- Wednesday, December 13, 2017

release-team@conference.openafs.org

Wednesday, December 13, 2017< ^ >

Room Configuration

Room Occupants

GMT+0
[03:01:04] mvita leaves the room
[03:09:58] mvita joins the room
[03:58:35] mvita leaves the room
[04:16:25] mvita joins the room
[05:03:28] mvita leaves the room
[13:26:51] meffie joins the room
[14:28:15] mvita joins the room
[15:56:40] wiesand joins the room
[15:58:04] <meffie> welcome back to jabber wiesand
[15:58:44] <wiesand> thanks mike
[15:58:54] <mvita> hail, hail, the gang's all here
[15:59:11] <wiesand> well we know ben has a parallel meeting
[15:59:30] <wiesand> so maybe we should start with dicussing the new time slot?
[15:59:37] <mvita> agreed
[15:59:41] <meffie> ok, good idea
[15:59:47] <wiesand> Looks like we have three slots all participants like. In GMT:
1) Tue    3:30 pm
2) Fri    2:00 pm
3) Fri    3:30 pm
My personal favourite is (2), followed by (3).
[15:59:57] <wiesand> (from my mail to release-team)
[16:00:23] <wiesand> is everyone fine with Fri 2:00pm GMT?
[16:00:42] <meffie> digging up a timezone converter
[16:00:45] <mvita> yes
[16:00:52] <mvita> -5 meffie
[16:01:04] <mvita> 9:00 am EST
[16:01:37] <meffie> ah, thanks. yes that time if fine.
[16:01:55] <meffie> *is
[16:02:08] <mvita> fine by me as well
[16:02:19] <wiesand> So, agreed. 2pm GMT Friday it is.
[16:02:39] <kaduk@jabber.openafs.org/barnowl> SGTM
[16:02:50] <wiesand> Shall we start this friday already?
[16:02:53] <wiesand> SGTM?
[16:03:00] <mvita> sounds good to me
[16:03:11] <wiesand> thanks…
[16:03:55] <wiesand> So, next meeting Friday the 15th?
[16:04:02] <mvita> okey
[16:04:15] <wiesand> my birthday btw...
[16:04:34] <mvita> Then we must have cake!
[16:05:02] <wiesand> np, but difficult to share over jabber
[16:05:27] <wiesand> Anyway, let's keep today's meeting short then.
[16:05:36] <wiesand> Regarding 1.6.x, what net?
[16:05:40] <wiesand> next
[16:05:48] <wiesand> 1.6.23 or 1.6.22.1?
[16:06:19] <mvita> probably 1.6.22.1 if I can finish getcwd changes requested by Andrew.
[16:06:45] <wiesand> ah, I was about to ask how close we are
[16:06:46] <mvita> I've been tied up with reactions to the security release
[16:07:14] <wiesand> Me too, and other stuff just as enjoyable…
[16:07:20] <kaduk@jabber.openafs.org/barnowl> Not too bad, I hope
[16:07:38] <kaduk@jabber.openafs.org/barnowl> (the reaction to the security release, I mean)
[16:07:49] <wiesand> No no.
[16:08:11] <mvita> would it be a breach of protocol to request a heads-up when a security release is in the wind?   No details need be shared… but it does affect my schedule a lot.
[16:08:59] <kaduk@jabber.openafs.org/barnowl> I think the heads-up comes in the form of an RT ticket in the security
queue ... for this particular one the release came a day after the
initial notification, since there were external forcing factors
[16:09:08] <wiesand> I felt a bit upset when I first got the news. But only for 5 minutes. The I realized that in this case there wasn't much of a head start possible.
[16:09:56] <wiesand> And after all, it was already done ;-)
[16:10:09] <meffie> yes
[16:10:54] <mvita> what's the RT number?
[16:11:56] <mvita> I'm not seeing it.
[16:12:15] <kaduk@jabber.openafs.org/barnowl> 134450
[16:12:34] <mvita> No permission to view ticket
[16:12:39] <mvita> <sigh>
[16:12:52] <kaduk@jabber.openafs.org/barnowl> That's odd.
[16:12:55] <meffie> are you logged in as "guest"?
[16:13:00] <kaduk@jabber.openafs.org/barnowl> I thought we gave you access
[16:13:03] <mvita> no, as mvitale
[16:13:10] <mvita> I thought so too.
[16:16:20] <wiesand> btw I can read the tickets in the queue, but I'm not a watcher and don't check for new tickets daily either
[16:16:39] <mvita> oh well, let's move on for now.
