12:01:32 #startmeeting 12:01:32 Meeting started Wed May 27 12:01:32 2015 UTC. The chair is ndevos. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:01:32 Useful Commands: #action #agreed #halp #info #idea #link #topic. 12:01:37 #info Agenda: https://public.pad.fsfe.org/p/gluster-community-meetings 12:01:41 #topic Roll Call 12:01:58 Hello all, who's joining us today? 12:02:00 * kkeithley is here 12:02:04 * overclk is here 12:02:38 * hchiramm_ is here 12:02:59 * spot is here 12:03:10 * raghu is here 12:03:12 * krishnan_p is here 12:03:16 hagarth might join a little later, we'll see if he makes it 12:03:24 * JustinClift waves 12:03:36 * tigert is here 12:03:38 * atinmu is here 12:03:38 * jdarcy particles. 12:03:44 very good, lets gets started 12:03:54 jdarcy, you had to! 12:03:55 #topic Action Items from last week 12:04:09 #info Subtopic: raghu to release 3.6.4beta1 by the end of next week 12:04:31 raghu: I think you and hchiramm_ did get that done? 12:04:41 yes.. packages are available 12:04:44 ndevos: Its done and the RPMs are also available. I just announced it in the mailing lists 12:04:48 * soumya is here 12:04:56 http://download.gluster.org/pub/gluster/glusterfs/qa-releases/3.6.4beta1/ 12:05:14 #info glusterfs 3.6.4beta1 has been made available and was announced 12:05:22 #info Subtopic: JustinClift to have basic forge v2 operational by Gluster Summit 12:05:27 Working on it slowly. 12:05:34 JustinClift: any ETA? 12:05:36 I seem to be putting time into anything non-gluster. :/ 12:05:39 raghu, I am just wondering how come 3.6.4 is out without considering BZ 1206429 12:05:48 In the next fortnight 12:05:49 raghu, the same was a blocker for 3.6.4 12:05:51 3.6.4beta1 12:05:55 atinmu, ^^ 12:06:19 * schandra is here 12:06:28 atinmu, I mean , its beta release 12:06:37 atinmu, hchiramm_, raghu: lets put that under the 3.6 topic a little later 12:06:52 hchiramm, I see 12:07:02 JustinClift: any ETA? 12:07:10 ndevos: In the next fortnight 12:07:32 Not because it's a lot of effort, but more because I seem to be avoid it unconciously :/ 12:07:43 * JustinClift is pretty sure it's just a burn-out thing 12:07:49 JustinClift: fortnight is 2 weeks, right? 12:07:55 Yep 12:08:04 fort(teen) nights 12:08:14 Unlikely this week. Pretty likely next week, if that helps. :) 12:09:02 #action JustinClift will have something to say about the Gluster Forge v2 on the June 10 meeting 12:09:11 Tx 12:09:17 #info Subtopic: tigert to announce new Gluster website layout demo to the mailing lists 12:10:08 tigert: ? 12:10:32 I'm not sure if that was sent out, but the new website is live for all I know 12:11:11 #action tigert to announce new Gluster website layout demo to the mailing lists (or drop from the agenda) 12:11:23 #info Subtopic: tigert will announce planet.gluster.org and include the rss urls too 12:11:47 #info announcement for planet.gluster.org was sent 12:12:08 #info Subtopic: hchiramm_ to provide an update on new documentation workflow in the mailing lists 12:12:18 ndevos, will send it today 12:12:19 hchiramm_: I did not see that email? 12:12:24 ok 12:12:53 #action hchiramm_ will send an email about the new documentation workflow to the lists TODAY (May 27) 12:13:19 hchiramm_: will you include a note on how pages in the wiki should be edited, or maybe, not edited? 12:13:31 will have a note about it 12:13:47 a detailed contribution page will be part of glusterdocs project 12:14:05 ok, thanks! 12:14:09 yw 12:14:38 hchiramm_: also, I was wondering where to put work-in-progress design notes, the wiki was quite nice for that 12:14:48 yes, true.. 12:14:57 but we cannot maintain wiki only for that 12:15:15 so we have to get those in docs projects as PRs in github 12:15:16 indeed, just give some suggestion what to use :) 12:15:37 or, maybe we should track them in etherpads and emails 12:15:49 but well, that gets off-topic now :) 12:15:59 true. :) 12:16:15 #action hchiramm_ will also include a not about the status and feature of the wiki 12:16:25 * msvbhat arrives late 12:16:29 *note 12:16:37 #chair hagarth 12:16:37 Current chairs: hagarth ndevos 12:16:45 hagarth: you're just in time! 12:16:51 #topic GlusterFS 3.7 12:16:56 ndevos: ok :) 12:17:01 open bugs: https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&id=glusterfs-3.7.1&hide_resolved=1 12:17:10 hagarth: its all yours now 12:17:34 If only there were a tool to track a document through multiple rounds of collaborative editing. We could call it a "content management system" or something like that. 12:17:42 ok, we have noticed a glusterd issue post upgrade to 3.7 12:17:56 we need to get that sorted out soon. 12:18:12 hagarth, what issue? 