12:01:29 <hagarth> #startmeeting
12:01:29 <zodbot> Meeting started Wed Jan 14 12:01:29 2015 UTC.  The chair is hagarth. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:29 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:42 <hagarth> #topic Roll Call
12:01:47 <hagarth> who do we have here today?
12:02:05 <hagarth> I expect the attendance to be a bit thin today as it is a holiday in India
12:02:09 * jimjag : Jim Jagielski
12:02:21 * raghu is here
12:02:21 * jdarcy is.
12:02:23 * aravindavk is here
12:02:31 <hagarth> jimjag: welcome!
12:02:34 * partner here
12:03:14 * jimjag thanks hagarth
12:03:22 <hagarth> I see ndevos lurking on the etherpad too, so he should be here shortly
12:03:40 <hagarth> #topic AIs from last meeting
12:03:57 <hagarth> AI - hagarth needs to send a cleanup patch for the MAINTAINERS file to remove inactive maintainers (add an alumni section?)
12:04:17 <hagarth> this is done - http://review.gluster.org/9445
12:04:20 <hagarth> reviews are welcome
12:04:45 <hagarth> AI - hchiramm will try to fix the duplicate syndication of posts from blog.nixpanic.net
12:05:04 * ndevos _o/
12:05:06 <hagarth> this is still WIP, why it is getting syndicated twice is still a mystery.
12:05:16 <hagarth> ndevos: right in time for your topic :)
12:05:30 <hagarth> AI - Debloper to write  a blog post about complete website revamp
12:05:43 <hagarth> I haven't seen this happen, will follow up with Debloper
12:06:08 <hagarth> AI - Working with Tuomas K (IRC: tigert) to create a stopgap landing page, Wiring needs to start for temp fix page
12:06:16 <hagarth> tigert: should we consider this done?
12:06:39 <hagarth> not sure if tigert is around, so moving on to next AI
12:06:41 <hagarth> AI - tigert to respond to Soumya Deb on gluster-infra about website.
12:06:58 <hagarth> this is done, we need to discuss further thought about website revamp.
12:07:15 <hagarth> AI - hagarth to contact Rikki Endsley about fixing the twitter glusterfs domain squatter problem
12:07:33 <hagarth> DONE and spot is pursuing with legal to claim @glusterfs back on twitter
12:07:47 <hagarth> AI - hchiramm_ to send an email on gluster-infra about gerrit upgrade
12:08:16 <ndevos> lala sent a related one, misc responded: http://www.gluster.org/pipermail/gluster-infra/2015-January/000674.html
12:08:32 <hagarth> ndevos: right, it would be cool to have an updated gerrit soon.
12:08:45 <hagarth> AI - ndevos will send an email about bugs that need patches (or reviews) for 3.5.x
12:09:12 <ndevos> done.
12:09:24 <hagarth> AI - ndevos will send out an announcment about the NFS with exports+netgroups feature (and lobby for reviews)
12:09:31 <ndevos> done too!
12:09:40 <hagarth> ndevos: thanks
12:09:50 <hagarth> AI - hagarth to open a feature page for (k)vm hyperconvergence
12:09:57 <hagarth> this is TBD, I will carry forward this AI
12:10:01 <hagarth> #action hagarth to open a feature page for (k)vm hyperconvergence
12:10:10 <hagarth> AI - hagarth to share results of community survey and solicit feature proposals from the community
12:10:14 <hagarth> ditto as above
12:10:17 <hagarth> #action hagarth to share results of community survey and solicit feature proposals from the community
12:10:30 <hagarth> AI - hagarth to carry forward discussion around backporting patches on gluster-devel
12:10:50 <hagarth> this is done, have initiated a thread on gluster-devel. please chime in with your thoughts.
12:10:59 <hagarth> AI - spot to investigate repetitive posts on social networking sites
12:11:21 <hagarth> spot is looking for some assistance here, maybe JustinClift can provide some help once he's back.
12:11:30 <hagarth> AI - spot will send out an email with upcoming Gluster events and meetups
12:11:50 <hagarth> not sure if spot is around atm, but don't recollect seeing this email
12:12:03 <hagarth> so .. #action spot will send out an email with upcoming Gluster events and meetups
12:12:23 <hagarth> whew, this  AI reviews almost seemed like a monologue today :)
12:12:39 <hagarth> #topic Gluster 3.6
12:13:00 <hagarth> raghu: updates?
