- Thursday, June 24, 2021

release-team@conference.openafs.org

Thursday, June 24, 2021< ^ >

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

Room Configuration

Room Occupants

GMT+0
[11:50:35] mbarbosa joins the room
[13:19:13] meffie joins the room
[13:42:43] Cheyenne joins the room
[15:03:29] <mvita2> could someone please administer a gentle kick to the upstream buildbot ?  It seems to be wedged.
[15:06:19] <kaduk@jabber.openafs.org/barnowl> We swapped DNS around to point to the new coordinator, and it seems
that things are confused as a result
[15:06:43] <kaduk@jabber.openafs.org/barnowl> I wonder if bouncing the process on the old machine would cause the
workers to make a new connection, hmm...
[15:07:38] <mvita2> flushing caches, stand by
[15:08:11] <kaduk@jabber.openafs.org/barnowl> (I'm in a meeting so am not going to pull out privileged credentials
at the moment)
[15:08:20] <mvita2> np
[15:15:51] mvita2 leaves the room
[15:22:21] <kaduk@jabber.openafs.org/barnowl> Okay, stopping the coordinator process on the old machine
[15:22:44] <kaduk@jabber.openafs.org/barnowl> And lo, things are queuing up on the new one!
[15:33:03] mvita2 joins the room
[15:33:52] <mvita2> thank you!
[15:38:27] <kaduk@jabber.openafs.org/barnowl> Hmm, looks like maybe we lost the ForceGerritBuild button in the
cutover, though.
[15:39:15] <Cheyenne> ::looking::
[15:44:05] mvita2 leaves the room
[15:46:52] <Cheyenne> might be a python2 -> python3 issue
[15:47:19] wiesand joins the room
[15:47:35] mvita2 joins the room
[15:48:43] <Cheyenne> actually -- might be a simple syntax error..
[16:00:54] <kaduk@jabber.openafs.org/barnowl> Anyway, it's the top of the hour -- greetings!
[16:00:59] <mvita2> greetings
[16:01:02] <Cheyenne> hello
[16:01:02] <meffie> oh, sorry i missed the excitement!
[16:01:11] <wiesand> Hi
[16:01:21] <mvita2> my internet connection has been unstable this morning so I may drop intermittently
[16:02:34] <meffie> by the low build numbers, it looks like the new buildbot is online! woot
[16:02:51] <wiesand> yay
[16:03:22] <mvita2> did the port number change?
[16:03:27] <Cheyenne> looks like the new python in the buildbot master is picking up a syntax error that was being skipped by the earlier python.  I have a simple fix I will be pushing in a minute
[16:03:29] <meffie> no
[16:03:39] <mvita2> can’t establish a connection to the server at buildbot.openafs.org:8010.
[16:03:53] <meffie> you should be using port 80
[16:04:26] <mvita2> hmm.   80 != 8010
[16:04:49] <meffie> apache is in front of the web ui
[16:04:56] <kaduk@jabber.openafs.org/barnowl> The backend port number changed, I think
[16:05:09] <kaduk@jabber.openafs.org/barnowl> Not 80.  443!
[16:05:14] <mvita2> apparently I am befuddled
[16:05:26] <mvita2> more coffee
[16:05:37] <kaduk@jabber.openafs.org/barnowl> I don't think I actually put a proxy in place for port 80, heh
[16:05:38] <meffie> d'oh, yes not 80, 443, https
[16:05:49] <mvita2> I see builders
[16:06:28] <meffie> (yes, the backend port did change because this one is running a newer version, still old, but newer)
[16:06:54] <wiesand> 443 seems to work for me (TM)
[16:06:58] <meffie> thank you ben for getting the apache config working.
[16:08:07] <wiesand> Not much news from me today.
[16:08:31] <kaduk@jabber.openafs.org/barnowl> Stephan, I added you (I hope) to
https://rt-02.central.org/rt/Ticket/Display.html?id=135238 for a
potential 1.8.8 addition
[16:09:01] <kaduk@jabber.openafs.org/barnowl> Andrew held off on pushing the patch to gerrit until we had a sense
for how we wanted to handle it for 1.8.x
[16:09:52] <wiesand> Seen it, thanks.
[16:10:02] <kaduk@jabber.openafs.org/barnowl> Great
[16:11:07] <meffie> can we consider gerrit 14650 for 1.8.8 ?
[16:11:07] <Cheyenne> Hopefully the forcerebuild problem was just the syntax error.  I just pushed a PR to github that fixed that.
[16:11:27] <wiesand> I have an extremely strong preference towards Andrew just pushing it to gerrit, and we then proceed as usual (with a swift 1.8.8 release).
[16:11:29] <meffie> thanks cheyenne, i'll take a lok.
