- Friday, June 12, 2020

release-team@conference.openafs.org

Friday, June 12, 2020< ^ >

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

Room Configuration

Room Occupants

GMT+0
[12:25:03] mbarbosa joins the room
[15:53:08] meffie joins the room
[15:57:50] yadayada joins the room
[15:59:20] cwills joins the room
[16:00:11] <meffie> greetings.
[16:00:18] <yadayada> Hi All
[16:00:21] <kaduk@jabber.openafs.org/barnowl> salutations
[16:00:42] <meffie> mark and cwills are occupied at the moment.
[16:00:56] mvita2 leaves the room
[16:01:33] <kaduk@jabber.openafs.org/barnowl> I have been a little sick for most of the week and had a hard time
just keeping up with everything for work+IETF; I basically didn't get
anything done for openafs.
[16:01:56] <meffie> sorry to hear, i hope you get well soon.
[16:02:16] <yadayada> yeah get well soon Ben
[16:03:21] wiesand joins the room
[16:03:39] <wiesand> Hello
[16:04:20] <kaduk@jabber.openafs.org/barnowl> Thanks!  I think my trend is improving, at this point
[16:04:45] <wiesand> Yes, get well soon. Hope it's nothing seiours?
[16:04:50] <meffie> good to hear
[16:05:05] <meffie> (about the improving that is)
[16:05:58] <wiesand> I'm a bit under the weather too today. Nothing serious, just the terrible air and atmosphere right before a thunderstorm and the rapidly falling air pressure…
[16:06:06] <kaduk@jabber.openafs.org/barnowl> Nothing serious, yes -- just a sore throat, and an occasional need for
a nap
[16:06:24] <wiesand> Sure it's not Corona?
[16:07:04] <kaduk@jabber.openafs.org/barnowl> Pretty sure
[16:07:25] <kaduk@jabber.openafs.org/barnowl> But I'm not leaving the house much either way
[16:08:33] <wiesand> That's not so much what I had in mind…
[16:09:45] <wiesand> These are bad times to seek treatment, but if you develop a cough or fever you probably should.
[16:10:31] <wiesand> So, I finally smoke tested and merged the changes foreseen for 1.8.6pre3
[16:10:44] <meffie> thank you
[16:10:46] <wiesand> I also updated 14224
[16:11:05] <wiesand> Some review of that and 14225 would be appreciated
[16:11:53] <wiesand> (and I rebased 14217, I think buildbot is still chewing on it)
[16:14:25] <wiesand> Cheyenne provided some useful feedback on 14240. I wonder whether there's more?
[16:15:29] <wiesand> NB if anyone (in particular a native speaker) feels like picking up the ball and taking it home please feel free to do so.
[16:16:40] <kaduk@jabber.openafs.org/barnowl> I have 14240 open in a tab to get back to
[16:18:34] <wiesand> pushed PS3 for 14224 - thanks
[16:21:07] <kaduk@jabber.openafs.org/barnowl> 14224 PS#3 looks good
[16:22:39] <kaduk@jabber.openafs.org/barnowl> Yadav, did you have any topics you wanted to bring up?  Otherwise we
may be looking at a short meeting today.
[16:22:50] <wiesand> 14217 buildbot failures look genuine?
[16:23:27] <yadayada> just one thing Ben, relating to 14239
[16:24:22] <meffie> re 14217, looks like yes
[16:24:35] <meffie> afsmonitor.c:3898:26: error: ‘terminationEvent’ undeclared (first use in this function)  code = LWP_WaitProcess(&terminationEvent);
[16:24:37] <yadayada> can you please review it. I know Andrew has suggested to use pthread_at_fork, but looks that might affect all fork. kindly let us know your thoughts and based on that we can move forward
[16:25:11] <meffie> but i though there was a prereq to fix that error? maybe just a rebase is needed?
[16:25:28] <meffie> i will look after the meeting.
[16:25:39] <kaduk@jabber.openafs.org/barnowl> I have 14239 open in a tab to look at today (just like 14240).
Sorry I did not do anything with it sooner.
[16:26:18] <meffie> i would like to see if we can use pthread_atfork for a permanent solution. i've started looking into that.
[16:27:00] <yadayada> ok, Thanks Mike
[16:27:08] <wiesand> NB I stumbled over this in the (public) Auristor release notes:
[16:27:11] <meffie> i'll propose that as an alternative, if that is ok?
