- Thursday, August 19, 2021

release-team@conference.openafs.org

Thursday, August 19, 2021< ^ >

kaduk@jabber.openafs.org/barnowl has set the subject to: openafs release team

Room Configuration

Room Occupants

GMT+0
[02:44:38] Cheyenne leaves the room
[11:41:09] mbarbosa joins the room
[12:53:13] meffie joins the room
[13:42:26] Cheyenne joins the room
[15:53:34] wiesand joins the room
[16:00:40] <kaduk@jabber.openafs.org/barnowl> greetings
[16:01:00] <meffie> hello
[16:01:10] <Cheyenne> hello everyone
[16:03:18] <wiesand> Hello
[16:04:09] <wiesand> I'm in the process of updating the wish list and the bookkeeping. Alas, I haven't gotten quite as far as I had hoped before the meeting.
[16:04:22] <kaduk@jabber.openafs.org/barnowl> I took a couple sick days this week, so not much from my end, either
[16:04:44] <wiesand> OUch, hope you're better already?
[16:05:21] <wiesand> But just to share what I have at hand:
[16:05:36] <meffie> Get well soon Ben
[16:06:10] <kaduk@jabber.openafs.org/barnowl> thanks
[16:06:41] <wiesand> There are two open changes already pulled up: The relatively new 14740 by Jeffrey (review welcome) and the postponed 14594 by Mike.
[16:07:44] <meffie> oh, i remember that day.
[16:08:17] <wiesand> The latter has a connection with 10831 (on the list for a long time) and IIRC more changes to server logging. It may be expedient to pull up more of that.
[16:08:43] <wiesand> Opinions welcome!
[16:10:04] <kaduk@jabber.openafs.org/barnowl> The logging stuff is probably worth having.  I am, however, not
entirely sure how to construct a stack of commits that is free of
conflicts while cherry-picking.
[16:10:19] <wiesand> Then there was mention a while ago of 14294 and 14295. These seem a bit hopeless to me since lots of other changes to the test framework haven't made it into 1.8 yet, and I'm not sure this topic is worth the effort.
[16:10:21] <meffie> it depends on if you want a backported set of patches.
[16:10:45] <wiesand> Ben: I'll give it a try. I *think* it's doable.
[16:11:27] <meffie> i think you'll need a backport, eg changes.
[16:12:00] <kaduk@jabber.openafs.org/barnowl> 14295 is a "sweeping change" that touches all tests.  It's just not
going to apply cleanly to the different branch.
Which I guess is the same thing Mike said
[16:12:37] <wiesand> I think cherry-picks for server logging won't be 100% clean but the modifications should be minor and comprehensible.
[16:12:57] <meffie> agreed
[16:13:10] <wiesand> Re 14294/5, I'm inclined to drop them from the list.
[16:14:22] <kaduk@jabber.openafs.org/barnowl> We may have to revisit if we get new tests we want to backport, but
dropping 14294/5 for now seems okay
[16:14:38] <meffie> well, i guess it depends on how must make checking you want to do on the stable branch. we try to add tests when we push to the master branch, so if the tests dont keep up, you will not have clean pullups in the future.
[16:15:05] <wiesand> More topics on the wish list are aix, cleanup-LINUX_ENV, 14662 (FDH_BLOCKSIZE), 14691 (LSM), 14465 & co. (RX SACK) and translate_et support for UAE
[16:16:35] <meffie> maybe we can pullup 14294 first, and do a 1.8 version of "tests: Accommodate c-tap-harness 4.7" if needed?
[16:17:05] <wiesand> Then there's 14595 (rx delays in multi_End_ignore) and kdump/afsinstall (14623/4)
[16:17:26] <kaduk@jabber.openafs.org/barnowl> > pullup 14294 first, and do a 1.8 version of "tests: Accommodate
> c-tap-harness 4.7" if needed?
that  would be the way to do it, I think
[16:17:26] <wiesand> I think 14294 was rather easy, I'll give it a try.
[16:20:12] <wiesand> I have 13540 and 14691 jotted down with a question mark behind - have to look at those again.