[16:12:11] <kaduk@jabber.openafs.org/barnowl> Excellent, that is what I was hoping for.  Do you want me to comment
on the ticket to that effect?
[16:12:37] <kaduk@jabber.openafs.org/barnowl> (Or I could just push it on his behalf, I suppose.)
[16:13:37] <wiesand> Please, go ahead. To be clear: with "proceed as usual", I mean "not a security release, just another bug fix".
[16:13:51] <meffie> +1
[16:14:12] <kaduk@jabber.openafs.org/barnowl> Agreed re "proceed as usual"
[16:14:35] <wiesand> :)
[16:15:25] <mvita2> that's a nice find
[16:15:47] <kaduk@jabber.openafs.org/barnowl> I guess that (other than that RT ticket), no news on 1.8 is good news?
I don't remember much else that we were going to wait for for
1.8.8-final.
[16:16:00] <meffie> can we consider gerrit 14650 for 1.8.8 ?
[16:16:29] <kaduk@jabber.openafs.org/barnowl> Mike mentioned 14650 (and I presume 14626?) as potentially for 1.8.8
as well
[16:16:52] <wiesand> I also received a mail saying that the 1.8.8pre1 announcement arrived with incorrect DMARC/DKIM headers and thus may have been blackholed for a number of recipients.
[16:16:56] <kaduk@jabber.openafs.org/barnowl> which I think I would be okay with (I just asked a question on 14626,
BTW)
[16:17:01] <meffie> 14626 is not as urgent, but it should go in too.
[16:17:56] <kaduk@jabber.openafs.org/barnowl> Hmm, presumably the usualy brokenness of DMARC for mailing lists :(
[16:18:50] <kaduk@jabber.openafs.org/barnowl> I think that upstream mailman has added some workaround (sender
rewriting) by now, but we may need to ask the gco admins to upgrade
[16:21:48] <wiesand> No objections to 14650 and 14626 in general. Would those require a pre2? Any potential new DOS issues due to those?
[16:23:01] <kaduk@jabber.openafs.org/barnowl> I think they are isolated enough to not need pre2.
For DoS, hmm, what does the XDR say...
[16:23:01] <meffie> i dont think so. they just avoid overwriting the stack with a nul char.
[16:24:03] <meffie> they dont change the xdr, it will still be limited to 257 bytes on the wire.
[16:24:12] <kaduk@jabber.openafs.org/barnowl> Yes, what Mike said
[16:26:41] <kaduk@jabber.openafs.org/barnowl> Slight complication w.r.t. RT 135238 ... there is a tiny merge
conflict due to 1d2a1002bd1bc8d82c05399c06836ede83f9eeea
[16:26:41] <meffie> overwriting the stack on the bos client program that is, not the bosserver...
[16:27:25] <kaduk@jabber.openafs.org/barnowl> But 1d2a1002bd1bc8d82c05399c06836ede83f9eeea seems to cherry-pick
cleanly to 1.8, so we could just take both of them, I think.
[16:27:52] <meffie> cool
[16:28:43] <wiesand> Not really cool a few days before the intended release date and in between pre1 and GA…
[16:29:37] <kaduk@jabber.openafs.org/barnowl> which one is the "not really cool" part?
[16:31:46] <wiesand> still digesting 1d2a1002bd1bc8d82c05399c06836ede83f9eeea…
[16:33:52] <kaduk@jabber.openafs.org/barnowl> I confirmed that the cherry-picked version does build (that is, that
the functions marked 'static' are indeed not used outside the file)
[16:34:49] <wiesand> ok looks like it's harmless (and we should have pulled it up two years ago)
[16:35:01] <kaduk@jabber.openafs.org/barnowl> Indeed.
[16:35:11] <kaduk@jabber.openafs.org/barnowl> Okay, I will go ahead and push the master version to gerrit
[16:36:03] <meffie> a search confirms as well
[16:39:26] <wiesand> Just to make sure I get this right: this is about the 24f95214ac46be2bbb39a3ecd0bf62b4fa490413 from Andrew showing up in an attachment to RT #135238, plus 14626 and 14650?
[16:42:24] <kaduk@jabber.openafs.org/barnowl> plus 13806 (style in afs_tokens.c)
[16:42:33] <wiesand> Hmm, I wonder whether the silence is a good sign here...
[16:42:35] <kaduk@jabber.openafs.org/barnowl> Those four would be the 1.8.8, is my understanding
[16:43:07] <wiesand> Ok, thanks.
[16:46:10] <Cheyenne> on a linux 5.13-rc note .. master and 1.8.x build cleanly on intel
[16:46:26] <meffie> (sorry was looking at gerrit)
[16:48:41] <wiesand> Thanks Cheyenne. I guess Linux 5.13 ETA is sunday? Since 1.8.7 is currently broken for current Linux kernels anyway, I think a few days to accommodate the fixes discussed above is fine.
