- Thursday, December 2, 2021

release-team@conference.openafs.org

Thursday, December 2, 2021< ^ >

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

Room Configuration

Room Occupants

GMT+0
[00:01:47] Cheyenne leaves the room
[13:00:59] mbarbosa joins the room
[14:57:59] Cheyenne joins the room
[16:48:43] wiesand joins the room
[16:57:31] meffie joins the room
[16:59:20] <meffie> good day
[17:00:01] <kaduk@jabber.openafs.org/barnowl> greetings
[17:00:18] <meffie> i hope you are feeling better kaduk@jabber.openafs.org/barnowl
[17:00:43] <kaduk@jabber.openafs.org/barnowl> Improving, at least.  Mostly back to normal in the past couple days
[17:01:25] <Cheyenne> Hello all
[17:01:33] <meffie> improving is good
[17:02:21] <wiesand> Sounds kind of serious though. (Hello All)
[17:04:07] <wiesand> Hope you had a nice Thanksgiving (NB why a Thursday?)
[17:04:24] <Cheyenne> Tradition...
[17:04:47] <kaduk@jabber.openafs.org/barnowl> deeply ingrained tradition, yes
[17:06:23] <wiesand> Not much visible progress here the past week, but I just deployed a test build based on the current HEAD of 1.8.x, + 14825 14830 14831 14841 14844 14828 - smoke tests look good. Could you confirm that's probably what we wan to release as 1.8.8.1 a.s.a.p.?
[17:08:09] <meffie> btw fedora 35 is out now and requires 14831 (at least)
[17:09:01] <kaduk@jabber.openafs.org/barnowl> Comparing against the list of open gerrits for 1.8.x that looks about
right.  Remind me why we want 14825 in that list, though?
[17:10:44] <Cheyenne> I think it makes the pull-in for 14831  "cleaner"
[17:10:56] <Cheyenne> (14831 linux 5.15)
[17:11:01] <kaduk@jabber.openafs.org/barnowl> +1
[17:12:36] <wiesand> It removes a potential regression due to 14503 and a path conflict with 14831, and is supposed not to change behaviour in other respects.
[17:13:17] <wiesand> But I'd accept deferring it to 1.8.9
[17:14:26] <kaduk@jabber.openafs.org/barnowl> I'm okay having it in 1.8.8.1; I just couldn't remember why we put it
there.
[17:15:09] <wiesand> It's one of those things which haven't been speeding me up lately…
[17:16:27] <wiesand> But I do think it makes sense to put it in now. If there's no objection, we now have at least a clear path how to proceed.
[17:16:47] <kaduk@jabber.openafs.org/barnowl> sounds good
[17:17:34] <wiesand> :)
[17:18:50] <wiesand> So, next on my todo list are "merge, update NEWS"
[17:20:11] <wiesand> Anything else regarding 1.8.x to discuss today?
[17:20:31] <meffie> i have nothing.
[17:21:30] <Cheyenne> (doing a 5.16 build against 1.8 along with the above list of gerrits)
[17:21:51] <meffie> thanks Cheyenne
[17:23:45] <wiesand> Regarding the issue brought up in [OpenAFS] Fwd: CRITICAL: RHEL7/CentOS7/SL7 client systems - AuriStorFS
v2021.05-10 released > OpenAFS versions? <https://lists.openafs.org/pipermail/openafs-info/2021-November/043164.html> , I guess we just have to wait for a fixed EL7 kernel?
[17:24:12] <Cheyenne> The problem reported to the mailing list concerning RHEL7 doesn't seem to be an issue with openafs.  I found the associated RH bugzilla and went through the openafs code paths.  If the reported condition happens, openafs internally catches it as an EIO
[17:25:00] <wiesand> Ah, I *had* some hope that Andrew's recent work in that are would help…
[17:25:37] <kaduk@jabber.openafs.org/barnowl> Is that something that we should be replying to say on the thread?
[17:26:53] <Cheyenne> (1.8.x + gerrits -- clean build against linux-5.16-rc3+)
[17:28:00] <wiesand> That's a tough one, politically ;-)
[17:28:30] <kaduk@jabber.openafs.org/barnowl> true
[17:29:09] <wiesand> Anyway, thanks a lot Cheyenne for checking up on bot 5.16 and that pending signal issue!
[17:29:49] <Cheyenne> The code in osi_rdwr (which is were the problem would occur).. has been working that way since at least 2006
[17:30:16] <wiesand> Huh?
[17:31:05] <Cheyenne> The code that would be affected by the RHEL7 bug is within osi_rdwr.
[17:32:27] <Cheyenne> and just by chance, the code has been checking for a short read and returning an EIO since at least 2005
[17:32:31] <Cheyenne> 2006 -- sorry
