far where it concerns missing deps, not busted buildroots
21:16:02 <pwhalen> I talked with Peter about that early, he says thats a shadow issue looking suspiciously similar to the repo issue we had
21:16:03 <jonmasters> nirik: I know you have a different mount setup and only one repo generation host
21:16:22 <jonmasters> nirik: but you increased the RAM on the VM doing that and it seems ok, right?
21:16:37 <nirik> yeah, currently the hub is doing all the newrepos... because builders can't get to the koji store.
21:16:46 <nirik> it's been fine since we gave it a bunch of memory tho.
21:16:55 <jonmasters> cool, indeed, just sharing for the logs
21:17:09 <nirik> on that front: there's a change request in, and hopefully friday we will have net for the other boxes.
21:17:10 <bconoboy> is the buildroot expiration issue fixed?
21:17:12 * jonmasters reads everything, I just don't always say stuff on #fedora-arm ;)
21:17:53 <jonmasters> nirik: you rock
21:18:03 <nirik> bconoboy: not sure. I'm not sure where to adjust it. ;(
21:18:07 <jonmasters> #info jonmasters to buy nirik a lot of his favorite beverage :)
21:18:13 * masta looks in late
21:18:14 <nirik> we should probibly ping dgilmore on that...
21:18:18 <masta> .fas parasense
21:18:18 <zodbot> masta: parasense 'Jon Disnard' <jdisnard@gmail.com>
21:18:31 <nirik> happy to help out. ;)
21:19:09 <bconoboy> dgilmore: yt?
21:19:32 <nirik> so, I think with the new net we just bring up the other machines on it, get them all set and add them in, only once they are happy do we move the current ones to the new net.
21:19:36 <bconoboy> #info There may still be an issue with buildroots expiring almost immediately after builds are finished- requires confirmation
21:19:37 <pwhalen> he mentioned he doubted he would be here (in Brno)
21:20:29 <jonmasters> nirik: on the SSL issue?
21:20:35 <jonmasters> nirik: is that a known issue?
21:20:41 <nirik> jonmasters: I think dgilmore fixed that... but I could be wrong.
21:20:43 <jonmasters> ok
21:20:56 <nirik> if folks could test that would be good.
21:20:57 <jonmasters> confirmed, he did
21:21:01 <jonmasters> I jsut chkecked
21:21:04 <bconoboy> ssl works for me
21:21:09 <bconoboy> login also works for me
21:21:12 <nirik> it was a ssl config that wasn't right from the move I think
21:21:22 <jonmasters> yea, it looked to be redirecting weirdly
21:21:46 <jonmasters> nirik: on builder ratios
21:21:55 <jonmasters> nirik: I think we need to ponder the 20/4 split
21:22:09 <jonmasters> nirik: when the other boxes come up, will we have all of them on Friday?
21:22:10 <nirik> yeah, happy to adjust. we could also do more v5's in the next chassis.
21:22:17 <jonmasters> nirik: if so, what is your thought/plan
21:22:18 <jonmasters> nirik: ?
21:22:35 <nirik> yeah, they will need to be installed, but that should be pretty quick.
21:22:36 <pwhalen> have we seen tasks waiting on v5?
21:22:54 <nirik> there seemed like a backlog of v5 stuff at first...
21:23:00 <nirik> but it seems now like it's pretty quiet.
21:23:19 <jonmasters> I'm happy with a small number of v5 builders
21:23:25 <jonmasters> but more than 4 in the end :)
21:23:36 <jonmasters> nirik: you'll provision all of the other chassis?
21:23:41 <jonmasters> nirik: so we'll have all 4 up?
21:23:42 <nirik> if we do 20/4 again for the next chassis then there would be 8.
21:23:51 <jonmasters> nirik: my only suggestion is of course to stripe things
21:23:53 <jonmasters> nirik: sure
21:23:54 <nirik> yeah. Did you folks want them all as builders?
21:24:03 <jonmasters> nirik: I'm assuming we'll want some for QE
21:24:06 <nirik> I thought some were hopefully going to be qa or the like.
21:24:08 <nirik> yeah
21:24:23 <jonmasters> nirik: and for other tasks, but hmmm...let's make them all builders for now to stress test then pull some off later?
