- Wednesday, September 6, 2017

release-team@conference.openafs.org

Wednesday, September 6, 2017< ^ >

Room Configuration

Room Occupants

GMT+0
[06:12:49] kadukoafs@gmail.com/barnowl772461F2 leaves the room
[06:16:06] kadukoafs@gmail.com/barnowl772461F2 joins the room
[07:17:33] kadukoafs@gmail.com/barnowl772461F2 leaves the room
[07:34:09] kadukoafs@gmail.com/barnowl772461F2 joins the room
[15:01:21] <kadukoafs@gmail.com/barnowl772461F2> $timeofday
[15:02:52] <mvita> hello
[15:03:07] <kadukoafs@gmail.com/barnowl772461F2> small crowd, today
[15:03:10] <mvita> sorry for the delay, Mike is out so I'm scrambling today
[15:03:53] <kadukoafs@gmail.com/barnowl772461F2> Don't worry about it
[15:04:40] <kadukoafs@gmail.com/barnowl772461F2> I think we're probably keeping 1.6 on hold until Stephan is back,
though IIRC I saw something go by that linux 4.13 finalized
[15:06:18] <mvita> yes, it did
[15:06:41] <mvita> our linux buildslaves are still floundering for non-linux reasons
[15:06:43] <kadukoafs@gmail.com/barnowl772461F2> Sounds like we're still okay, then
[15:06:51] <kadukoafs@gmail.com/barnowl772461F2> Oh, hmm.
[15:06:56] <mvita> but we need to get them fixed so we can know if we are okay
[15:07:02] <kadukoafs@gmail.com/barnowl772461F2> *nods*
[15:07:09] <mvita> I should have some time to look at that later today
[15:07:16] <kadukoafs@gmail.com/barnowl772461F2> Cool
[15:07:28] <mvita> or make someone else look at it, since I am "virtual mike" this week
[15:07:48] <kadukoafs@gmail.com/barnowl772461F2> mwahahahaha?
[15:08:38] <mvita> <twirls mustache>
[15:08:43] <mvita> <strokes chin>
[15:08:49] <kadukoafs@gmail.com/barnowl772461F2> :)
[15:09:06] <kadukoafs@gmail.com/barnowl772461F2> I guess we can move on to 1.8, then
[15:09:22] <mvita> thanks for beta!
[15:10:01] <kadukoafs@gmail.com/barnowl772461F2> The tag and announcement went out, though I guess I'm supposed to
update the web with a link to the announcement
[15:10:04] <kadukoafs@gmail.com/barnowl772461F2> you're welcome!
[15:10:11] <kadukoafs@gmail.com/barnowl772461F2> I also put it into debian experimental
[15:10:32] <mvita> oh, nice.
[15:11:05] <kadukoafs@gmail.com/barnowl772461F2> Apparently the rxevent signal test fails on armhf, which I didn't get
a real chance to look at.
[15:11:24] <mvita> oh, who reported that?
[15:11:29] <kadukoafs@gmail.com/barnowl772461F2> But assuming that can be fixed, and no issues are reported, I expect
to put it into debian unstable in a week or two
[15:11:56] <kadukoafs@gmail.com/barnowl772461F2> It's automatic -- I upload a source package and optionally one
architecture binary package, and the buildd's try to do binary builds
for the other architectures.
[15:12:10] <kadukoafs@gmail.com/barnowl772461F2> So I get an email and link to things like
https://buildd.debian.org/status/fetch.php?pkg=openafs&arch=armhf&ver=1.8.0~pre2-1&stamp=1504597342&file=log
if things fail.
[15:12:42] <kadukoafs@gmail.com/barnowl772461F2> IIRC at pre1 time there were several more test failures, which
prompted the jenkins hash endianness fixes from Anders.
[15:13:40] <mvita> wow
[15:13:44] <kadukoafs@gmail.com/barnowl772461F2> Any thoughts on how long we should wait for feedback (and receive
none) before calling it a final release?
[15:14:01] <kadukoafs@gmail.com/barnowl772461F2> I'm currently leaning towards the 1-2 month timescale.
[15:14:09] <mvita> I know Mike will want to weigh in - he'll be back next week.
[15:14:41] <kadukoafs@gmail.com/barnowl772461F2> Sure.  There's no rush, on that timescale.
[15:15:05] <mvita> we have been working a lot on various automated tests as well, and I'd like to set up a test cell for 1.8 for those
[15:15:20] <kadukoafs@gmail.com/barnowl772461F2> > automated tests
Ooh, that's highly exciting
[15:16:55] <mvita> I discovered how Ceph does it yesterday, and now I am itching to adapt their concepts too
[15:17:04] <kadukoafs@gmail.com/barnowl772461F2> :)
[15:17:29] <mvita> docs.ceph.com/teuthology/docs/README.html <http://docs.ceph.com/teuthology/docs/README.html>
[15:18:57] <kadukoafs@gmail.com/barnowl772461F2> thanks for the link
[15:19:00] <mvita> we have started discussing Solaris futures with some sites;  I think we are going to make some internal illumos test machines and build slaves
