12:02:05 <jdarcy> #startmeeting Gluster community weekly meeting
12:02:05 <zodbot> Meeting started Wed Mar 30 12:02:05 2016 UTC.  The chair is jdarcy. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:02:05 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:02:05 <zodbot> The meeting name has been set to 'gluster_community_weekly_meeting'
12:02:15 <jdarcy> OK, who's here?
12:03:05 * atinm waves
12:03:18 * poornimag is here
12:03:23 <jdarcy> For a moment I thought this was going to be a *really* short meeting.
12:03:28 <jdarcy> ;)
12:03:30 * kshlm here
12:04:01 * jiffin is here
12:04:02 <jdarcy> #topic Next week's host
12:04:03 * ndevos attends
12:04:14 <kshlm> kkeithley, Can you OP me here for a little while? I'd like to update the topic.
12:04:42 <jdarcy> Any volunteers to host next week?
12:05:04 <kshlm> I can if there are no else. But I'd like to see some other hosts :)
12:05:41 <jdarcy> I believe we'll be sticking with this time for next week, instead of alternating.
12:05:41 <ndevos> what would be the time of next weeks meeting?
12:05:59 <kshlm> That's something we need to discuss.
12:05:59 <jiffin> jdarcy: shyam was ready to host next week meeting right(at 15 UTC)?
12:06:01 <jdarcy> That's a statement of my understanding, not advocacy for a particular outcome.
12:06:19 <kshlm> I believe I put it up as a topic of discussion.
12:06:19 <jdarcy> jiffin: OK, so we are going to keep alternating?  Cool.
12:06:46 <jdarcy> Yeah, let's come back to that during Open Floor.
12:07:03 * msvbhat joins bit later to the meeting
12:07:05 <kshlm> Cool. So I think I'll volunteer next week.
12:07:12 * ira is here.
12:07:28 <jdarcy> Thank you, kshlm.  I'll nudge Shyam to see if he's "willing" to take a turn.
12:07:36 <jdarcy> #topic Last week's AIs
12:07:39 <kshlm> shyam can take up the next time we do the meeting at 15UTC (If we do it)
12:07:41 * rjoseph is here
12:07:53 <jdarcy> kkeithley to send a mail about using sanity checker tools in the codebase
12:08:41 <jdarcy> I think I saw patches, not sure about email.  kkeithley?
12:09:04 <ndevos> I think pkalever is involved in that too, not sure how they divided the work
12:09:26 <kshlm> This AI has been here since the new year IIRC.
12:09:36 <jdarcy> I guess we'll keep that one for next week.
12:09:44 <jdarcy> AI: kshlm & csim to set up faux/pseudo user email for gerrit, bugzilla, github
12:10:01 <kshlm> We'll probably get to it after the jenkins move.
12:10:07 <jdarcy> OK.
12:10:14 <jdarcy> AI: msvbhat to announce DiSTAF HOWTO on mailing lists
12:10:47 <msvbhat> Ah, I forgot that
12:10:52 <msvbhat> I will announce now.
12:11:05 <kshlm> msvbhat, The howto is complete?
12:11:17 <msvbhat> kshlm: Yes.
12:11:21 <jdarcy> Also related: msvbhat to add links to DiSTAF on the dev workflow page.
12:11:47 <kshlm> msvbhat, Please make sure you do that before you disappear for the next 2 weeks.
12:12:15 <kshlm> :)
12:12:23 <jdarcy> Where *is* the correct dev workflow page, anyway?  Google takes me to what seem to be old pages.
12:12:29 <msvbhat> jdarcy: Not done too :( But i remember searching for dev workflow page in github
12:12:45 <msvbhat> jdarcy: Yes, That was my confusion too.
12:12:52 <ira> Is there one?
12:12:59 <jdarcy> OK, then.
12:13:02 <msvbhat> kshlm: Make sure that I announce the howto to ML?
12:13:19 <kshlm> That and add the link.