12:13:19 <raghu> I have done 3.6.2beta2. Tarball is ready. waitinf for the rpms.
12:13:34 <raghu> s/waitinf/waiting
12:13:50 <hagarth> is hchiramm_ going to build the RPMs?
12:14:30 <hagarth> raghu: all further patches on release-3.6 now move to 3.6.3?
12:14:36 <ndevos> lala is planning to build the rpms, hchiramm_ to check and put them on download.gluster.org
12:15:06 <raghu> hagarth: yeah. All further patches on release-3.6 will be moved 3.6.3.
12:15:33 <hagarth> raghu: ok, can you please open a new tracker for 3.6.3 and move dependent bugs from the 3.6.2 tracker to that one?
12:15:35 <ndevos> hmm, Mac OSX will still be broken then, http://review.gluster.org/9145
12:15:40 <raghu> The good news is that all USS patches passed regression in the last regression that was triggered after last week's meeting and I have merged them too
12:15:53 <hagarth> I observed xavih also adding a dependency to 3.6.2 tracker today
12:16:02 <hagarth> ndevos: maybe we could have an early release of 3.6.3
12:16:10 <raghu> hagarth: sure
12:16:32 <kkeithley> It's not like we hear lots of screaming about Mac OS X
12:16:41 <kkeithley> Or do we?
12:16:46 <hagarth> #action raghu to open a 3.6.3 tracker and move dependent bugs from the 3.6.2 tracker to that.
12:17:10 <ndevos> no, I dont think so, we have very little Max OSX useres - probably because it just doesnt compile
12:17:11 <kkeithley> And the patch is there
12:17:21 <ndevos> *users
12:18:00 <hagarth> maybe we could produce a special build for mac with this patch?
12:18:02 <ndevos> raghu: maybe you can list the brokenness of OS X in the release notes?
12:18:26 <raghu> ndevos: sure. will do it
12:18:43 <ndevos> I think that could work - add the patch to any OS X builds we provide (do we?)
12:18:50 <jdarcy_> We could call it a remix.
12:19:21 <hagarth> ndevos: we can check what harsha and dennis would prefer for builds
12:19:48 <partner> if you build rpms why don't you do debs on the same go? its pretty annoying when each release comes with different set of packages and platforms
12:20:10 <partner> sounds like the process is fully manual?
12:20:16 <ndevos> hagarth: I dont know if they provide builds, or only make sure it compiles/works...
12:20:19 <kkeithley> We build debs for debian and Ubuntu
12:20:28 <kkeithley> Yes, but then so is building rpms
12:20:54 <partner> yes there are debs around when i ask long and loud enough :)
12:20:55 <kkeithley> Automation would be nice
12:20:59 <partner> indeed
12:21:13 <hagarth> can we form a working group for automation?
12:21:19 <ndevos> is there a demand for beta .deb's?
12:21:36 <kkeithley> semiosis is a volunteer. It's hard to twist his arm.  When it's apparent that he's not going to get to it then I do it.
12:21:39 <partner> ndevos: don't know about that but possibly it will keep bunch of beta testers out also
12:22:01 <ndevos> partner: yeah, I guess so
12:22:40 <hagarth> should we take this discussion further on -infra?
12:22:49 <ndevos> we have nightly builds for rpms, fully automated - something similar could be setup for debs?
12:22:59 <hagarth> ndevos: +1
12:24:10 <ndevos> https://forge.gluster.org/bugzappers/nightly-builds/blobs/master/README for the rpms, someone with .deb knowledge should be able to adapt that
12:24:25 <hagarth> #action carry forward discussion on automated builds for various platforms in gluster-infra ML.
12:25:01 <hagarth> anything more on 3.6?
12:25:08 <raghu> nope.
12:25:14 <hagarth> ok, moving on
12:25:30 <hagarth> #topic GlusterFS 3.5.x
12:25:59 <hagarth> ndevos: any updates here?
12:26:07 <ndevos> still waiting for reviews
12:26:13 <ndevos> no response on http://www.gluster.org/pipermail/gluster-devel/2015-January/043452.html yet
12:26:30 <ndevos> nothing planned to be released in the (very) near future
12:26:51 <hagarth> hmm, I think this chimes in well to the thread that I started
12:27:08 <hagarth> how do we get more traction for reviews/patches on release branches?