12:18:23 we are intending to release 3.7.1 with a fix for that issue alone. 12:18:38 kshlm: stale frame->cookie issue for which atinmu has sent a patch. 12:18:54 Oh. Okay. 12:18:56 thats a glusterd issue only, right? 12:19:02 ndevos, yes 12:19:10 if all goes well, we might release 3.7.1 this week and push out all fixes targeted for 3.7.1 to 3.7.2 12:19:18 Just waiting for you to give the Green flag for merging. 12:19:37 since there is a good number of patches on 3.7.2, I intend to do 3.7.2 soon (approx 2 weeks from now) 12:19:52 kshlm: waiting for all regression tests to run clean to be able to do that. 12:20:03 atinmu: what is the BZ for that particulat 3.7 bug? 12:20:22 s/there is/there are/ 12:20:24 hagarth, what abt the logging changes ? 12:20:32 * atinmu checks for the BZ 12:20:34 will it be part of 3.7.1 or 3.7.2 ? 12:20:44 hchiramm_: certainly not for 3.7.1 12:20:54 ndevos, https://bugzilla.redhat.com/show_bug.cgi?id=1223213 12:21:02 thanks atinmu! 12:21:28 atinmu: this happens for those who upgrade from < 3.6 right? 12:21:43 hagarth, yes 12:21:44 hagarth: how likely is the "emergency" release of 3.7.1 going to be? 12:22:13 ndevos: I am quite certain .. about 90% + :) 12:22:30 hagarth: any reasons not to do it? 12:22:34 ndevos: just waiting for the moratorium to be lifted. 12:22:38 ndevos: can't think of any 12:22:53 hagarth, and we can hit this problem even in a non-upgrade case if there is a rpc failure 12:23:02 okay, we're in a meeting now, so we can decide what to do, and when :) 12:23:14 atinmu: right, it is good to have this issue addressed. 12:23:28 Um, an emergency release shouldn't wait for the moratorium 12:23:36 That kind of defeats and emergency ;) 12:23:39 s/and/an/ 12:23:52 #agreed there will be a hot-fix 3.7.1 release which only includes fix(es) for bug 1223213 12:24:04 did someone write a test case for this issue? :) 12:24:13 JustinClift: we are very close to remove the moratorium 12:24:21 If it's an emergency fix, can we put it out today anyway? 12:24:39 JustinClift: tomorrow :) 12:24:42 k ;) 12:24:52 spot: yes, it has been tested well 12:25:08 hagarth: i meant more in the sense of "prevent this from happening again in the future" 12:25:14 #action ndevos moves all glusterfs-3.7.1 blockers to 3.7.2 so that the intended bugs for 3.7.2 are still on track 12:25:30 #info glusterfs-3.7.2 is planned to be released in approx. 2 weeks 12:25:37 spot, Emmanuel spotted this bug in his setup 12:25:48 atinmu: any thoughts on adding a regression test for the issue? 12:26:02 manu rocks! 12:26:17 spot, I setup a 3.5 cluster and then upgraded one of the node to 3.7 and then things started failing 12:26:39 hagarth: thats what I was getting at. :) 12:27:07 atinmu: automated regression tests? can it be done on a single vm environment? 12:27:25 hagarth, as of now we don't have any tests for heterogeneous cluster (I am talking about .t) 12:28:01 atinmu: rpc failure? 12:28:09 can that happen in a homogeneous cluster? 12:28:40 hagarth, yes we could test that 12:28:45 Could it happen using the cluster.t or similar mock-cluster stuff in our regression testing? 12:29:13 atinmu: let us consider adding a regression test for this 12:29:37 #action atinmu will think about an automated regression test for bug 1223213 12:29:40 JustinClift, it could have, but manu raced all of them :) 12:30:17 I am also interested in adding automated upgrade tests to distaf 12:30:26 JustinClift: Probably not. You really need the nodes to be separate wrt software versions, which wouldn't be the case with the cluster.rc stuff. 