12:13:22 <jdarcy> #action jdarcy to find/create "official" dev workflow page
12:13:35 <kshlm> #link https://gluster.readthedocs.org/en/latest/Developer-guide/Development-Workflow/
12:13:56 <ndevos> #link http://gluster.readthedocs.org/en/latest/Developer-guide/Developers-Index/
12:14:04 <jdarcy> OK, nvm.
12:14:09 <jdarcy> #inaction ...
12:14:19 <ndevos> thare are actually two, one simple one and one .... more complete?
12:14:22 <msvbhat> Ah, Okay. Will do that now
12:14:28 <jdarcy> AI: jdarcy to share his notes from his Bangalore discussions with the community
12:14:59 <jdarcy> Not done.  I got dragged into another downstream fire.  Will try to do it this week.
12:15:16 <kshlm> Next week then.
12:15:21 <jdarcy> AI: ndevos to reach out to packagers about 3.5.9 packaging
12:15:36 <ndevos> yes, I did that
12:15:40 <jdarcy> Yay!
12:15:59 <ndevos> 3.5.9 was packaged for NetBSD a while already, and I think Kaleb build RPMs for many versions
12:16:12 <jdarcy> AI: kshlm to update hagarth of 3.7.10 date
12:16:22 <kshlm> I did that.
12:16:42 <jdarcy> Yay again.  Setting a great example here, kshlm and ndevos.  :)
12:16:54 * ndevos \o/
12:16:58 <kshlm> Thank you :)
12:17:00 <jdarcy> AI: kshlm to ask rabhat for an update on 3.6.10
12:17:09 <kkeithley> yes, RHEL/CentOS+Fedora RPMs and Ubuntu Launchpad packages for 3.5.9 are available
12:17:23 <kshlm> Done that as well.
12:17:28 <kshlm> But rabhat hasn't replied.
12:17:29 * kkeithley is wondering where all our packaging volunteers are
12:17:57 <jdarcy> kkeithley: You mean within this group, or within distros etc.
12:18:00 <jdarcy> ?
12:18:47 <kkeithley> well, in general yes, but I was specifically thinking of the volunteers we lined up in BLR last year
12:20:09 <jdarcy> No idea.  Should we poke them again?
12:20:21 <kkeithley> and apropos of not very much, Debian packages for 3.5.8 were never made, and nobody ever complained, so maybe those don't matter any more?
12:21:29 <kkeithley> within distros? I'm only aware of pmatthai (sp?) in Debian and Emmanuel in NetBSD.  Who else is there?
12:21:38 <jdarcy> Maybe we should use the ML to gauge interest and/or call for volunteers?
12:21:52 <jdarcy> kkeithley: Do we have someone for FreeBSD?
12:22:06 <kkeithley> I don't believe so, no?
12:22:13 <kkeithley> s/?//
12:22:31 <ndevos> people that are interested in doing the packaging are subscribed to http://www.gluster.org/mailman/listinfo/packaging already...
12:22:51 <kshlm> y4m4 used to do it for FreeBSD, IIRC.
12:22:53 * post-factum is late
12:23:05 <jdarcy> Quiet list.
12:23:39 <jdarcy> Is there an AI we should add that would help?
12:24:04 <ndevos> FreeBSD seems to be on 3.7.6, Craig Butler updated it last time
12:24:09 <ndevos> #link https://www.freebsd.org/cgi/ports.cgi?query=glusterfs&stype=all
12:24:22 <jdarcy> OK, moving on then...
12:24:23 <kkeithley> we have a team, the release announcements get sent, would additional nag mails make any diff?
12:24:52 <jdarcy> I don't know.  Suspect not, but figured I'd ask.
12:24:59 <jdarcy> #topic GlusterFS 3.7
12:25:12 <kshlm> hagarth isn't here.
12:25:27 <post-factum> 3.7.10 should be tagged today...
12:25:30 <kkeithley> fwiw, starting with 3.8 the RHEL/CentOS pkgs will be done in the CentOS Storage SIG and we'll just point to those, the same way we do today for Ubuntu and SuSE.