12:27:49 <ndevos> I'm wondering about that too...
12:28:08 <32NAAW3WY> o/
12:28:39 <hagarth> I wonder if we should follow the linux kernel model and have a well defined patch acceptance window
12:28:53 <hagarth> gothos: thanks for identifying yourself :)
12:29:08 <ndevos> hagarth: like, no negative reviews, just merge it?
12:29:17 <gothos> hagarth: yeah sorry, the bounces stole my nick not yet fully configured
12:29:25 <ndevos> but, we still need someone to do the backports then...
12:29:47 <hagarth> ndevos: yeah and review the backports too
12:30:12 <hagarth> gothos: no worries, good to have you around
12:30:15 <ndevos> I doubt that'll work for the stable branches
12:30:48 <ndevos> or would giving deadlines make developers more aware of posting patches and doing reviews?
12:31:18 <hagarth> ndevos: I think that might work, time bound model for releases works in quite a few communities
12:32:11 <ndevos> hagarth: from your email http://www.gluster.org/pipermail/gluster-devel/2015-January/043488.html
12:32:24 <ndevos> the #1 point already has the most difficulties, I think
12:32:38 <hagarth> ndevos: agree
12:33:03 <ndevos> getting reviews for things that have active maintainers is relatively easy
12:33:31 * ndevos looks at rdma changes and shakes his head
12:34:09 <hagarth> ndevos: right, how do we get our active maintainers to be more active? :)
12:34:52 <ndevos> well, doing backports is a little different task...
12:35:16 <hagarth> to answer that, one option would be to have more maintainers if the active maintainers do not have bandwidth
12:35:33 <ndevos> I tend to pick/clone important bugs, so the maintainers should be aware of the need for backports
12:35:34 <hagarth> the bottleneck for release branches is two fold:
12:36:00 <hagarth> 1. identifying what needs to be backported
12:36:21 <hagarth> 2. doing the backports and review + merge them
12:36:29 <ndevos> indeed
12:37:08 <raghu> true
12:37:08 <ndevos> the identifing is currently completely reactive - bugs get filed against a stable version, fix can be available in master
12:37:59 <hagarth> should we ask spot to double up as a program manager for these releases? ;)
12:38:38 <ndevos> maybe, but we definitely need someone that tries to keep an eye on it
12:38:59 <hagarth> ndevos: let me think about this a little more
12:39:29 <ndevos> probably we should include "is a backport of this patch needed?" in the review criteria
12:39:50 <aravindavk> ndevos: that is good idea
12:39:59 <jdarcy> +1
12:40:12 <raghu> +1
12:40:18 <hagarth> ndevos: +1, wonder if we can automate that into our gerrit workflow
12:41:12 <ndevos> hagarth: not fully at the moment, I think, but the reviewer could go to the bug and clone it for the (possibly) affected releases
12:41:47 <hagarth> ndevos: please chime in with these ideas on the thread, let us arrive at a consensus and document the process.
12:42:06 <ndevos> hagarth: sure, only just now noticed you sent an email about it :)
12:42:21 <hagarth> ndevos: thanks!
12:42:28 <ndevos> #idea probably we should include "is a backport of this patch needed?" in the review criteria
12:42:56 <hagarth> ndevos: can you also publish a schedule for 3.5.4? that might spur some action for reviews and backports
12:43:22 <ndevos> sure, but I'll need to think about that a little more
12:43:37 <hagarth> ndevos: sure
12:43:40 <hagarth> anything more on 3.5?
12:43:49 <ndevos> not from me
12:44:07 <hagarth> ok, let us move on
12:44:13 <hagarth> #topic GlusterFS 3.4
12:44:25 <hagarth> kkeithley: any updates here?
12:44:34 <kkeithley> no, nothing per se
12:44:36 <hagarth> the memory leak probably needs investigation still
12:44:49 <kkeithley> just wrapping up the first pass through memory leak
12:45:10 <hagarth> kkeithley: ok
12:45:16 <kkeithley> there's no leak with all performance.* off
12:45:32 <hagarth> kkeithley: interesting, anything with md-cache only being off?