[16:16:57] <wiesand> looking into rt…
[16:17:16] <wiesand> seems like F27 updated to Linux 4.14
[16:17:30] <wiesand> so it's really high time for that 1.6.22.1 release
[16:18:47] <kaduk@jabber.openafs.org/barnowl> debian also took 4.14
[16:20:18] <wiesand> Ok. Let's wait till Friday and see where we are regarding EL7.4 getcwd. Worst case, issue a 1.6.22.1 with 4.14 support only.
[16:20:40] <meffie> ok, thanks.
[16:22:02] <wiesand> That .1 and probably a .2 should soak up the resources we have to spend on 1.6 for the rest of the year :-(
[16:22:13] <kaduk@jabber.openafs.org/barnowl> How worried are we about dropping the lock during iteration?
[16:22:48] <wiesand> Sounds like a major worry to me.
[16:23:49] <wiesand> We should either be sure it's no problem or have sufficient testing.
[16:24:47] <mvita> oh, sorry, got distracted by RT - catching up…
[16:25:00] <wiesand> Rushing out a "stable" release that introduces a new, possibly more severe problem due to that isn't helpful.
[16:25:54] <kaduk@jabber.openafs.org/barnowl> We could attempt to kick the can by only using the new code on
"sufficiently new" environments
[16:26:02] <kaduk@jabber.openafs.org/barnowl> But I'm not sure that I'm terribly excited about that route
[16:26:05] <mvita> the next patchset will not drop the lock _as often_ - but we will still drop it once in a while - so Andrew's suggestions have merit and I will address them.
[16:26:42] <mvita> "sufficiently new" is exactly what is so difficult to determine accurately
[16:27:36] <wiesand> I'm not going to propose a configure switch…
[16:27:50] <meffie> heh
[16:28:16] <mvita> dropping the lock is not dangerous - it's not restarting iteration after it's been dropped that can cause problems.
[16:29:00] <wiesand> Though I could deal with it, the SL packaging could, …
[16:30:26] <kaduk@jabber.openafs.org/barnowl> I mean, we think that this is what we ought to be doing always, so at
some point we ought to bite the bullet and ... always do it.
[16:33:06] <meffie> yes
[16:33:58] <meffie> seems we need some prerelease testing to be more confident
[16:34:18] <meffie> instead of trying to rush out a 1.6.22.1
[16:34:51] <mvita> the original reporter is willing to do some additional testing
[16:35:09] <mvita> Ben replied to him yesterday, I will reply to him today.
[16:35:44] <meffie> great, thanks. we should find people willing to test  prerelease on older kernels too.
[16:36:04] <mvita> yes, I think that's what he was offering, to try an older kernel
[16:36:07] <mvita> IIRC
[16:36:13] <meffie> oh, excellent!
[16:37:10] <wiesand> I can offer EL6 and EL7.3
[16:37:25] <meffie> nice
[16:37:33] <wiesand> if and when I get around to it…
[16:38:36] <wiesand> on to 1.8? master?
[16:39:05] <mvita> Here you go, Stephan:   http://bytesdaily.blogspot.com/2011/11/round-tuit.html
[16:39:15] <meffie> so, is the target for getcwd fixes then 1.6.23pre1 ?
[16:39:25] <meffie> for next week?
[16:40:22] <wiesand> mike: it depends
[16:40:31] <kaduk@jabber.openafs.org/barnowl> For 1.8 I think we want the RHEL7.4 getcwd thing and another ... couple
weeks' test period?  I don't think there is much anything else to wait
for, specifically.
[16:40:43] <meffie> just trying to get my notes straight.
[16:40:49] <kaduk@jabber.openafs.org/barnowl> I am probably going to put 1.8 into debian unstable over the holiday
break
[16:41:02] <wiesand> whenever we're confident it won't introduce a severe regression, the getcwd fix could go out
[16:41:43] <wiesand> that could be for 1.6.22.1 this week, 1.6.22.2 a bit later, or after a 1.6.23pre1 testing phase
[16:42:43] <meffie> ok, thanks
[16:43:08] <meffie> great news ben! thanks
[16:43:17] <wiesand> That's my point of view at least.
[16:43:47] <wiesand> Yes, good news re 1.8 in debian unstable.
