- Wednesday, May 21, 2014

release-team@conference.openafs.org

Wednesday, May 21, 2014< ^ >

Room Configuration

Room Occupants

GMT+0
[12:39:38] wiesand joins the room
[12:40:22] <wiesand> test
[13:40:29] meffie joins the room
[13:57:57] kaduk joins the room
[13:58:57] <kaduk> I guess I can poke people about grand.
[13:59:38] <shadow@gmail.com/barnowl7628413F> oh, that it's out of space?
[14:00:51] <wiesand> I was wondering whether we should move volumes and remove additional ro replicas from grand.
[14:00:55] <kaduk> I took Stephan at his word, yes.
[14:01:15] <wiesand> Right now, it's not feasible to add another replica for a release there.
[14:01:32] <shadow@gmail.com/barnowl7628413F> we could unreplicate old releases as a stopgap
[14:01:51] deason joins the room
[14:02:14] <wiesand> most of the older rw's are there though, so we'd need to move a couple too
[14:02:33] <wiesand> but I'm not sure how important the second replica is
[14:04:04] <wiesand> Ok, shall we get 1.6.8 out the door today?
[14:04:22] <kaduk> I skimmed through the draft announcement and have nothing to say about it.
[14:04:45] <shadow@gmail.com/barnowl7628413F> not even that it looks ok? ;)
[14:05:03] <wiesand> *that* bad?
[14:05:24] <kaduk> shadow: :)
[14:06:01] <deason> looks fine to me
[14:06:33] <deason> er, maybe it's worth mentioning that the two known issues are not new?
[14:07:20] <wiesand> "turned out not to be regressions introduced in this release" too weak?
[14:09:24] <shadow@gmail.com/barnowl7628413F> well, when you send the announcement i will approve it, and we will be
done
[14:09:55] <deason> I somehow completely skipped over that phrase
[14:09:56] <deason> okay, nevermind
[14:10:13] <wiesand> Sent.
[14:11:04] <shadow@gmail.com/barnowl7628413F> not there yet
[14:11:43] <wiesand> now it is
[14:12:02] <wiesand> (got the "awaits approval" reply)
[14:13:10] <wiesand> Just received it. Web stuff looks good too. Thanks!
[14:13:52] <wiesand> After the release is before the release. So on to 1.6.9 :)
[14:14:45] <meffie> yay
[14:14:54] <wiesand> Christof brought up 11092. Is that considered suitable for 1.6?
[14:15:31] <meffie> it makes sense to me (but i've not actually tried it)
[14:15:34] <deason> yes
[14:16:21] <kaduk> I don't immediately see a reason why it wouldn't be.
[14:16:56] <wiesand> Ok, so that's on the tentative list too. Anything else, besides what was already discussed last week?
[14:17:23] <kaduk> While I was putting together the 1.6.8 freebsd packages, I noticed that we never pulled up the --disable-kauth configure option to 1.6.
[14:17:45] <deason> mike sent a list; not sure if it anything needs discussion here?
[14:18:33] <meffie> mostly janitorial stuff.
[14:19:14] <kaduk> I do still have one local patch in the freebsd packaging, but I'm not sure whether it's actually needed (something supposed to be a parallel build fix for venus/test, IIRC)
[14:20:40] <wiesand> disable-kauth: which one is it?
[14:20:48] <wiesand> build fix: 1.6-only?
[14:21:11] <deason> parallel make fixes would be really nice to get in...
[14:21:31] <meffie> the disable kauth effects the rpm spec, iirc
[14:21:35] <wiesand> Unless they break more than they fix, which has happened.
[14:21:59] <wiesand> But that's not an objection.
[14:21:59] <deason> maybe I should phrase that like "it would be really nice to fix parallel make" :)
[14:22:10] <meffie> yes :)
[14:22:22] <kaduk> disable-kauth was 8457
[14:22:58] <meffie> i think the current patchsets for parallel make are much less disruptive now.
[14:22:58] <kaduk> It has no effect on anything unless you change how you invoke configure.
