- Thursday, April 29, 2021

release-team@conference.openafs.org

Thursday, April 29, 2021< ^ >

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

Room Configuration

Room Occupants

GMT+0
[12:08:27] mbarbosa joins the room
[13:23:30] Cheyenne joins the room
[13:54:30] mbarbosa leaves the room
[13:54:32] mbarbosa joins the room
[15:15:29] meffie joins the room
[15:59:59] <kaduk@jabber.openafs.org/barnowl> greetings
[16:00:13] <Cheyenne> Good morning Ben
[16:00:20] <meffie> greetings
[16:00:45] <mvita2> oh, it's now
[16:00:51] <mvita2> heh
[16:01:07] <mvita2> I forgot to update my calendar
[16:01:17] <mvita2> good thing I'm always lurking here
[16:01:51] <meffie> fyi we have a hard stop at the top of the hour
[16:03:51] <kaduk@jabber.openafs.org/barnowl> I would become triple-booked at the top of the hour, too, so let's
finish on time
[16:04:04] <kaduk@jabber.openafs.org/barnowl> (I become double-booked at quarter-past)
[16:05:43] <kaduk@jabber.openafs.org/barnowl> In addition to only being a 6-day gap between meetings, this week is
also the IESG retreat, so I have not gotten much done, myself.
[16:06:02] <kaduk@jabber.openafs.org/barnowl> I think I saw Stephan pull up a few more things for 1.8.x and we got
some reviews, though
[16:06:18] <meffie> i am currently working on buildbot master documentation and playbook cleanup.
[16:07:19] <meffie> kaduk if you log into either hosts running the buildbot, look for a new README file
[16:07:45] <kaduk@jabber.openafs.org/barnowl> Ooh, that sounds wonderful -- thanks!
[16:07:56] <Cheyenne> linux 5.12 is now "ga", I ran master and 1.8.x builds against linux.5.13rc1 (pre) without any problems.
[16:09:16] <kaduk@jabber.openafs.org/barnowl> Maybe I should have sent a calendar invite to the list, to make sure I
got the time right (still no Stephan yet, right?)
[16:09:17] <meffie> we are still running buildbot 1.x, and have 2.x on the "new" rhel7, but buildbot 3.x was recently released. looks like configuration changes (at least) are needed for 3.x
[16:09:33] <kaduk@jabber.openafs.org/barnowl> The price of progress, I guess?
[16:09:48] <meffie> yes
[16:10:07] <meffie> do we want to cut over to the new rhel7 box soon?
[16:10:26] <kaduk@jabber.openafs.org/barnowl> That would be ideal, yes
[16:11:13] <kaduk@jabber.openafs.org/barnowl> (IIRC the old buildbot version has an embedded ... Twistd(?) component
that has some CVEs against ist.)
[16:12:24] <meffie> do we just tell people running bots to point to the new host, or change the host name so it is transparent?
[16:12:48] <kaduk@jabber.openafs.org/barnowl> It's probably best to change the host name, but I think we have to
coordinate with jhutz/Chaskiel for DNS changes.
[16:13:30] <kaduk@jabber.openafs.org/barnowl> I believe the DNS is hosted at Amazon Route53 at the moment, but the
account does a lot of central.org stuff as well as openafs.org stuff,
so I don't have useful access to it (yet?)
[16:13:34] <meffie> ok, i'm not sure what is best. at the moment i'm updating both (with ansible)
[16:14:17] <meffie> i just have one worker pointed at the new buildbot.
[16:15:14] <kaduk@jabber.openafs.org/barnowl> Are we in a spot where it would "just work" to cut over the DNS?
[16:15:25] <meffie> yes
[16:15:27] <kaduk@jabber.openafs.org/barnowl> in terms of worker passwords/etc?
[16:15:28] <kaduk@jabber.openafs.org/barnowl> cool
[16:16:47] <meffie> i should recheck the gerrit integration is still working, it's been a while since i looked.
[16:18:00] <kaduk@jabber.openafs.org/barnowl> *nods*