[16:21:03] <wiesand> Mentioned by Mark: 12183 and 14195, + maybe the rest of the cache-metrics topic
[16:21:46] <kaduk@jabber.openafs.org/barnowl> 14691 was already mentioned in the earlier message, I think (the LSM
fix)
[16:21:58] <kaduk@jabber.openafs.org/barnowl> So I don't think there's much of a question mark about 14691
[16:22:39] <kaduk@jabber.openafs.org/barnowl> 12183 and 14195 should be pretty self-contained, and are worth having
[16:22:58] <wiesand> Ah, sorry - that's what I mean by "not quite as finished as I'd like"
[16:23:34] <wiesand> 14161 (rx: Indent ifdef maze in rx_kernel.h) seems clean if accompanied by "FBSD: Remove support for 8.x and 9.x" which was already approved here a while ago
[16:23:57] <wiesand> + more FBSD pullups including the one mention by Måns
[16:24:49] <meffie> btw, mans freebsd12 buildbot builder is running now (as a nightly builder), 11 and 13 are on the way.
[16:25:07] <wiesand> Then there's "previousPacket handling" 14658..14661
[16:25:11] <kaduk@jabber.openafs.org/barnowl> > what I mean by
no worries
[16:25:26] <wiesand> It's good to have those builders, thanks!
[16:25:41] <kaduk@jabber.openafs.org/barnowl> Yeah, I should prioritize looking at the previousPacket stuff
[16:25:46] <kaduk@jabber.openafs.org/barnowl> the current state is really bad
[16:26:32] <wiesand> And that's all :-)
[16:27:08] <wiesand> Actually it's kind of enough to warrant a stable release IMO
[16:27:44] <kaduk@jabber.openafs.org/barnowl> :)
[16:27:47] <wiesand> But I'm sure there's more to come. At least there's no lack of items to work on ;-)
[16:29:13] <meffie> thanks wiesand
[16:29:15] <wiesand> Please feel free to comment on those, here or on -devel or by PM, whatever you prefer.
[16:29:26] <kaduk@jabber.openafs.org/barnowl> Does anyone want to bring up topics for master, or should we call the
meeting early?
[16:29:53] <Cheyenne> good linux builds on master and 1.8.x with current linux repo (5.14-rc6+)
[16:30:00] <meffie> thanks Cheyenne.
[16:30:14] <wiesand> Thanks Cheyenne!
[16:30:50] <kaduk@jabber.openafs.org/barnowl> Thanks Cheyenne!
[16:30:58] <meffie> not urgent and not for master, but i did want to mention that 'git fetch git://git.openafs.org/openafs.git' is still down
[16:31:21] <meffie> down == hangs indefinitely
[16:31:53] <wiesand> I think it's just very very slow. My git pull attempts today succeeded, but really took ages to complete.
[16:31:56] <kaduk@jabber.openafs.org/barnowl> Huh, I did get a fetch from that URL to work once.
[16:31:59] <kaduk@jabber.openafs.org/barnowl> Really slow, yes
[16:32:42] <meffie> i can pull from gerrit.openafs.org with ssh: and with https:
[16:32:45] <kaduk@jabber.openafs.org/barnowl> That flow is going through a git 1.7 git-daemon, and git 1.7 seems to
behave very inefficiently on the git repo that the current version of
gerrit's replication plugin created
[16:33:38] <kaduk@jabber.openafs.org/barnowl> I tweaked the gerrit and apache config to use a newer git version
where possible, and rate-limited inetd serving git:// URLs to avoid
stressing the whole machine, but that is the best I can do until
services are migrated to the EL7 system
[16:34:24] <meffie> thanks ben
[16:34:55] <meffie> i hope you feel better soon. i have nothing more.
[16:35:00] <kaduk@jabber.openafs.org/barnowl> thanks
[16:36:06] <kaduk@jabber.openafs.org/barnowl> so, motion to adjourn?
[16:36:35] <Cheyenne> sounds good
[16:36:47] <wiesand> motion sustained.
[16:37:03] <wiesand> Thanks a lot everyone, and get well soon Ben.
[16:44:15] meffie leaves the room
[19:31:17] wiesand leaves the room
[21:51:57] mbarbosa leaves the room
[22:35:25] Cheyenne leaves the room
[22:35:49] Cheyenne joins the room
[22:36:21] Cheyenne leaves the room