[17:33:46] <wiesand> Could aaa7043a154d35838e65bc28473355c452339bcc possibly make a difference?
[17:34:47] <Cheyenne> No..
[17:34:56] <Cheyenne> different condition.
[17:35:18] <wiesand> (and please forgive me if the idea is completely off and stupid, which is quite possible ;-)
[17:35:52] <Cheyenne> Oh.. no.. not a problem.. I took a look through that as well..
[17:36:38] <Cheyenne> turns out that all reads related to the reported problem go through the osi_rdwr function
[17:39:04] <wiesand> Ok, thanks. I'm leaning towards not picking up a fight on openafs-info. But it's not for me to decide, and I certainly don't want to keep anyone from reporting his findings on that list ;-)
[17:39:33] <Cheyenne> understand
[17:40:44] <wiesand> Maybe a private mail to Giovanni would be a good idea…
[17:42:35] <meffie> did anyone see my mail about xmpp auth?
[17:43:10] <Cheyenne> I did find one minor bug while digging through the code -- fixed with gerrit 14848 -- very specific edge case
[17:43:19] <wiesand> [mumbling] yes…
[17:43:22] <kaduk@jabber.openafs.org/barnowl> > xmpp auth
Oops, it came in while I was not keeping up with things
[17:43:35] <kaduk@jabber.openafs.org/barnowl> I can create local accounts on jabber.openafs.org for people, I think.
[17:43:37] <meffie> no worries.
[17:43:48] <meffie> excellent news
[17:44:37] <wiesand> I don't have a good answer though. Personally, I'm using a launchpad/google combo to authenticate to gerrit…
[17:45:55] <meffie> i use launchpad for gerrit. but currently using the sinenomine.net xmpp server or jabber logins. that's going away soon.
[17:46:18] <wiesand> ah, and for jabber I think I have a local account there
[17:46:42] <meffie> a jabber.openafs.org account?
[17:47:23] <kaduk@jabber.openafs.org/barnowl> It looks like it, yes.
[17:47:44] <kaduk@jabber.openafs.org/barnowl> We could presumably also consider moving to a different chat
technology for meetings (suggestions welcome)
[17:48:04] <wiesand> probably - I'd have to dig it up, too long ago and since it's been working…
[17:48:39] <meffie> well, anything but slack :) or discord. :)
[17:50:31] <meffie> i thought we liked the automatic logging already set up for jabber. (even though the index is hard to navigate)
[17:50:57] <wiesand> Mike already brought up moving to video (zoom, probably), and yes that might be more contemporary. But I'm a dinosaur here and really like the availability of the transcripts including the gerrit numbers…
[17:51:33] <wiesand> So, frankly, I'm very happy with what we have right now!
[17:51:36] <Cheyenne> IRC ??
[17:51:54] <meffie> (i also have an 8x8 account now for video with jitsi)
[17:52:07] <kaduk@jabber.openafs.org/barnowl> I don't have any particular complaints about what we have now, either;
just wanted to make sure we're not being too narrow-minded.
[17:53:06] <meffie> if it is not too much trouble to create logins on jabber.openafs.org, then i dont mind continuing to us jabber.
[17:53:41] <kaduk@jabber.openafs.org/barnowl> I just have to log in and run a couple commands (if I remember
correctly), which is not a big deal.
[17:54:04] <meffie> excellent news. thank you ben. sorry for the extra stuff to do. :(
[17:55:06] <meffie> if we only had a system that could handle authentication, maybe with tickets...
[17:58:57] <wiesand> like kerberos? ;-)
[17:59:04] <meffie> heh.
[17:59:15] <meffie> motion to adjourn?
[17:59:22] <kaduk@jabber.openafs.org/barnowl> sounds good
[17:59:40] <meffie> hope  you get well soon kaduk.
[18:00:01] <kaduk@jabber.openafs.org/barnowl> (I was busy with IESG stuff this week, and the only thing I did was to
merge 14838 this morning.  Sorry for not being more proactive about
transitioning the meeting topic over to master and adjourning.)
[18:00:07] <kaduk@jabber.openafs.org/barnowl> thanks meffie
[18:00:29] <wiesand> Let's adjourn then. Thanks a lot everyone! And Ben, get well soon.
[18:00:39] <Cheyenne> Thanks for 14838 -- my builds will be quieter
[18:00:57] <Cheyenne> take care everyone :)
[18:43:17] wiesand leaves the room
[20:55:10] meffie leaves the room
[22:04:40] mbarbosa leaves the room
[23:47:23] Cheyenne leaves the room