12:01:05 <hagarth> #startmeeting
12:01:05 <zodbot> Meeting started Wed Jun 10 12:01:05 2015 UTC.  The chair is hagarth. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:05 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:12 <hagarth> #chair ndevos, hagarth
12:01:12 <zodbot> Current chairs: hagarth ndevos
12:01:28 <hagarth> #agenda - https://public.pad.fsfe.org/p/gluster-community-meetings
12:01:35 <hagarth> #topic Roll Call
12:01:41 <hagarth> who do we have here today?
12:01:49 * jdarcy is
12:02:03 * atinmu is partially there
12:02:08 * hchiramm_ is here
12:02:35 <hagarth> ok, I think some more folks will get active during the course of the meeting.
12:02:44 <hagarth> #topic AI review from last week
12:02:52 <hagarth> AI - JustinClift will have something to say about the Gluster Forge v2 on the June 10 meeting
12:03:01 * raghu` is here
12:03:08 <hagarth> I know that JustinClift has been busy migrating repos from forge to github
12:03:19 * msvbhat is here
12:03:24 <hagarth> not sure about the current status
12:03:43 <hagarth> but I expect that we will retire forge soon
12:03:59 <hagarth> hope none of us have concerns about retiring forge..
12:04:23 * overclk is here..
12:04:35 <hagarth> next AI - hchiramm_ will send out a call (email and blog?) for volunteers for helping with the packaging  [TBD]
12:04:35 <hchiramm_> hagarth, are all the projects getting migrated to github?
12:04:41 <hagarth> hchiramm_: yes
12:04:49 <hchiramm_> hagarth, ok.
12:04:52 * spot is here, sorry for being a bit late
12:04:53 <hchiramm_> reg#AI, hagarth, will do this week ..
12:04:56 * pranithk here
12:05:26 <hagarth> hchiramm_: ok
12:05:31 <hagarth> #action hchiramm_ will send out a call (email and blog?) for volunteers for helping with the packaging
12:05:48 <hagarth> #action JustinClift to provide an update on forge v2
12:06:04 <hagarth> next AI - tigert summarize options for integrating a calendar on the website, call for feedback
12:06:11 * partner meh
12:06:26 <hagarth> tigert is unable to attend this week. Update on this AI is available in the agenda etherpad.
12:06:41 <hagarth> Pasting that here to make it simpler
12:06:59 <hagarth> (I  am unable to attend this week, so summarizing here: I have not had time  to look into this in depth yet. However, we do have the Red Hat events  calendar that is pulling data from yml files in community.redhat.com: http://community.redhat.com/events/ and the event files are in https://github.com/OSAS/rh-events - maybe we could do something based on this? Information about local events needs to be collected for this to work at all,
12:07:14 <hagarth> if you have more thoughts on this one, please talk to tigert.
12:07:28 <hagarth> next AI is more of a reminder
12:07:33 <hagarth> AI - EVERYONE Announce participation and times of Gluster related presentations on the community lists
12:07:54 <hagarth> if you are talking about Gluster anytime in the near future, please drop a mail on the community lists
12:08:13 <hagarth> that is about AIs from last week
12:08:27 <hagarth> next AI - GlusterFS 3.7
12:08:30 <hagarth> oops
12:08:33 <hagarth> #topic GlusterFS 3.7
12:09:01 <hagarth> 3.7.1 was released last week. We have had some patches accrued in gerrit and have also merged patches in release-3.7
12:09:15 <hagarth> I expect that we will have 3.7.2 shortly, maybe sometime next week.
12:09:42 <hagarth> if you have fixes that you want to see in 3.7.2, please try to get them in by early next week.
12:10:28 <overclk> hagarth, provided gerrit dosen't block us from reviews ;)
12:10:34 * ndevos looks around
12:10:41 <hagarth> #info GlusterFS 3.7.2 to be released in the week of June 15
12:10:50 <hagarth> overclk: absolutely .. is gerrit acting up now?
12:10:52 <overclk> it's been insanely slow today
12:11:11 <hagarth> overclk: +1, I think we should do an infra discussion at the end of the meeting today.
12:11:12 <msvbhat> Yeah, for me as well. Extremley slow
12:12:01 <hagarth> I could not login to gerrit this afternoon
12:12:10 <hagarth> with some luck and persistence I finally could :)
12:12:12 <overclk> hagarth, sure. if you ask me, "gerrit" the real blocker :)
12:12:15 <atinmu> hagarth, its damn slow
12:12:17 * overclk ducks..
12:12:29 <hagarth> s/gerrit/gerrithub/ ? :)
12:12:45 <hagarth> anyway let us discuss this later and try to evolve a quick fix for these issues.
12:12:52 <hagarth> overclk: good one!
12:12:58 <hagarth> any more questions on 3.7?
12:13:45 <hagarth> oh btw, we still have some presentations/demos of features to be done for 3.7.
12:14:09 <hagarth> https://public.pad.fsfe.org/p/gluster-3.7-hangouts has the list
12:14:38 <hagarth> if you are interested in doing a hangout with the broader community on your features, please drop a note on the mailing lists preferably with an invite
12:15:10 <hagarth> that's about 3.7
12:15:15 <hagarth> ndevos: do you want to take over
12:15:20 <hagarth> ?
12:15:30 <ndevos> hagarth: sure, thanks!
12:15:40 <ndevos> #topic GlusterFS 3.6
12:15:46 <raghu`> Got the ack for the pending patches today. I have merged the pending patches in release-3.6 branch. Going to make 3.6.4beta2 today.
12:16:11 <ndevos> raghu`: you really want to do a beta2, or just release it?
