Home
release-team@conference.openafs.org
Friday, January 31, 2020< ^ >
meffie has set the subject to: openafs release team
Room Configuration
Room Occupants

GMT+0
[12:00:58] mbarbosa joins the room
[12:51:27] mvita leaves the room
[15:00:04] mvita joins the room
[15:35:27] mvita leaves the room
[15:36:44] meffie joins the room
[16:12:26] mvita joins the room
[16:31:59] mvita leaves the room
[16:46:17] kaduk@jabber.openafs.org/barnowl joins the room
[16:49:59] wiesand joins the room
[16:58:10] <meffie> good afternoon/morning
[16:58:28] <kaduk@jabber.openafs.org/barnowl> greetings
[16:58:36] <wiesand> good evening
[17:00:38] <wiesand> I made some progress after the last meeting, over the weekend.
[17:00:54] <wiesand> Merged away the better part of the 1.8 backlog.
[17:01:32] <wiesand> The remaining changes in the queue mostly look ok to me, but require more review.
[17:01:36] <meffie> yes, i saw, that's wonderful.
[17:02:24] <wiesand> Here's a list of those marked "requires more review" in my books, in order of priority (first = highest):
[17:02:41] <wiesand> 13738 13750 13894 13900 13901 13904 13968 13974 13975 13895 13896 13897 13906 13908 13938 13951 14018 14019 14031..14037
[17:02:44] <kaduk@jabber.openafs.org/barnowl> It was nice to see that stuff getting in, even if it does mean more
mail for me to go through
[17:03:59] <wiesand> We're also back to getting a buildbot +1 on 1.6 changes (if they build that is), thanks to the windoze work.
[17:04:21] <meffie> i'll do notes today and add those numbers to the list.
[17:05:01] <wiesand> Alas, the windows builders have become pretty slow - a turnaround time of > 1h means you rebase a handful of changes, and that's what you get done for the day.
[17:05:23] <wiesand> mike: thanks
[17:05:51] <kaduk@jabber.openafs.org/barnowl> Hmm, I will try to ping Alejandro and see if he can take a look (but
it may be issues with the underlying hypervisor/resourcing)
[17:06:00] <kaduk@jabber.openafs.org/barnowl> We could also consider not doing all 4 builds all the time
[17:06:18] <kaduk@jabber.openafs.org/barnowl> even (e.g.) 32-bit-checked plus 64-bit-unchecked would still be pretty
good coverage
[17:07:38] <wiesand> I can't judge on that. Just wanted to mention that it slowed me down.
[17:08:12] <meffie> let me know if there is anything i can change on the buildbot master (e.g. remove some of them from gerrit builds)
[17:08:17] <kaduk@jabber.openafs.org/barnowl> I think I've noticed "waiting for the windows builders" to be painful
as well.
[17:08:55] <kaduk@jabber.openafs.org/barnowl> Mike: the bit about commenting-out the 32-bit free and 64-bit checked
builds was directed at you :)
[17:09:38] <meffie> ah!! sorry i am dense.
[17:09:39] <kaduk@jabber.openafs.org/barnowl> Since IIUC there is a single VM that provides all the logical builders
[17:10:11] <wiesand> Interestingly, I think build time was ~20m initially, and then gradually went up to >60m. So yes, looks like it's not a fundamental problem but rather a resource issue on the hypervisor.
[17:10:41] <kaduk@jabber.openafs.org/barnowl> Or windows needing defragmentation or something like that :)
[17:11:13] <wiesand> windows keeps wowing me", yes ;-)
[17:12:04] <kaduk@jabber.openafs.org/barnowl> Well that's expected -- the Windows on Windows subsystem for 32-bit
compat on 64-bit hosts is a WoWer
[17:12:31] <wiesand> ah, it's a feature then
[17:14:11] <wiesand> and then I noticed 14046 and the corresponding RT #135065
[17:15:34] <kaduk@jabber.openafs.org/barnowl> I'm glad you noticed, as I am just flying blind here
[17:16:13] <wiesand> Yes, I tend to use "conservative systems" for cutting releases. And yes, I'm aware that the system used does impact the resulting configure stuff and more (I'm using an updated pod2man for example).
[17:16:47] <meffie> (buildbot config changed, will restart in a bit)
[17:17:34] <wiesand> Do we need to standardize on an OS environment in which "makerelease" is run?
[17:18:12] <kaduk@jabber.openafs.org/barnowl> I don't think so
[17:18:30] <wiesand> I could probably comply using containers
[17:18:55] <wiesand> (as long as it's Linux ;-)
[17:19:01] <kaduk@jabber.openafs.org/barnowl> It would be ridiculous to mandate (e.g.) that it be run on debian, or
RHEL
[17:19:41] <wiesand> kind of, right
[17:20:28] <wiesand> mandate a minimum autoconf/automake/pod2man instead?