[14:23:02] <wiesand> Yes, Mike's work on the parallel build system is another topic. But I recall Jeffrey objecting to pulling that into -stable.
[14:23:36] <meffie> these mostly just add the missing deps
[14:23:52] <kaduk> I don't remember whether I ever actually submitted this venus/test patch into gerrit.
[14:23:58] <meffie> and run compile_et separately for headers and c
[14:24:40] <wiesand> Mike: are they merged on master yet?
[14:24:44] <meffie> (for master, i'd imagine we want something that actually generates accurate deps)
[14:25:31] <kaduk> (The venus patch thingy is available at http://svnweb.freebsd.org/ports/head/net/openafs/files/patch-Makefile.in)
[14:25:41] <meffie> 5 are still in gerrit for master
[14:25:52] <wiesand> 8457 looks harmless
[14:27:08] <wiesand> I think we want 11128. Ill pull it up.
[14:27:16] <kaduk> I guess to do the kauth thing "properly" we should also pull in the commit to not install manpages, either, which is (looking)...
[14:27:38] <wiesand> Right.
[14:27:57] <kaduk> ...10993, and is not yet merged to master
[14:28:41] <meffie> > 8457
i think you'll need to change the config options in the rpm spec, if i'm remembering correctly, looking for the gerrit ...
[14:29:36] <wiesand> I don't think we'll stop packaging them in the 1.6 series?
[14:30:19] <kaduk> Right, I just want people to have the option to disable it if they want.
[14:30:44] <kaduk> (I already disable it for freebsd, but no longer need to carry the patch to do so, because of some features of the ports system.)
[14:31:21] <meffie> yes, you need to update the configure options to build them, or the rpmbuild fails
[14:31:43] <wiesand> I'm too dumb to get the svnweb to show me a diff :-(
[14:32:08] <meffie> (or i'm just remembering wrong.)
[14:32:18] <kaduk> wiesand: "diff to previous 300897" has a link
[14:32:50] <deason> the file _is_ a diff, so you want http://svnweb.freebsd.org/ports/head/net/openafs/files/patch-Makefile.in?revision=340725&view=co yes?
[14:32:51] <wiesand> Which shows me " - No changes -"
[14:32:52] <kaduk> meffie: I disbelieve that the rpmbuild must know about all options that could possibly be given to configure; no one in their right mind would write a system that broken.
[14:33:11] <deason> meffie: it wouldn't need to change if the default behavior doesn't change
[14:33:23] <kaduk> I'm sorry that I can't send better links; my mouse seems to be not capable of clicking in the VM that's running pidgin, so I am keyboard-only.
[14:33:41] <wiesand> That's what I think, but I'll keep it in mind. No big deal if it is a problem.
[14:34:07] <meffie> yep.
[14:34:53] <meffie> thought we changed the defaults on master for kauth.
[14:35:05] <kaduk> Oh, I see, the "diff to previous" is for a commit that only changes svn:merginfo, so there is no content.
[14:35:06] <meffie> the configure defaults, that is.
[14:35:11] <kaduk> Just "view" the file, yes.
[14:35:20] <wiesand> venus build fix: Thanks Andrew. Looks innocious. But has to go through master if it hasn't already.
[14:35:25] <kaduk> We did change the kauth configure defaults on master.
[14:35:38] <deason> so don't change the defaults on 1.6; done
[14:35:45] <kaduk> in 10962, the kauth default changed on master
[14:35:47] <wiesand> Makes sense.
[14:36:05] <meffie> ok
[14:36:15] <kaduk> I guess I can look at that venus thing and whether master needs it or not
[14:39:59] <wiesand> Ok, so that's the list.
[14:40:12] <meffie> (your link worked for me, i see the diff)
[14:44:53] <wiesand> Obviously, things with a high urgency:complexitiy ratio can be added later. But for the time being, I think we should work on getting merged what's already pulled up and pulling up what has been discussed. Quite a list already anyway.