12:25:45 <kshlm> post-factum, yes
12:25:49 <ndevos> hagarth mentioned in one email that he does not have time to do 3.7.10 this week :-/
12:25:54 <post-factum> :(
12:25:59 <jdarcy> #info starting with 3.8 the RHEL/CentOS pkgs will be done in the CentOS Storage SIG and we'll just point to those, the same way we do today for Ubuntu and SuSE.
12:26:08 <ndevos> not sure if any of the other maintainers is stepping in to get it done
12:26:15 <kshlm> Yeah, so we need a volunteer to get it done this week.
12:26:24 <post-factum> skoduri: are you here?
12:26:30 <ndevos> #link http://thread.gmane.org/gmane.comp.file-systems.gluster.maintainers/508/focus=519
12:26:31 <skoduri> post-factum, yes..
12:26:48 <post-factum> skoduri: bothering you again about rpc fixes :). any news?
12:26:52 <skoduri> post-factum, was about to ping..we need http://review.gluster.org/#/c/13824/ to be merged
12:26:59 <post-factum> oh
12:27:14 <post-factum> that is the last of 3?
12:27:25 <skoduri> post-factum, its waiting on regressions..
12:27:28 <kshlm> I can volunteer for this one time.
12:27:49 <skoduri> post-factum, no.. we have skipped one as it dint get in even in master
12:27:53 <jdarcy> Thanks again, kshlm.
12:28:00 <atinm> kshlm, thank you!
12:28:07 <ndevos> kshlm, great, thanks!
12:28:15 <post-factum> skoduri: i see, so no luck to get it into .10
12:28:59 <skoduri> post-factum, right..but we definitely need http://review.gluster.org/#/c/13824/ as another patch http://review.gluster.org/#/c/13507/ got merged which had dependency on this :-|
12:29:01 <kshlm> So as the release manager, I'll be aiming to do the release later tonight.
12:29:18 <kshlm> I'll send an update asking for any pending changes to be merged soon.
12:29:28 <skoduri> kshlm, I re-triggered the regression tests..
12:29:30 <skoduri> kshlm, sure
12:29:52 <atinm> I've been seeing loads of regression failures today and informed misc about it, not sure whether that's gonna slow down this tagging
12:30:04 <post-factum> skoduri: http://review.gluster.org/#/c/13507/ is merged into 3.7
12:30:07 <atinm> I think we still have bunch of patches lined up for 3.7.10
12:30:13 <jdarcy> atinm: Any particular cause, or the usual random selection of causes?
12:30:31 <atinm> jdarcy, in middle of the run the slaves went offline
12:30:33 <skoduri> post-factum, right and it has dependency on #13824 which hasn't got merged
12:30:41 <post-factum> ah, got it now
12:30:42 <jdarcy> Ugh.
12:30:49 <post-factum> but there was one more patch
12:30:58 <ndevos> atinm: its not a problem if some patches do not make it in 3.7.10, there is 3.7.11 in a few weeks anyway?
12:31:04 <atinm> jdarcy, misc suspects that underlying virt host has crashed
12:31:20 <kshlm> I was rebooting disconnected slaves a little while before, but I don't believe that should have caused any failures.
12:31:25 <skoduri> post-factum, the 3rd patch is independent of these 2 and can be taken in next release
12:31:38 <atinm> ndevos, but if they are attached to the 3.7.10 tracker then?
12:31:44 <post-factum> skoduri: looking now at my list in tracking bug
12:32:04 <post-factum> http://review.gluster.org/#/c/13456 -- merged
12:32:18 <post-factum> http://review.gluster.org/#/c/13592 -- not merged
12:32:20 <kshlm> I kindof remember seeing a rackspace email yesterday. Didn't read it, but they could have been performing maintenance.
12:32:26 <ndevos> atinm: they'll get moved to the new 3.7.11 one :)
12:32:30 <post-factum> http://review.gluster.org/#/c/13658 -- not merged
12:32:32 <atinm> ndevos, :)
12:32:33 <jdarcy> kshlm: Yeah, I have that same vague memory.