12:45:53 <kkeithley> and just finishing the last run of single performance enabled. no leaks, at least not on the order of when everything is enabled
12:46:24 <hagarth> kkeithley: ok, let us discuss this further in next week's meeting
12:46:36 <kkeithley> okay
12:46:38 <hagarth> anything more on 3.4?
12:46:45 <kkeithley> no
12:46:53 <hagarth> ok, moving on
12:47:01 <hagarth> #topic GlusterFS next
12:47:23 <jdarcy> Nothing to report.  I'll try to gather status on the pieces for next week.
12:47:30 <hagarth> jdarcy: ok
12:47:47 <hagarth> on the topic of 3.7, there are a few new feature pages
12:48:10 <hagarth> the notable one being Netgroups style exports for gNFS
12:48:37 <hagarth> this is essentially a forward port of work that was done in Facebook on 3.4.x
12:48:48 <jdarcy> Do we need a separate agenda item for Manila/Cinder?  They seem to be proposing features completely outside of the upstream release cycle.
12:49:01 <hagarth> thanks to ndevos and jiffin for their relentless efforts in forward porting
12:49:24 <hagarth> jdarcy: good point, I will invite the OpenStack developers to the next meeting
12:49:51 <hagarth> or rather have them attend these meetings so that they understand community workflow better :)
12:50:29 <ndevos> oh, I would like to see something like portable POSIX ACLs in 3.7 too - its more an improvement and not a new feature
12:50:53 <ndevos> and, I'll be working on that, would it need a feature page?
12:51:04 <hagarth> ndevos: yes, that would be helpful
12:51:14 <ndevos> oka
12:51:15 <ndevos> y
12:51:43 <hagarth> any questions or thoughts on 3.7, 4.0 ?
12:51:44 <jdarcy> If they want to get anything in upstream, they should seriously consider attending these meetings.  ;)
12:51:50 <ndevos> #action ndevos to create a feature page for more portable POSIX ACLs
12:51:54 <hagarth> jdarcy: +1 :)
12:52:17 <hagarth> figure nothing more about $TOPIC, moving on
12:52:31 <hagarth> #topic other agenda items
12:52:41 <hagarth> #topic Upcoming talks at conferences
12:53:00 <hagarth> I think quite a few of us have upcoming talks at various conferences
12:53:20 <hagarth> can we add it to the etherpad so that it would be easy for spot or anybody to provide publicity to these talks?
12:53:48 <ndevos> or maybe start a separate etherpad for conferences/talks?
12:54:09 <hagarth> ndevos: +1
12:54:15 <kkeithley> and highlight upcoming conferences that we ought to present at!
12:54:35 <kkeithley> s/we/one of us/
12:54:46 <ndevos> yes, and include dates for related talks, like jdarcy's Red Hat Storage one
12:55:06 <hagarth> kkeithley: certainly, per last week's meeting spot is also looking at arriving at a event calendar that we should be attending.
12:55:18 <hagarth> ndevos: the hangout you mean?
12:55:37 <jdarcy> A real calendar on the site would be cool.  Great way to show how active we all are.
12:55:57 <hagarth> jdarcy: +1
12:55:59 <ndevos> hagarth: maybe yes, I did not see it yet :-/
12:56:30 <hagarth> ndevos: ah ok, I think we should send out the hangout link on gluster mailing lists
12:56:43 <hagarth> that will also give users a feel of what 4.0 is going to be about
12:56:58 <ndevos> yes, we should, I only noticed it in my twitted feed
12:57:04 <ndevos> *twitter
12:57:18 <hagarth> hchiramm_ and others are also working on a Gluster meetup in Pune
12:57:42 <hagarth> here's the link - http://www.meetup.com/glusterfs-India/events/219327028/
12:58:12 <hagarth> quite a bit of action
12:58:19 <hagarth> anything else for today?
12:58:50 <ndevos> https://public.pad.fsfe.org/p/gluster-events
12:59:01 <hagarth> ndevos: awesome!
12:59:04 <ndevos> please update that with your conferences ;)
12:59:21 <hagarth> I think that should be it for now.
12:59:29 <hagarth> Thanks for attending today and talk to you all next week!
12:59:38 <jimjag> cheers all
12:59:44 <hagarth> #endmeeting