[01:45:40] --- wiesand has become available [02:04:09] --- Simon Wilkinson has become available [02:27:32] --- Simon Wilkinson has left [02:27:41] --- Simon Wilkinson has become available [02:51:05] --- Simon Wilkinson has left [02:59:16] --- Simon Wilkinson has become available [03:27:18] --- Simon Wilkinson has left [03:28:42] --- Simon Wilkinson has become available [03:29:21] --- Simon Wilkinson has left [03:29:21] --- Simon Wilkinson has become available [03:29:21] --- Simon Wilkinson has left: Lost connection [03:29:28] --- Simon Wilkinson has become available [05:19:11] --- Simon Wilkinson has left [05:21:04] --- Simon Wilkinson has become available [06:34:00] --- Marc Dionne has become available [07:00:46] Hello [07:02:05] mornin'. [07:02:20] greetings [07:02:36] Let's start. Marc: any Linux news? [07:02:54] on the current mainline front, still good as of this morning [07:03:27] Great, thanks. Looks like we'll get away this time. [07:03:31] --- deason has become available [07:03:57] Simon: Just in case you want to leave asap: Anything on 1.6.7? [07:04:56] Meanwhile: Gerrit 10876. [07:05:26] I'm wondering whether I should simply repush it with the FIXES line added? [07:05:51] Or maybe I'll mail Hartmut.. [07:06:42] if you wanna... but it doesn't seem like it needs a lot of concern [07:07:22] But maybe it would help getting it merged so we can pull it up [07:07:48] Any thoughts on https://rt.central.org/rt/Ticket/Display.html?id=131825 ? [07:08:49] Looks really weird. Can this possibly be an OpenAFS problem? [07:09:28] I can ask for debug info, at least, but I don't really know [07:09:55] It could possibly be an OpenAFS problem. I don't really see how, and a hardware issue is much more likely, but I don't think it can be ruled out 100%. [07:10:07] I wonder whether the vicepb hardware just isn't up to random IO. [07:11:00] Last item I had under "problem reports" is what Andrew observed regarding gerrit 10804 [07:11:11] I can confirm this. [07:12:20] I can eventually get to looking into it a bit more and providing a fix, but of course anyone else doing so is greatly appreciated :) [07:12:41] I'd consider this a showstopper for pre1. [07:13:46] So, revert it for the time being? [07:14:55] maybe submit a change to revert it, but only merge if we don't have anything else [07:15:13] that is, by the time we need to release anything [07:15:22] well if you revert you bring back the getcwd ENOENT problem. and with that reverted i can easily make some bad things happen in conjunction wtih fs rm. so maybe a different solution would be best [07:15:51] I feel like reverting it would just make a different set of people sad. Er, ==Marc [07:16:10] It's 4 work days before EAKC. [07:16:51] Sp, probably no pre1 before. No big deal. [07:17:12] reverting it brings back a problem, but as far as we know it's a problem that has been around for a long time [07:17:28] the problem it seems to introduce is new, so it's a regression [07:17:44] The problem that has been around for a long time got a lot worse recently. [07:18:15] (and it may be 4 work days away, but some of us have quite a lot of time on planes ahead of us :) [07:19:25] Yeah. So there's a chance we can bake pre1 in the CERN restaurant next week ;-) [07:20:59] Andrew, thanks for the WIP bind-mount changes. Haven'd had a closer look yet. [07:21:14] Any thoughts on whether this is going in the right direction? [07:21:39] How much work will this need to be finished? [07:22:02] Is it "suitable for a stable release" at all? [07:22:33] not now, no [07:22:45] I don't know how long; it's take... some time [07:22:56] Ah... [07:23:34] Does it build and work as is? [07:23:35] as a direction it seems like a good way to go. unfortunate that is has to be more complex because of kernel gpl restrictions [07:23:50] I wouldn't expect it for 1.6.8 [07:24:00] I don't. [07:24:13] it builds at least on rhel 6.5; it "works", but... well, it has quotes around it [07:24:13] But can I try/test it as it is now? [07:24:24] Ok, answered. [07:24:31] you can try it , but it will print some messages to the log [07:24:48] most of the problems I've been trying to work around involve shutting it down; it tends to work pretty well just as its running [07:24:51] i'll try to give it a spin, but time will be short in the next few weeks with EAKC and all [07:25:32] there are some specific notes and stuff in the last commit in the stack; you might have something to say about that stuff even without looking at the code [07:26:29] I'll try too. But let's move on - this is not for the next stable anyway. [07:26:44] the biggest problem I don't really know how to solve is what to do if a user has created a mountpoint on a dir in afs, and that dir becomes unreachable (a parent dentry has been d_drop'd/d_invalidate'd) [07:27:05] but there are several issues noted in that last commit [07:27:27] the general "scheme" I think works, though, which took some time to figure out, so once we have that, the rest might go a bit more quickly [07:27:29] and yeah sorry, moving on [07:28:03] Ok, but thanks for the additional info. So: "1.6.7 security