12:32:49 <post-factum> so, http://review.gluster.org/#/c/13658 is the 4th :)
12:33:37 <skoduri> post-factum, http://review.gluster.org/#/c/13658 is the 3rd one :)
12:33:52 <post-factum> skoduri: i'm getting lost in those numbers :)
12:33:55 <skoduri> http://review.gluster.org/#/c/13507/ is the backport http://review.gluster.org/#/c/13592
12:34:06 <skoduri> post-factum, you are looking at master branch fixes..
12:34:12 <post-factum> aha!
12:34:40 <post-factum> so, again. 13456 merged, 13507 merged, and waiting for 13658
12:34:52 <jdarcy> Seems like we have some patches to get sorted out before 3.7.10, but should happen soon?
12:35:14 <skoduri> post-factum,  13456 merged, 13507 waiting on regression, and 13658 (master) probably taken in the next release
12:35:25 <post-factum> skoduri: plus 13824
12:35:46 <post-factum> just to make sure what i will cherry-pick
12:35:56 <jdarcy> #topic  GlusterFS 3.6
12:36:10 <skoduri> post-factum, yes...we can take it on gluster-dev post this meeting :)
12:36:24 <jdarcy> Anything new in 3.6-land?
12:36:26 <kshlm> Just checked those rackspace mails. They did perform network maintenance which caused a lot of the slaves to lose network for some time.
12:36:30 <post-factum> skoduri: thanks so much for clarifying my in-head mess
12:37:09 <jdarcy> I don't think I've seen rabhat, so I'm guessing no news.
12:37:22 <kshlm> He's been missing for some weeks.
12:37:26 <jdarcy> #topic GlusterFS 3.5
12:37:45 <jdarcy> kshlm: Some pretty severe distraction has been happening there.
12:37:54 <ndevos> 3.5.9 is ready, needs to be announced still
12:38:15 <ndevos> no changes have been sent for 3.5.10 yet, but there are ~10 days left for that
12:38:30 <kshlm> jdarcy, Maybe someone else should manage the release then.
12:39:42 <jdarcy> kshlm: A fair point.
12:40:01 <jdarcy> ndevos: What else needs to happen prior to announcement of 3.5.9?
12:40:20 <ndevos> jdarcy: no, not really, I just need to find a few minutes to get it done
12:40:28 <jdarcy> ndevos: OK, cool.
12:41:23 <jdarcy> #topic GlusterFS 3.8
12:41:39 <ndevos> ah, me again!
12:41:54 <ndevos> so we did get only few updates of features since last week
12:42:30 <jdarcy> When does that window close, again?
12:42:30 <ndevos> jiffin and I will start moving the features that do not send updates to a new "at risk" category on the roadmap
12:42:51 * msvbhat leaves the meeting and will read the chatlogs later
12:43:02 <ndevos> all changes need to be in by the end of April, but the earlier the better
12:43:21 <ndevos> and we need to see some regular progress, for example on updates to the designs/docs
12:43:36 <ndevos> #link https://www.gluster.org/community/roadmap/3.8/
12:43:54 <ndevos> many of the features are not even complete on that page
12:44:07 <ndevos> basic things like "summary" are often missing
12:44:15 <jdarcy> ndevos: Indeed.
12:44:45 <ndevos> and it's really easy to send a pull request, just click the "edit this page" link on the bottom
12:45:00 <jdarcy> IIRC, consensus was reached that experimental (e.g. 4.0) features would be disabled/removed on the 3.8 branch when it happens.  Correct?
12:45:55 <ndevos> yes, correct, after we have a release-3.8 branch, features that are not planned for 3.8 (on the roadmap!) will be disabled/removed
12:46:32 <jdarcy> You seem well on top of everything.  Looks like it'll be a solid release.
12:46:40 <kkeithley> how are we deciding disabled vs removed?