[17:21:22] <kaduk@jabber.openafs.org/barnowl> Perhaps.  I don't have particular insight into how those versions
would be determined, though.
[17:21:47] <kaduk@jabber.openafs.org/barnowl> We could presumably start by tabaulating what the current version is
on some popular distro/versions
[17:25:00] <wiesand> makes sense
[17:26:27] <wiesand> EL7: autoconf-2.69, automake-1.13.4
[17:27:05] <wiesand> EL6: autoconf-2.63, automake-1.11.1
[17:28:02] <wiesand> U18: autoconf-2.69, automake-1.15.1
[17:28:52] <wiesand> How to know which versions know about ppc64le?
[17:29:35] <kaduk@jabber.openafs.org/barnowl> I want to say autoconf-2.69, but am not sure why.
[17:30:20] <wiesand> I agree that these problems should come up very infrequently only though. And maybe those on these bleeding edge platforms should indeed be prepared to run regen.sh and report issues…
[17:30:53] <wiesand> I vow not to cut tarballs on EL6 any more ;-) (quite possible I did this for 1.8.4)
[17:31:48] <wiesand> We should have a 1.6.6pre1 "rather soon" - that would be an occasion to ask the reporter for feedback.
[17:32:13] mbarbosa leaves the room
[17:32:30] <wiesand> (planning to run makerelease on EL7, thus with autoconf-2.69)
[17:32:34] <kaduk@jabber.openafs.org/barnowl> <oblig>
commit d294e9cea5bf7c8b748355537c2f423eb177b035 (tag:
openafs-stable-1_6_6pre1)
Author: Stephan Wiesand <stephan.wiesand@desy.de>
Date:   Thu Oct 17 13:48:26 2013 +0200
    Make OpenAFS 1.6.6pre1
</oblig>
[17:32:56] <wiesand> hmm, pity we scared away Marcio…
[17:33:28] <meffie> 1.6?
[17:33:56] <meffie> you mean 1.8.6?
[17:33:57] <wiesand> er, 1.8.6, yeees
[17:34:06] <meffie> ah, ok.
[17:34:57] <meffie> the commit tag is wrong then?
[17:35:03] <kaduk@jabber.openafs.org/barnowl> hence my snarky <oblig />, yes
[17:35:31] <wiesand> Hmm, why?
[17:35:57] <meffie> oh, ben was kidding. i see.
[17:36:12] <meffie> sorry, dense today ;)
[17:36:24] <wiesand> and I'm not getting the "oblig"
[17:37:24] <kaduk@jabber.openafs.org/barnowl> Sorry; "oblig" is short for "obligatory", in that it's clear what you
meant to say but the actual characters on the stream don't match that,
so it's an obligatory correction
[17:37:38] <wiesand> thanks :)
[17:38:47] <kaduk@jabber.openafs.org/barnowl> sorry for the disruption
[17:38:51] <wiesand> still, how about that plan? I'll cut 1.*8*.6pre1 on a fully updated EL7 with the standard autoconf/automake and we wait/ask for feedback?
[17:39:01] <kaduk@jabber.openafs.org/barnowl> That sounds good to me
[17:39:20] <meffie> excellent thank you
[17:39:54] <wiesand> If that fails, I can use an U18 container or whatever.
[17:41:30] <wiesand> Besides review of candidate changes, 1.8.6pre1 needs a NEWS updates. This should be more feasible than usual for me to do because I can copy a lot from what Mike put together for 1.9 ;-)
[17:42:02] <meffie> woot
[17:42:19] <wiesand> Which brings me to "that's what I have on the stable series for today - on to 1.9/master?"
[17:42:55] <kaduk@jabber.openafs.org/barnowl> not a whole lot noteworthy; I had to rebase a couple things while
reviewing, but mostly this week I spent trying to catch up to things I
had earmarked to look at at last week's meeting.
[17:43:46] <kaduk@jabber.openafs.org/barnowl> But once I clear that I plan to tidy up the 1.9 NEWS
[17:44:23] <meffie> yay
[17:45:14] mbarbosa joins the room
[17:46:02] <wiesand> Anything else to discuss today?
[17:46:38] <kaduk@jabber.openafs.org/barnowl> I don't think I have anything else.
[17:47:10] <wiesand> Let's adjourn then. Thanks a lot, and have a nice weekend!
[17:47:32] <kaduk@jabber.openafs.org/barnowl> Thanks everyone!
[17:47:48] <meffie> thanks! have a good weekend.
[17:47:55] wiesand leaves the room
[17:48:01] meffie leaves the room
[18:04:31] mbarbosa leaves the room
[18:06:12] mbarbosa joins the room
[18:09:47] mvita joins the room
[18:19:41] mbarbosa leaves the room
[18:39:20] mbarbosa joins the room
[20:26:22] mbarbosa leaves the room
[20:44:58] mbarbosa joins the room
[21:56:48] mbarbosa leaves the room