12:31:16 anyway, that is the update on 3.7. 12:31:23 jdarcy: Ahhh, gotcha 12:32:02 #idea have regular (nightly/weekly) distaf runs for automated update testing 12:32:11 hchiramm_: are you pushing dockerfiles for 3.7? 12:32:15 This sounds like a job for . . . containers. 12:32:20 hagarth, I have pushed that change 12:32:25 need to rebuild and see 12:32:35 hchiramm_: isn't rebuilding automatic? 12:32:53 jdarcy, in this particular case, we don't necessarily need different software versions for the failure. But ... 12:32:55 the repo which is forked under gluster , dont have it 12:33:00 hagarth, ^^.. 12:33:07 hagarth: you want to continue with the rest of the meeting, or shall I? 12:33:13 ndevos: please go ahead 12:33:26 #topic GlusterFS 3.6 12:33:41 raghu: you released 3.6.4beta1, right? 12:33:46 I have just announced 3.6.4beta1 in the mailing list 12:33:57 and it is missing a patch that should have been included? 12:34:00 ndevos: yes. Also I have pushed some more patches 12:34:26 ndevos: That is a blocker for 3.6.4. I will definitely include that patch for 3.6.4 (probably for beta2 itself) 12:34:35 ndevos: but it does not have a +1 yet 12:34:38 * ndevos scrolled up and found bug 1206429 in the history 12:35:06 ndevos: so waiting for it to get reviewed. Have asked kshlm to review it. Once it gets an ack will merge it 12:35:21 #info glusterfs-3.6.4beta2 is needed, a few more patches will get included, also fixing bug 1206429 12:35:55 raghu: sounds like a good plan to me, when do you think beta2 can be made? 12:36:19 ndevos: most likely by end of next week 12:36:22 kshlm: when do you have time to review the patch(es) for bug 1206429 ? 12:36:27 raghu: ok 12:37:05 ndevos: I have already merged few more patches after beta1 is done. http://review.gluster.org/#/q/status:open+project:glusterfs+branch:release-3.6 are the list of patches for release-3.6 12:37:18 #agreed glusterfs-3.6.4beta2 is targetted for the end of next week (June 5) 12:37:19 I'm doing it now. 12:37:40 thanks kshlm! 12:37:59 raghu: anything else to add? 12:38:13 ndevos: nope. I need some help with review of a patch 12:38:47 ndevos: its for release-3.6 branch. I have reviewed it and given a +1. But need one more +1 before merging it. 12:39:11 I have added some reviewers. If one of you can review, it would be great 12:39:33 #info Developers are encouraged to review backports of patches too, not only patches for the master branch 12:39:42 http://review.gluster.org/#/c/8351/ 12:40:19 raghu: okay, those links should make it into the meeting minutes, hopefully people read those 12:40:51 #topic GlusterFS 3.5 12:41:02 open bugs blocking the next release: https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&hide_resolved=1&id=glusterfs-3.5.4 12:41:28 a beta was made available a while back, nobody reported any issues or succcess about it 12:41:46 I'll probably release glusterfs-3.5.4 later this week 12:42:13 #info glusterfs-3.5.4 will likely be released later this week 12:42:31 any questions for 3.5? 12:42:56 I consider that as a "no" 12:42:58 #topic GlusterFS 4.0 12:43:18 open bugs for 4.0: https://bugzilla.redhat.com/showdependencytree.cgi?id=glusterfs-4.0 12:43:33 jdarcy: care to comment? 12:44:04 Yeah, I need to add feature requests for all of the 4.0 stuff. 12:44:39 I also need to write up some of the latest DHT2 ideas. 12:44:51 #info Features for 4.0 should add "glusterfs-4.0" to the "blocks" field in the bug report 12:44:56 Other than that, still just waiting until others can pull free from 3.x stuff. 12:46:13 okay, any thing else? 12:46:19 or questions from others? 12:46:54 ... silence? 12:46:59 * spot drops a pin 12:47:07 where's the kaboom 12:47:10 #topic Open Floor / Bring your own topic 12:47:16 there's supposed to be an earth shattering kaboom 12:47:19 #info subtopic https://glusternew-tigert.rhcloud.com 12:47:48 marked as "To be discussed on-list" 12:47:55 I guess that is kinda old stuff 12:48:03 because the glusternew is the new gluster.org now 12:48:33 #info glusternew has replaced the old gluster.org site, glusternew *is* the new gluster.org 12:48:54 thanks tigert! 