12:47:06 <ndevos> kkeithley: depends on the complexity of removing the code, and how much time I feel like spending on removing it
12:47:38 <ndevos> assistance with that is welcome, and jdarcy already offered some help - thanks!
12:48:30 <jdarcy> Disabling is usually easier - just comment/ifdef out the options or similar.  Removing usually means screwing around with configure/specfile/etc. as well.
12:49:12 <post-factum> will it be possible to update to 3.8 from 3.7?
12:49:13 <jdarcy> Yes, I'm glad to help for the parts I know something about - e.g. glusterfs itself, not so much NFS/Samba/etc.
12:49:20 <ndevos> jdarcy: indeed, but I do not want to give anyone the idea to un-comment/ifdef some things and try to get it included in 3.8.x
12:50:07 <jdarcy> ndevos: They'd still need to get something through Gerrit for that to happen, though.
12:50:11 <ndevos> post-factum: that is the plan
12:50:29 <post-factum> ndevos: but with downtime?
12:50:50 <ndevos> jdarcy: yes, they should, but that is not too bad
12:51:06 <jdarcy> Do any of the 3.8 features require downtime for expensive conversions and such?  Seems like rolling upgrade should be possible.
12:51:16 <jdarcy> That is, BTW, one of the things the feature pages are supposed to address.
12:51:17 <ndevos> post-factum: I expect so, at least for normal functionality
12:52:00 <ndevos> jdarcy: I have not heard or read about any such requirements
12:52:53 <post-factum> ndevos: ok, will it be possible to update clients to 3.8 first, and then update the servers or vice-versa? just not to stop the world
12:53:13 <jdarcy> ndevos: Yeah, once the feature list has settled we should (dangerous word) be able to go through the feature pages and see what the upgrade requirements are.
12:53:54 <ndevos> post-factum: at the moment I am not aware of any incompatibilities that would prevent either way of upgrading
12:54:13 <jdarcy> post-factum: What he said.  ;)
12:54:36 <jdarcy> #topic GlusterFS 4.0
12:54:37 <ndevos> jdarcy: yes, and hence the repeated requests to the feature owners/developers to get their documentation in order :)
12:55:02 <jdarcy> Approximately zilch to report here.  AFAIK all of the 4.0 developers have been totally slammed with other tasks.
12:55:20 <atinm> jdarcy, :-\
12:55:53 <atinm> jdarcy, but still one good news from GD2 is kshlm pushed the txn framework changes and its looking good to me :)
12:56:10 <kshlm> Finally!
12:56:15 <jdarcy> Under the circumstances, a heroic effort.  Thanks, kshlm.
12:56:25 <kshlm> It's usable in a single node system right now.
12:56:44 <kshlm> I think I'll be spending a little more time on GD2 from now on hopefully.
12:56:58 <jdarcy> Yay!
12:57:12 <kshlm> I don't seem to have any other **URGENT** issues coming my way. And I hope it stays that way.
12:57:34 <atinm> kshlm, I will do my best to keep you in that state ;)
12:57:35 <jdarcy> kshlm: Yeah, we're down to *URGENT* (one star).
12:57:48 <jdarcy> #topic Open Floor
12:57:55 <kshlm> jdarcy, :)
12:58:12 <jdarcy> Item: Should we continue to rotate meeting times?
12:58:25 <kshlm> I asked this.
12:58:33 <kshlm> So what does the community feel?
12:58:42 * ira -1's.
12:58:57 <kshlm> We were trialling it for a month, and the months done.
12:58:59 * post-factum is ok with that
12:59:03 <ndevos> this time (12:00 UTC) works best for me, 3 hour later is more difficult
12:59:13 <ira> It is very confusing to know how to plan around this meeting.
12:59:21 <atinm> the question here is did we see any huge change w.r.t attendance?
12:59:32 <jdarcy> I don't think we were getting any benefit from the later time.
12:59:34 <kshlm> IMHO, I didn't see much changes.