21:24:34 <jonmasters> nirik: might as well put them to use until we have the QE stuff setup?
21:24:43 <jonmasters> nirik: after all, you can tear them down pretty quick, right?
21:25:01 <nirik> sure.
21:25:25 <nirik> ideally I'd love to have a few to run infrastructure stuff on. ;) Real world testing.
21:25:34 <nirik> but yeah, we can reallocate as we go
21:26:54 <pwhalen> anything on the move?
21:27:51 <pwhalen> #topic 3) Mass rebuild - Status update
21:28:14 <pwhalen> from Peter's email:
21:28:17 <pwhalen> We're about 1/3 of the way through with around 2750 build up to this morning UK time. koji-shadow has been a little temperamental but I believe I've got on the top of this and it should start to pick up speed for the final 2/3rds or so.
21:28:27 <nirik> primary finished yesterday.
21:28:49 <nirik> sounds like things are cranking along. ;)
21:28:50 <jonmasters> we need also to ponder the aarch64 patching fallout
21:28:59 <jonmasters> Dennis didn't get chance to do that so we'll need another plan
21:30:23 <ahs3> if i can get a source package list, i can scan the archives and at least generate the patches independently of the rebuild
21:30:28 <jonmasters> indeed so
21:30:39 <jonmasters> we'll sync with Dennis to work out a secondary contingency option
21:30:53 <jonmasters> then we'll bypass the secondary bypass regulatory and override the mainframe
21:31:20 <jonmasters> <cue scrolling HTML5 green text from stage left>
21:31:28 <pwhalen> :)
21:31:42 <ctyler> jonmasters: sounds suspiciously like a webapp there!
21:31:53 <jonmasters> ctyler: I said that just for you ;)
21:32:08 <jonmasters> ctyler: I'll eat that hat eventually ;)
21:32:18 <jonmasters> ok so other mass rebuild status?
21:32:27 <jonmasters> anything really gnarly need me to look?
21:33:05 <jonmasters> I'm going to mostly poke at kernel stuff - sent peter initial comments on the 3.8 config
21:33:45 <pwhalen> #topic 4) ARM Tech Talks - suggestions for future talks
21:34:00 <jonmasters> so apparently we want me to do a flyswatter talk
21:34:03 <jonmasters> what else?
21:34:22 <jonmasters> John Dulaney is doing a Chromebook talk this Fri
21:34:25 <pwhalen> thanks jonmasters for the last talk
21:34:26 * ctyler has some ideas floating around but not ready to commit yet
21:34:36 <jonmasters> #info John Dulaney is doing a Chromebook Tech Talk this Fri
21:34:49 <seano> needs to be posted to the list. :)
21:34:51 <pwhalen> jonmasters, do we know when?
21:35:46 <jonmasters> I am assuming at 15:00EST on Fri or so, checking
21:36:19 <jonmasters> #info link for tech talks https://fedoraproject.org/wiki/Architectures/ARM/Talks/ARMTechTalks
21:37:16 <pwhalen> I'll find out when for Friday and post something to the list, we should make it a consistent time
21:38:22 <pwhalen> anything else for tech talks?
21:38:42 <pwhalen> #topic 5) Open Floor
21:38:56 <jonmasters> pwhalen: thanks
21:39:42 <pwhalen> anything else? we have lots of time
21:40:01 <jonmasters> I'm happy to wrap up and we can followup by email
21:40:19 <seano> i really don't have anything. :)
21:40:21 <jcapik> fedoraproject.org doesn't seem to work
21:40:32 <ctyler> WAT? A #fedora-arm meeting finishing early?
21:40:37 <jonmasters> jcapik: seems to be some kind of proxy slowness or something - does load eventually
21:40:56 <jcapik> jonmasters: 502 bad gateway :]
21:41:06 <pwhalen> took a while, but loaded here
21:41:08 <jonmasters> oh, just sluggish when I tried
21:41:35 <masta> we need  to start pushing our kickstarts upstream
21:41:36 <pwhalen> ok, unless anyone objects we'll end now
21:41:55 * masta eof
21:42:26 <jcapik> c.u.
21:42:37 <pwhalen> #endmeeting