[16:18:05] <meffie> thank you
[16:18:16] <kaduk@jabber.openafs.org/barnowl> But that should be relatively straightforward to get back if it has
stopped working
[16:18:19] wiesand joins the room
[16:18:43] <wiesand> Hi, sorry I'm late.
[16:18:44] <meffie> yes
[16:19:53] <meffie> hello.
[16:20:34] <wiesand> No idea where those lat 20 inutes went…
[16:20:44] <kaduk@jabber.openafs.org/barnowl> I know the feeling...
[16:24:51] wiesand leaves the room: Replaced by new connection
[16:25:29] wiesand joins the room
[16:25:53] <wiesand> And right now there was trouble with my internet connection :-(
[16:26:10] <kaduk@jabber.openafs.org/barnowl> (My other meeting started, so I will try to type a couple things a bit
asynchronously, and sorry for disrupting a logical flow of the
meeting)
[16:26:27] <wiesand> Last thing I was about to write before I got disconnected:
[16:26:52] <kaduk@jabber.openafs.org/barnowl> I see there's a 'macos-prefpane' topic that showed up; some of it
looks like generic cleanup (dead code removal), but how high of a
priority is the "real changes" part of the stack?
[16:27:09] <wiesand> Since the last meeting I pulled up the vol-bless topic. That's kind of a no-brainer and will be merged later.
[16:28:43] <wiesand> I also pulled up bigsur (with minor edits) and a related change. That's not a no-brainer at all, at least not to me.
[16:28:45] <wiesand> https://gerrit.openafs.org/#/q/status:open+project:openafs+branch:openafs-stable-1_8_x+topic:macos-bigsur
[16:29:08] <mvita2> the real meat of the macos-prefpane topic is to prevent some confusing behavior during installs
[16:29:46] <mvita2> it's not urgent, but it is intended to improve user experience
[16:30:10] <wiesand> I marked two of those changes -2, but please don't let that keep you from reviewing the stack. If you convince me it's correct, or propose a reasonable alternative, fine.
[16:30:23] <meffie> cool thanks
[16:30:25] <kaduk@jabber.openafs.org/barnowl> Thanks, Mark (re macos-prefpane)
[16:30:57] <kaduk@jabber.openafs.org/barnowl> In terms of macos-bigsur for 1.8.x, I think we are going to be stuck
with something fairly disruptive just based on the scale of changes
that Apple is forcing on us
[16:31:13] <kaduk@jabber.openafs.org/barnowl> So I don't know if Stephan is concerned about it for 1.8.8
specifically or ever
[16:32:32] <wiesand> Well, in the best of all worlds, the stack I propose works and makes sense. It is fairly minimal after all.
[16:33:18] <kaduk@jabber.openafs.org/barnowl> "the stack you propose" including the ones currently marked -2?
[16:33:34] <wiesand> yes, it does ;-)
[16:33:47] <kaduk@jabber.openafs.org/barnowl> got it :)
[16:36:43] <kaduk@jabber.openafs.org/barnowl> I also wanted to ask if anyone had gotten a chance to take a look at
14465 (RX_ACK_TYPE as bit field)
[16:37:20] <mvita2> I looked but did not grok.
[16:37:48] <mvita2> it would have been helpful to see an example of the intended future use, but I guess that will not be forthecoming
[16:37:59] <kaduk@jabber.openafs.org/barnowl> ok, thanks for letting me know.  (Do you want more
background/explanation?)
[16:38:07] <wiesand> NB 14599 is the kind of change we discussed last week: IMO that one should be pulled up a.s.a.p.
[16:39:15] <mvita2> oh, I see there is more discussion in 14465 that wasn't there when I looked the first time
[16:39:22] <mvita2> let me read that and then we'll see
[16:39:31] <kaduk@jabber.openafs.org/barnowl> *nods*