12:49:01 tigert: I think we should drop a note about how to contribute to this site 12:49:15 yes 12:49:26 easiest is "click the github link on the bottom" 12:49:52 tigert: please post that in an email to the lists :) 12:49:58 wilco 12:49:58 tigert: yes, making it obvious could help :) 12:50:18 #action tigert will send an email to the lists on how to contribute to the new gluster.org site 12:50:36 #info subtopic: "earth shattering kaboom" from kkeithley aka purple/unnamed in the etherpad 12:50:47 kkeithley: BOOM!? 12:50:51 I was named, what happened? 12:51:03 there, fixed it 12:51:04 :) 12:51:09 :D 12:51:47 kkeithley: please have your say, I'm not going to copy/paste all that 12:51:57 Lala has moved on to other things. He was building the RPMs. hchiramm_ had a possible candidate lined up. Any news? 12:52:23 kkeithley, not yet. 12:52:32 okay 12:52:41 may be we need to poll it from gluster community members .. 12:52:47 who ever interested.. 12:52:52 sure 12:52:58 yeah, we definitely need a replacement 12:53:17 The CentOS Storage SIG was also Lala's. I think he's still involved with that. I'll check with him 12:53:22 #info volunteers for assisting with the RPM packaging of GlusterFS are very welcome! 12:53:38 yes, we have some actions there.. so for now , we can forget that 12:53:45 such a volunteer would need to be able to build in Fedora Koji, FYI 12:53:55 yep 12:54:08 Which is not hard 12:54:24 :P 12:54:35 And I'll check with Lala about where things stand with the CentOS Storage SIG repos 12:54:41 hchiramm_, kkeithley: who of you want to send a request for volunteers to the lists? 12:54:42 in SIG 12:54:53 we have builds/targets and tags 12:55:02 however we need to see the long term plan for el builds .. 12:55:09 we dont want to maintain 2 repos 12:55:17 I mean download.g.org and centos sig 12:55:48 right. We can just have a readme for EL rpms that points at the storage sig. Just like we do for the Ubuntu PPA 12:55:52 hchiramm_: yes, please start a discussion with the options and considerations on the list 12:55:54 may be in next sig meeting we can discuss more about this.. 12:56:18 ndevos, ok.. 12:56:28 hchiramm_: if we need a temporarily builder, i can do it 12:56:43 spot++ thanks 12:57:14 atm I'm the temporary builder, but we need a long term solution. One that scales better. ;-) 12:57:23 true :) 12:57:27 kkeithley: let me know if i can help you out. 12:57:44 thanks, will do 12:57:51 spot: whatever you do, DON'T LOOK AT THE SPEC FILE 12:58:03 ndevos: yeah. i'm sure it's awesome. :) 12:58:14 most awesome! 12:58:28 :P :P 12:58:50 okay, Is semiosis missing in action? I'm hearing rumblings that 3.6.3 and 3.7.0 aren't in the PPA. Anyone know? 12:58:58 hchiramm_, kkeithley: so, who's doing the call for volunteers? 12:59:06 * ndevos would like to #action that one 12:59:11 I nominate hchiramm_ 12:59:16 ndevos, will do :P 12:59:19 kkeithley, :) 12:59:53 #action hchiramm_ will send out a call (email and blog?) for volunteers for helping with the packaging 13:00:10 (thats not restricted to RPMs) 13:00:36 because debian packaging is so much fun 13:00:50 Nobody knows about semiosis and the PPA? 13:00:56 * ndevos does not 13:01:17 Is it too early to think about alternatives to the PPA? 13:01:24 s/think/talk/ 13:01:27 * hchiramm_ does not 13:01:31 I think there is a gluster-team shared PPA 13:01:50 that should not be restricted to semiosis, others should be able to get into that team somehow 13:01:56 do we know who is on that team? 13:02:10 last time I checked, it was semiosis and JoeJulian 13:03:08 #idea add more people to the Ubuntu PPA for (https://launchpad.net/~gluster) the Gluster packages 13:03:55 any other last minute topics? we've already run a little longer than planned 13:04:41 no? well, in that case we'll finish the meeting 13:04:52 I hope to see you attending next week again, thanks! 13:04:56 #endmeeting