[15:19:24] <kadukoafs@gmail.com/barnowl772461F2> Sure.  The Oracle situation is pretty sad.
[15:19:32] <mvita> tragic
[15:20:51] <mvita> do you think it would be helpful for openafs.org to have an illumos/smartos build slave?
[15:21:01] <kadukoafs@gmail.com/barnowl772461F2> Still, I have the solaris chain of stuff in gerrit open in a browser
tab
[15:21:28] <kadukoafs@gmail.com/barnowl772461F2> It probably would be helpful, since that seems the direction things
will go in the future
[15:21:29] <mvita> yes, I have some homework on those after Andrew's reviews
[15:22:02] <kadukoafs@gmail.com/barnowl772461F2> I guess I can preview that "you're both wrong" (and both right) about
lock ordering
[15:22:38] <mvita> I meant to abandon the lock ordering ones long ago and forgot about them
[15:22:50] <mvita> he's right that only obtain order matters
[15:23:00] <kadukoafs@gmail.com/barnowl772461F2> Ah.  I'm not opposed to some tweaks going in, with the right commit
messages
[15:23:39] <kadukoafs@gmail.com/barnowl772461F2> Anyway, that's on my list to look at, though I don't expect to have
too much time this week for openafs stuff.
[15:24:12] <kadukoafs@gmail.com/barnowl772461F2> You saw that I refreshed the rxgk patches in gerrit?  I managed to get
my overall branch rebased and building, though I didn't really do any
runtime testing yet.
[15:24:27] <kadukoafs@gmail.com/barnowl772461F2> But most of it is pretty orthogonal to the core code, so it "ought to"
work.  Famous last words, I know.
[15:24:28] <mvita> well, maybe you should hold off until I do a new PS
[15:24:55] <mvita> (hold off on reviewing the solaris stack)
[15:25:13] <mvita> No, I missed your rxgk updates
[15:25:18] <mvita> thank you!
[15:25:23] <kadukoafs@gmail.com/barnowl772461F2> I expect I can manage to hold off for that, yes
[15:25:27] <mvita> I will definitely try them out.
[15:25:32] <kadukoafs@gmail.com/barnowl772461F2> Yay!
[15:25:51] <kadukoafs@gmail.com/barnowl772461F2> What's in gerrit is supposed to be enough for the intra-dbserver
connections to be speaking rxgk (well, vl and pt)
[15:26:37] <mvita> okay - that's where I'm spending a lot of time these days, so very apropos
[15:26:44] <kadukoafs@gmail.com/barnowl772461F2> heh :)
[15:26:51] <mvita> stress testing on ubik
[15:27:19] <kadukoafs@gmail.com/barnowl772461F2> As far as getting something on "actual" clients to exercise rxgk, the
fastest path may be via bos.  So resurrecting the pthread-bos patchset
is also on my radar.
[15:27:49] <mvita> oh, I thought that was in 1.8?  or at least on master, done.
[15:28:24] <kadukoafs@gmail.com/barnowl772461F2> I think it ended up being too risky for 1.8
[15:28:30] <mvita> lots of logging and signal considerations there, iirc
[15:28:35] <kadukoafs@gmail.com/barnowl772461F2> Though maybe some parts of it went in
[15:29:36] <mvita> LWP can't die fast enough in my view - I've been deep in the guts of it for the past week.
[15:30:43] <kadukoafs@gmail.com/barnowl772461F2> Yeah.
[15:30:59] <kadukoafs@gmail.com/barnowl772461F2> Plenty of other work to do on master, moving tfoo to foo and trying to
convert the stragglers
[15:31:25] <mvita> yes: vos and the other little clients
[15:32:52] <kadukoafs@gmail.com/barnowl772461F2> Any other topics for today?
[15:33:30] <mvita> no, except to ask if there's anything we can help w/ on 1.8.x this week
[15:33:35] <mvita> (other than testing)
[15:33:55] <mvita> I think my highest priority is getting the linux builders fixed so we know where we stand
[15:33:56] <kadukoafs@gmail.com/barnowl772461F2> I think we're on "test test test" at this point.  Though "review
review review" is never wrong ;)
[15:34:17] <kadukoafs@gmail.com/barnowl772461F2> That sounds good, too.
[15:34:39] <mvita> that should be enough to keep me out of trouble for another week
[15:35:18] <kadukoafs@gmail.com/barnowl772461F2> more than enough, especially when you are acting as Mike
[15:35:51] <kadukoafs@gmail.com/barnowl772461F2> Thanks for being here!
[15:35:53] <mvita> okay, that's all I have for today.  Thank you Ben, have a good one.
[15:35:59] <kadukoafs@gmail.com/barnowl772461F2> You, too
[22:29:50] mvita leaves the room