[14:45:27] <kaduk> But no shiny new features? ;)
[14:45:46] <meffie> he he
[14:46:07] <wiesand> 9020 is a feature?
[14:46:49] <meffie> well, andrew says it should be just how it works, or something :)
[14:47:05] <wiesand> What shiny new features are on offer?
[14:47:40] <kaduk> I guess I should send my patch to make the freebsd disk cache less broken to master.
[14:48:21] <meffie> i think all the volscan patches and man pages have been merged on master.
[14:50:11] <wiesand> Man pages usually don't need much discussion. I try to catch those, but when I fail just pull them up.
[14:50:23] <wiesand> I haven't been following volscan lately.
[14:50:46] <meffie> that's a lot of changes to volinfo tho. but that's a debugging tool.
[14:52:16] <wiesand> volscan is not on 1.6 yet is it?
[14:53:02] <meffie> no, just on master
[14:53:11] <jhutz@jis.mit.edu/owl> > most of the older rw's are there though, so we'd need to move a couple
I'm not sure why any RW's are there, except maybe that they were moved
while the other server was upgraded. Feel free to move RW's to
penn-central.
[14:53:41] <wiesand> jhutz: will do, thanks
[14:54:52] <wiesand> volscan would be a nice feature, if it really doesn't touch upon more than volinfo
[14:55:25] <meffie> ok, i'll push the patchsets to 1.6.x for consideration.
[14:56:12] <wiesand> No hurry though. There's plenty to merge and rebase right now. But we can consider volscan.
[14:56:35] <meffie> thank you
[14:57:01] <wiesand> I'd like to have it on my systems ;-)
[14:57:24] <kaduk> wiesand: you mean, you don't run master on your systems ;)
[14:57:41] <meffie> tsk tsk
[14:57:43] <wiesand> yes, shame on me ;-)
[14:59:21] <wiesand> Ok. Anything else to discuss today?
[14:59:34] <kaduk> I have something.
[14:59:46] <wiesand> Ok
[14:59:48] <kaduk> A big thanks to Stephan for running another successful release!
[15:00:30] <wiesand> No, huge thanks to all of you.
[15:01:57] <meffie> yes, i second that motion, thank you wiesand.
[15:03:34] <wiesand> Fine. I think we're done for today.
[15:03:36] <meffie> oh, are the token cache fixes slated for 1.6.9?
[15:03:59] <wiesand> Simon brought them up last week I believe?
[15:04:03] <meffie> yes
[15:04:53] <wiesand> So they are, if ready in time, review is postive etc. as usual.
[15:05:05] <meffie> ok, thanks.
[15:05:50] <wiesand> Fine. I think we're done for today?
[15:05:54] <meffie> have a good evening.
[15:06:12] <kaduk> Yup, thanks.
[15:06:14] <wiesand> Will try to get out the minutes a bit more timely this time..
[15:06:21] <wiesand> Thanks a lot everyone!
[15:06:30] wiesand leaves the room
[15:06:46] meffie leaves the room
[15:09:13] deason leaves the room
[16:09:42] deason joins the room
[16:10:09] <deason> the 1.6.8 release pages currently contain a bunch of nonfunctioning links; see 11157
[16:10:09] wiesand leaves the room
[16:10:09] wiesand leaves the room
[16:25:12] stephan.wiesand joins the room
[16:27:28] <stephan.wiesand> Yeah, I screwed up :-( Pulled the make_www_release command from history, and thus invoked it with -d which allows for testing the page set locally. Obviously, the result looked fine when I checked it before submitting. Thanks for catching and fixing this so quickly.
[17:47:41] shadow@gmail.com/barnowl7628413F joins the room
[17:50:38] shadow@gmail.com/barnowl7628413F leaves the room
[18:44:10] stephan.wiesand leaves the room
[19:27:53] shadow@gmail.com/barnowl7628413F joins the room
[19:30:42] shadow@gmail.com/barnowl7628413F leaves the room
[19:32:18] deason leaves the room
[19:32:18] deason joins the room
[21:37:43] kaduk leaves the room