release" [07:28:48] Asking once more: any news? ETA? Is "the Wednesday after EAKC" realistic? [07:29:14] Alternately, "did the reporter come up with draft text?" [07:29:58] unless Simon Wilkinson responds, I don't know if you're going to get any answers right now [07:30:41] Probably not. Lets keep hoping, and move on... [07:31:16] 1.6.8: anything coming up to include (besides 10804 fix and other things already discussed?) [07:31:41] I don't know that we explicitly discussed the xcode 5.1 changes. [07:32:05] I was just about to beg for review of those. [07:32:30] http://gerrit.openafs.org/#q,status:open+project:openafs+branch:openafs-stable-1_6_x+topic:xcode51,n,z [07:32:51] doesn't seem like they need discussion, though; just need review [07:32:54] Having buildbot automatically verify changes again would be good... [07:33:04] ==deason [07:33:05] did you say you wanted someone to build on older os x for something, though? [07:33:36] It would be nice, but may not be ncessary. Also, I think that someone (CUDave?) was starting to do so. [07:33:58] But, basically, I am only build-testing these, I haven't been run-testing them. [07:34:02] but what was it? that stack on 1.6 for xcode51? [07:34:21] More likely the stack on master. [07:35:20] can you give me a change to build? or just 'master' right now? [07:35:39] master should be fine, for this purpose. [07:36:13] gerrit 10926 was the last commit in the chain targetted for master. [07:36:32] but they're all merged now, so it doesn't really matter, unless we are worried about there having been other regressions. [07:37:32] okay [07:37:48] --- Simon Wilkinson has left [07:38:41] Hmm [07:39:34] --- Simon Wilkinson has become available [07:40:10] Simon? [07:40:41] Oh well. [07:40:46] Anything else? [07:41:26] remind everyone of the proposed schedule in the next couple of weeks? [07:41:44] Ok: [07:41:59] er, I meant "for the next couple of weeks" [07:42:17] on the plane: try to fix/enhance 10804 [07:42:46] Wednesday evening in Restaurant #1: release-team meeting [07:43:09] Try to get pre1 ready there. [07:43:26] Wednesday after EAKC: 1.6.7 release. Maybe. [07:43:52] Anything else depends on pre1 feedback, I think. [07:43:58] okay, but announce that a security release is coming during eakc? [07:44:15] and 1.6.8pre1 is released... during eakc? or just "soon" whenver it's ready [07:44:44] Well, if/when we know the 1.6.7 release date, it should be announced. [07:45:25] If we have a smoke tested solution for the 10804 problem: Wednesday evening tag and upload pre1. [07:46:43] Without a solution for 10804, I see no point in issuing pre1. [07:47:46] i'm testing something that may help [07:47:56] Differing opinions welcome, of course. [07:48:11] Marc: great. [07:49:38] sorry I am late. I had a conflicting meeting. I believe that Simon still wants to target the Wednesday after EAKC for the security release. [07:49:56] Ah, Input... Hello Jeff [07:50:11] You want a minute for reading up? [07:50:45] However, I do not know what the status of write-ups etc are. [07:51:21] I forget, did we agree to a one week pre-announcement for security fixes or two weeks? [07:52:04] Not sure we specified it. I believe I asked for one week is possible. [07:52:23] I thought it was preannounce at eakc and release the wednesday after, which suggests ~1 week [07:52:41] One day is still much better than not at all. [07:52:54] if one week, a release on the April 2nd is still possible with a pre-announce on March 26th at EAKC and on the mailing lists. [07:53:12] --- Simon Wilkinson has left [07:53:17] I think that was our plan last time it was discussed. [07:53:27] --- Simon Wilkinson has become available [07:55:24] that continues to be the plan unless something prevents it [07:55:45] The very nature of plans ;-) [07:55:46] it really is just a question of Simon having the time to do all of the behind the scenes work [07:56:20] Understood, and accepted. [07:57:34] Ok. Anything else to discuss today? [07:57:50] --- Simon Wilkinson has left [07:57:57] not here [07:59:36] So let's close the meeting. Thanks a lot everyone! [07:59:46] See you at CERN [08:00:25] Yeah. First release-team meeting pitcher will be on me :-) [08:01:09] --- wiesand has left [08:02:37] --- deason has left [08:02:38] --- meffie has become available [08:02:51] --- deason has become available [08:03:16] --- Marc Dionne has left [08:06:46] (sorry, unavoidable conflict, second pitcher will be on me.) [08:12:04] --- meffie has left [08:41:51] --- meffie has become available [08:43:20] --- meffie has left [08:54:24] --- Simon Wilkinson has become available [08:58:06] --- Simon Wilkinson has left [09:13:01] --- stephan.wiesand has become available [11:36:45] --- stephan.wiesand has left [11:40:56] --- deason has left [11:40:57] --- deason has become available [17:37:31] --- shadow@gmail.com/barnowl7628413F has left [17:37:47] --- shadow@gmail.com/barnowl7628413F has become available [18:28:17] --- deason has left [22:04:50] --- shadow@gmail.com/barnowl7628413F has left [22:09:40] --- shadow@gmail.com/barnowl7628413F has become available [23:23:49] --- Simon Wilkinson has become available