[16:27:13] <wiesand> '''While debugging a Linux SIGBUS issue, it was observed that receipt of an ICMP network error in response to a transmitted packet could result in termination of an unrelated rx call and could mark a server down. If the terminated call is a StoreData RPC, permanent data loss will occur. All Linux clients derived from the IBM AFS code base experience this bug. The v0.195 release prevents this behavior.
[16:27:28] <kaduk@jabber.openafs.org/barnowl> For a while OpenSSL had a pthread_atfork() handler in the tree to
cause the RNG to diverge between parent and child, but we had to
revert it.  I need to look up what the issue was that made us revert
the pthread_atfork() usage in OpenSSL.
[16:28:22] <meffie> ok, hopefully our logging library is much simpler
[16:28:24] <kaduk@jabber.openafs.org/barnowl> meffie: pushing an alternative approach in a different gerrit number
is always useful
[16:28:50] <meffie> ok, thank you.
[16:29:57] <kaduk@jabber.openafs.org/barnowl> wiesand: an interesting thing to see; thanks for pointing it out
[16:30:02] <meffie> (it could be possible to have a targeted fix for stable, and a longer term change to avoid the general class of problems)
[16:31:14] <kaduk@jabber.openafs.org/barnowl> re 14217 and terminationEvent, 144217 itself is renaming
terminationEvent to be cm- or fs-specific.  So I think that no amount
of rebasing will help
[16:31:35] <kaduk@jabber.openafs.org/barnowl> (s/44/4/; this new keyboard is a little enthusiastic)
[16:32:32] <meffie> cwills will need to look closer then.
[16:33:55] <wiesand> Something is weird about buildbot's response to 14224 PS3
[16:34:44] <kaduk@jabber.openafs.org/barnowl> PS3?  It's only 15 minutes old.
[16:35:08] <meffie> (i still think 14217 ps2 is not based correctly)
[16:36:33] <meffie> openafs-stable-1_8_x/14224
[16:36:52] <wiesand> 14217 is something else, and quite probably my fault
[16:36:55] <meffie> that windows builders are currently building 14224
[16:37:05] <kaduk@jabber.openafs.org/barnowl> I see what meffie sees
[16:37:24] <kaduk@jabber.openafs.org/barnowl> (I mean, as far as what buildbot's doing.  Nothing creepy.)
[16:37:44] <meffie> HA!
[16:37:55] <wiesand> ah, ok… did I mention that I'm not exaclty on top of things today?
[16:38:13] <kaduk@jabber.openafs.org/barnowl> BTW, meffie, there was a weird thing in the buildbot log for the
solaris10_x86 builder build of 14149
[16:38:58] <kaduk@jabber.openafs.org/barnowl> Maybe it was just a one-off fluke, but could be worth checking to see
how that builder is doing if it happens again
[16:39:37] <kaduk@jabber.openafs.org/barnowl> Anyway, it sounds like we have a few things to do but not much left to
talk about.  Motion to adjourn?
[16:39:49] <yadayada> nothing from my side
[16:40:03] <meffie> i'll look, thanks!
[16:40:39] <wiesand> rebased 14217 on top of 14226, hopefully that will help
[16:41:43] <meffie> nothing else here.
[16:42:00] <wiesand> I intend to ask Ben for the 1.8.6pre3 tag once 14224/5 are merged, which should be soon. Any objections?
[16:42:15] mbarbosa leaves the room
[16:42:29] <meffie> none here, thank you
[16:42:33] <kaduk@jabber.openafs.org/barnowl> Good eye, wiesand; that does actually look likely to help.  Sorry for
my earlier confusion.
[16:42:37] mbarbosa joins the room
[16:43:27] <kaduk@jabber.openafs.org/barnowl> I will be on the lookout for a tag request.
[16:43:52] <wiesand> Thanks. But please do let your recovery take precedence!
[16:44:17] <kaduk@jabber.openafs.org/barnowl> Okay.
[16:44:21] <kaduk@jabber.openafs.org/barnowl> Thanks everyone!
[16:44:24] <yadayada> Thanks all
[16:44:38] <wiesand> That's all from me today. Let's adjourn. Thanks a lot everyone@
[16:44:44] <wiesand> s/@/!/
[16:45:33] wiesand leaves the room
[16:47:02] meffie leaves the room
[17:19:52] yadayada joins the room
[17:29:00] yadayada leaves the room
[20:14:13] yadayada leaves the room
[21:46:07] mbarbosa leaves the room