[16:50:36] <kaduk@jabber.openafs.org/barnowl> So I guess this is the part of the meeting where I have to remember my
topics for master/1.9?
[16:51:15] <kaduk@jabber.openafs.org/barnowl> (Though I think at least a couple of them have already come up :)
[16:51:21] <Cheyenne> I pushed a rebase for the linux_env cleanup
[16:51:40] <Cheyenne> (cleanup-LINUX_ENV topic)
[16:51:46] <kaduk@jabber.openafs.org/barnowl> I saw -- thanks!
[16:52:08] <kaduk@jabber.openafs.org/barnowl> I guess I also wanted to ask if anyone had further workshop-followup
after having the week to dig into things a bit more
[16:52:38] <mvita2> I passed the ubik app cache question to Marcio
[16:52:45] <mvita2> he understands it better
[16:53:01] <kaduk@jabber.openafs.org/barnowl> Makes sense; thanks
[16:53:33] <kaduk@jabber.openafs.org/barnowl> I seem to have https://gerrit.openafs.org/#/c/6895/ open in my browser
but don't remember exactly why ... presumably it is something that I
should be reviewing.
[16:54:05] <mvita2> Oh, I have seen this many times in testing
[16:54:37] <mvita2> I should probably apply that patch and test it
[16:54:49] <kaduk@jabber.openafs.org/barnowl> That seems like it would be useful to try, yes
[16:55:12] <mvita2> I don't think I knew there was a fix for it
[16:56:01] <wiesand> Is gerrit 6895 related to RT #135240?
[16:56:54] <kaduk@jabber.openafs.org/barnowl> They don't seem obviously related on first look
[16:57:21] <mvita2> No permission to view ticket
[16:57:21] <kaduk@jabber.openafs.org/barnowl> Or rather, I think they are unrelated, but would have to actually
review the gerrit change in order to assert that will full confidence.
[16:57:51] <kaduk@jabber.openafs.org/barnowl> Mark: that's expected :-/
[16:58:00] <mvita2> <nod>
[16:59:19] <kaduk@jabber.openafs.org/barnowl> I think the only other item I had was in Andrew's stack in gerrit
leading up to rxgk-phase2 ... starting at 14318 there are some test
framework improvements, and I left some review comments there.
Any thoughts on whether I should wait for a response from Andrew vs
pushing my own thoughts on the changes?
[17:00:24] <meffie> (fyi, mvita2 dropped due to power outage)
[17:01:26] <meffie> i'm not sure if we should wait or not. i've been basing things on the changes in gerrit, so it would be good to have them.
[17:01:48] <meffie> changes in gerrit -> the changes for tests:
[17:03:53] <kaduk@jabber.openafs.org/barnowl> I'm also not sure how soon I would actually get a chance to make my
proposed changes ... so maybe I will just plan to push a revised
version if one does not already exist, at the time when I have the
time to do so.
[17:04:01] mvita2 leaves the room
[17:04:01] <wiesand> power outage?!
[17:04:35] <kaduk@jabber.openafs.org/barnowl> (We also hit the end of the hour; I don't know if anyone had to drop
for another meeting)
[17:06:23] <meffie> ok, i can ping andrew about the tests: gerrits tomorrow
[17:07:20] <meffie> motion to adjourn ?
[17:07:30] <wiesand> I see 14651 is up there - will pull it up once it's merged unless someone beats me to it.
[17:07:45] <wiesand> Motion to adjourn sustained.
[17:09:16] <wiesand> Thanks a lot everybody! Hope we'll get 1.8.8 out during next weeks meeting or even before.
[17:11:02] <kaduk@jabber.openafs.org/barnowl> Thanks everyone!
[17:11:07] <kaduk@jabber.openafs.org/barnowl> Looking forward to 1.8.8 :)
[17:11:56] meffie leaves the room
[17:14:05] wiesand leaves the room
[17:16:23] <Cheyenne> ben, if there are still problems with the forcegerritrebuild, could you email me any errors from the buildbot master log (probably in the httpd logs)
[19:02:07] mvita2 joins the room
[19:25:48] meffie joins the room
[19:26:33] <meffie> kaduk@jabber.openafs.org/barnowl: looks like the new host is trying to build all the gerrits it missed since it was put on stand by last year?
[19:26:57] <meffie> if so, i might just try to flush the database..
[19:45:01] <meffie> flushed... force gerrit build seems to be working now...
[19:49:23] meffie leaves the room
[21:23:21] mbarbosa leaves the room
[21:56:29] <kaduk@jabber.openafs.org/barnowl> Thanks, meffie.
[22:30:17] Cheyenne leaves the room