12:59:42 * anoopcs +1's the current time.
12:59:43 <ndevos> but, I also do not need to attend every week, and it may benefit people in other locations
12:59:53 <atinm> Then it doesn't make sense to me :(
12:59:54 <jdarcy> Personally either's completely doable, but it is kind of nice to have it at a stable time.
13:00:07 * ira +1's jdarcy.
13:00:17 <jdarcy> Could we do this as a poll?
13:00:45 <kshlm> I agree.
13:00:47 <ndevos> I'm sure we could, but the last time the poll had many more votes than people attending the meeting...
13:01:38 <jdarcy> We seem to have a lot of sentiment in favor of this time, plus a concern that might only be theoretical about the later time being better for people in other locations.
13:02:14 <ndevos> I guess 12:00 UTC it is, and will stay
13:02:28 <jdarcy> Seconded.
13:02:36 <kshlm> I feel like we had lower attendance for both times. May be it was just this month, and everyone was busy.
13:03:15 <jdarcy> I'll note that on the etherpad etc.
13:03:22 <ira> kshlm: Confusion dones that.
13:03:30 <kshlm> I'm fine with  both times, btw.
13:03:35 <jdarcy> I know I showed up at the wrong time once.
13:04:03 <jdarcy> Item: 3.6 maintainer
13:04:06 <kshlm> I tried to be clear enough about the times of the next meetings, when I hosted.
13:04:26 <ira> kshlm: Humans have a hard enough time remembering 1 time ;)
13:04:51 <jdarcy> I'll poke rabhat first, but assuming he's not against the idea, should I send email to the list looking for a new volunteer?
13:05:24 <ndevos> jdarcy: yes, check with rabhat and send an email to the maintainers list asking for assistance with 3.6
13:05:48 <jdarcy> #action jdarcy to follow up on 3.6-maintainer situation
13:05:58 <jdarcy> Any other new items?
13:06:19 <ndevos> #info JoeJulian and I will be presenting at http://www.incontrodevops.it/program-idi2016/ on Friday :)
13:06:37 <ndevos> not together though, different talks
13:06:41 <kshlm> ndevos, Cool!
13:07:30 <jdarcy> ndevos: Cool.  And that segues into announcements...
13:08:00 <jdarcy> (1/3) If you're attending any event/conference please add the event and yourselves to Gluster attendance of events: https://public.pad.fsfe.org/p/gluster-events
13:08:00 <kshlm> I'd like to know the reason why we don't have more maintainers attending meetings.
13:08:11 <jdarcy> (2/3) Put (even minor) interesting topics on https://public.pad.fsfe.org/p/gluster-weekly-news
13:08:23 <jdarcy> (3/3) Use the following etherpad for backport requests  https://public.pad.fsfe.org/p/gluster-backport-requests
13:08:51 <jdarcy> kshlm: I think the same people who become maintainers also tend to be key for other things, and get forcibly distracted.
13:09:18 <kshlm> I'd like to think that taking part in community discussions is a maintainer reponsibility.
13:09:33 <jdarcy> kshlm: Agreed.
13:09:54 <jdarcy> In fact...
13:10:13 <jdarcy> #agreed taking part in community discussions is a maintainer reponsibility.
13:10:22 <jdarcy> (That should get it into the minutes)
13:10:33 <jdarcy> (For all the good *that* does)
13:10:42 <post-factum> :D
13:11:01 <jdarcy> Last call for new business.
13:11:34 <kshlm> I'll take this up on the maintainers list.
13:11:36 <kshlm> I
13:11:45 <kshlm> i'd like to get some feedback.
13:11:54 <kshlm> That's all I had.
13:12:17 <jdarcy> #info Next meeting: 12:00 UTC, chaired by kshlm
13:12:44 <jdarcy> kshlm: It shouldn't hurt to raise the issue, if only so people have something to show their distractors.
13:13:07 <jdarcy> Thanks for putting up with my random chairmanship, everyone.
13:13:13 <jdarcy> #endmeeting