[16:43:28] <mvita2> I have been doing a lot of OpenAFS work under Solaris lately and have repeatedly seen the known crash fixed by https://gerrit.openafs.org/#/c/12183; so I need to rebase that topic and get it moving again.
[16:45:38] <kaduk@jabber.openafs.org/barnowl> (my standing question about anything I should prioritize is also open,
if people have thoughts)
[16:46:52] <mvita2> for me, just 12183 (after I rebase and make a few tweaks) - I'd like to see that in 1.8.x. some day
[16:47:27] <mvita2> oh, and 14608 which is new
[16:48:19] <kaduk@jabber.openafs.org/barnowl> Oh, oops, I +2'd an early 12183 rev but it's not merged
[16:48:54] <kaduk@jabber.openafs.org/barnowl> I started looking at 14608 last night but concluded that I was too
tired to trust myself to think about ubik protocol stuff
[16:49:20] <mvita2> understood
[16:51:26] <mvita2> maybe I should add jhutz to the reviewer list for 14608?  what do the rest of you think?
[16:51:31] Cheyenne leaves the room
[16:52:26] <wiesand> Earmarked 12183 for inclusion in 1..x
[16:52:30] <wiesand> 1.8.
[16:52:52] <kaduk@jabber.openafs.org/barnowl> asking jhutz for review in gerrit can't hurt, but I don't expect it to
be the most effective mechanism to get his attention
[16:53:08] <mvita2> *nod*
[16:53:23] <wiesand> I have another question on "backporting".
[16:54:16] <meffie> ok
[16:54:36] <wiesand> Part of the 1.8 bigsur stack is 14599 "rx: Indent ifdef maze in rx_kcommon.c"
[16:54:53] <mvita2> (reminder:  many of us have a hard-stop in 5m)
[16:55:23] <wiesand> That's the kind of "cleanup/refactoring"change we discussed last week. IMO those should be pulled up swiftly.
[16:55:58] <kaduk@jabber.openafs.org/barnowl> That seems likely, yes
[16:56:10] <wiesand> A similar change is 14161 "rx: Indent ifdef maze in rx_kernel.h"
[16:56:28] <kaduk@jabber.openafs.org/barnowl> E.g., in openssl when we did some include file restructuring we did it
on all the current branches at basically the same time, to avoid
endless merge conflicts trying to backport things
[16:56:56] <wiesand> Alas, that one is not a clean cherry-pick. It does become a clean cherry-pick when based upon a4e9365 (FBSD: Remove support for 8.x and 9.x)
[16:57:02] <kaduk@jabber.openafs.org/barnowl> I think there were four or five ifdef-reindents that landed close to
each oterh
[16:57:50] <wiesand> Now, is removing FBSD 8/9 support from 1.8.x an option?
[16:57:56] <kaduk@jabber.openafs.org/barnowl> I think so
[16:58:06] Cheyenne joins the room
[16:58:19] <meffie> yes
[16:58:26] <kaduk@jabber.openafs.org/barnowl> The last FreeBSD 9.x release went out of support at the end of 2016
[16:58:32] mbarbosa leaves the room
[16:58:51] mbarbosa joins the room
[16:59:20] <wiesand> Right, but then this is the "stable series". Anyway, I take it that removing FBSD<10 support from 1.8.x is ok. Thanks!
[16:59:54] <wiesand> That were the more urgent thins I had to discuss today.
[17:00:07] <meffie> ok, thank you.
[17:00:12] <kaduk@jabber.openafs.org/barnowl> Bye people with a hard stop
[17:00:18] <wiesand> Please review the bigsur stack.
[17:00:38] <meffie> ok thank you, notes to be posted today. ;)
[17:01:23] <wiesand> And I hope the new meeting slot will allow me to get some openafs work done tomorrow.
[17:01:37] <wiesand> Let's adjourn. Thanks a lot everybody!
[17:01:54] Cheyenne leaves the room
[17:01:54] Cheyenne joins the room
[17:04:00] <kaduk@jabber.openafs.org/barnowl> Yes, thanks everyone!
[17:07:54] Cheyenne leaves the room
[18:30:08] wiesand leaves the room
[19:49:59] Cheyenne joins the room
[20:24:22] meffie leaves the room
[21:50:24] mbarbosa leaves the room