[16:45:18] <kaduk@jabber.openafs.org/barnowl> I may also have time over break to actually address Andrew's rxgk
comments
[16:45:31] <wiesand> More great news.
[16:45:47] <meffie> thanks ben, is there anything else (besides getcwd) you need for 1.8.0 ?
[16:45:49] <kaduk@jabber.openafs.org/barnowl> but I don't want to plan on spending too much time for my own stuff,
since we're staying at the in-laws and there will certainly be other
events planned
[16:46:08] <meffie> that we could do that is?
[16:46:12] <kaduk@jabber.openafs.org/barnowl> The only other thing would be the usual "test test test".
I've been running it on my laptop for a while now.
[16:46:30] <kaduk@jabber.openafs.org/barnowl> But I should probably put it on a server I run, too.
[16:46:40] <meffie> ok, we've been running 1.8.0pre2 fileservers since it was released.
[16:46:51] <meffie> (on centos)
[16:47:23] <meffie> i'll ping more people to run pre3
[16:47:51] <wiesand> Are we finished for today? Shall we adjourn till Friday?
[16:48:19] <kaduk@jabber.openafs.org/barnowl> I think so.
[16:48:57] <meffie> ok, i'll get back in the habit of posting notes to openafs-devel
[16:49:11] <meffie> i'll include our new meeting time in the post
[16:49:20] <kaduk@jabber.openafs.org/barnowl> (Oh, and the ticket probably wasn't showing up because I 'resolved' it
already)
[16:49:24] <wiesand> That would help. Thanks a lot Mike!
[16:49:32] <meffie> ben, can people request jabber accounts if they request?
[16:49:36] <kaduk@jabber.openafs.org/barnowl> Thanks, Mike!
[16:49:43] <meffie> er need to join?
[16:49:57] <kaduk@jabber.openafs.org/barnowl> I'm open to giving more people jabber accounts, though reserve the right
to exercise some discretion
[16:50:26] <meffie> yes, we dont want any spammers or such.
[16:50:34] <wiesand> Getting the password to them could be more difficult than in my case though
[16:50:53] <kaduk@jabber.openafs.org/barnowl> Also true.
[16:50:56] <meffie> that's what gpg is for :)
[16:51:06] <kaduk@jabber.openafs.org/barnowl> I should probably clean up after myself there, too.
[16:51:35] <kaduk@jabber.openafs.org/barnowl> Or, I guess you could, too, but that might feel weird.
[16:51:48] <meffie> should we kill the ##openafs-release-team channel, or leave it as a backup?
[16:51:59] <kaduk@jabber.openafs.org/barnowl> I don't really have an opinion
[16:52:25] <mvita> no opinion
[16:52:49] <meffie> ok, i'll kill it. one less place to lurk. thanks for fixing the jabber issues ben!
[16:53:09] <wiesand> Yes thanks a lot!
[16:53:27] <kaduk@jabber.openafs.org/barnowl> You're welcome!  I probably should have done it a lot sooner.
[16:54:08] <wiesand> CU Friday then
[16:54:11] <mvita> Oh, I almost forgot.
[16:54:20] <wiesand> ok?
[16:54:21] <mvita> Meffie fixed the Linux builders
[16:54:29] <meffie> oh, yes.
[16:54:37] <kaduk@jabber.openafs.org/barnowl> yay; thank you, Mike!
[16:54:39] <mvita> they are now failing for real reasons - there are a few changes needed for 4.15
[16:54:43] <mvita> nothing major yet
[16:54:56] <meffie> thanks mark.
[16:55:05] <mvita> I am tracking them and will submit patches after the getcwd stuff is squared away
[16:55:12] <wiesand> Looking forward to success reports then. Thanks!
[16:55:14] <mvita> unless someone else beats me to it
[16:55:38] <meffie> (thanks, mark, more for the notes;)
[16:56:10] <meffie> have a good day, enjoy the snow if you got some
[16:56:40] <wiesand> bye
[16:56:41] wiesand leaves the room
[16:57:50] <kaduk@jabber.openafs.org/barnowl> thanks everyone!
[17:05:26] meffie leaves the room
[18:38:13] meffie joins the room
[19:59:06] meffie leaves the room
[21:42:05] meffie joins the room
[22:16:20] meffie leaves the room
[23:47:48] mvita leaves the room