12:16:35 * ndevos will probably not do any beta's for 3.5 anymore, nobody provides feedback on those
12:16:51 <raghu`> ndevos: I can do it. But just to make sure that if there are any issues in beta2, then v can fix it before making GA
12:17:25 <raghu`> ndevos: Probably for 3.6.5 I can skip beta and directly do GA
12:17:25 <hagarth> raghu`, ndevos: I think we should define some tests that need to be run pre-GA.
12:17:33 <ndevos> raghu`: sure, its your decision :)
12:18:18 <ndevos> hagarth: we should have those tests in Jenkins, I do very limited testing on 3.5 myself
12:18:24 <hagarth> ndevos: +1
12:18:52 <raghu`> ndevos: +1
12:19:10 <hagarth> ndevos: let us probably launch a discussion on gluster-devel? msvbhat and RaSTar_away have expressed strong interest on this.
12:19:42 <ndevos> #info glusterfs 3.6.4beta2 will be made available later today, feedback from testers wanted
12:20:06 <hagarth> any other pending bugs in 3.6 before we do 3.6.4?
12:20:26 <raghu`> hagarth: nope. All the pending patches have been taken in.
12:20:27 <ndevos> hagarth: yes, you want to start the discussion, or shall we ask msvbhat to pick that up?
12:21:07 <hagarth> ndevos: anything works .. maybe we'll have an AI on msvbhat and rastar
12:21:25 <raghu`> there are no patches (except one patch from xavih, but I think that patch has not gone into master yet)
12:21:51 <hagarth> raghu`: right .. maybe we should do a quick check of bugs logged on 3.6.x in bugzilla and see if something's worth fixing?
12:21:56 <ndevos> #action msvbhat to start a discussion about more featured automated testing (distaf + jenkins?)
12:21:59 <raghu`> if there are no bugs in beta2, then I am planning to call it 3.6.4
12:22:11 <itisravi> raghu`: I have one bug
12:22:24 <raghu`> hagarth: sure.
12:22:31 <ndevos> #info itisravi has a bug
12:22:51 <itisravi> raghu`:  I haven't sent http://review.gluster.org/10776 on 3.6. I will send it today, it is good to have it in.
12:22:53 <raghu`> itisravi: can you tell me the bugid?
12:23:14 <itisravi> raghu`: I did not backport it to 3.6 WIll do it today ASAP.
12:23:34 <raghu`> itisravi: hmm. Let me take a look at it.
12:23:40 <ndevos> #info Bugs blocking 3.6 are listed on https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&id=glusterfs-3.6.4&hide_resolved=1
12:24:38 * itisravi reminds himself to send important AFR patches from 3.7 to 3.6 as well.
12:24:40 <ndevos> raghu`: are we done with 3.6?
12:24:55 <ndevos> itisravi: *and* 3.5 too!
12:25:05 <raghu`> ndevos: yeah. The list says almost all the bugs are in MODIFIED state.
12:25:17 <itisravi> ndevos: bug doesn't occur in 3.6 as it is afr v1 :-)
12:25:25 <raghu`> ndevos: there is one bug which is in NEW state. Have to take a look at it. Probably for 3.6.5
12:25:51 <ndevos> itisravi: sure, but you never know if you find anything that would affect 3.5, if you do, please send a backport :)
12:26:01 <itisravi> ndevos: Will do :-)
12:26:41 <raghu`> that is all I had for 3.6.4
12:26:55 <ndevos> raghu`: anything in NEW would be pushed out to 3.6.5, I think?
12:27:03 <ndevos> okay, yes, lets move on
12:27:07 <raghu`> ndevos: yeah. sure
12:27:16 <ndevos> #topic GlusterFS 3.5
12:27:44 <ndevos> glusterfs 3.5.4 has been release last week and packages have been provided
12:28:05 <hagarth> ndevos: good stuff on getting 3.5.4 out!
12:28:06 <ndevos> there are 1 or 2 fixes requested for the next 3.5 update
12:28:35 <ndevos> I guess 3.5.5 can be done next week, or the week after - but surely in June
12:28:52 <ndevos> #info glusterfs-3.5.5 is expected later in June
12:29:05 * msvbhat was afk for sometime to get Tea. Now reading chatlogs
12:29:22 <ndevos> #info Bugs planned for 3.5.5 are listed on https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&hide_resolved=1&id=glusterfs-3.5.5
12:29:25 <hagarth> msvbhat: we managed to add multiple AIs on you while you were away ;)
12:29:50 <ndevos> any questions about 3.5?
12:30:17 <ndevos> nope.
12:30:18 <ndevos> #topic GlusterFS 4.0
12:30:38 <ndevos> #info Bugs/features for 4.0 are collected through https://bugzilla.redhat.com/showdependencytree.cgi?id=glusterfs-4.0
12:30:46 <ndevos> jdarcy: anything to add?
12:30:48 <msvbhat> hagarth: Yeah, Saw that now :) will send
12:30:55 <jdarcy> Nope.
12:31:14 <ndevos> anyone else has questions about 4.0?
12:31:38 <hagarth> I have a thought
12:31:51 <ndevos> very nice, care to share?
12:31:57 <hagarth> I think we need to update gluster.org with planned roadmaps for all gluster.next releases
12:31:58 <jdarcy> It happens.  I usually try to get rid of those.
12:32:35 <ndevos> hagarth: I think someone else had that thought already, I'm pretty sure we have a bug against the website for that
12:33:04 <jdarcy> That's a very good idea.  First, we should probably agree on what's in and out for 4.0
12:33:44 <ndevos> well, maybe not, I only see bug 976129: Upstream Gluster could do with a decent Feature Matrix on the website
12:33:59 <hagarth> jdarcy: yes, let us do that. Maybe keep the scope minimal and have something like "more to follow" as a placeholder.
12:34:33 <ndevos> something like the Planning40 page in the wiki?
12:34:59 <jdarcy> It's the best place to start, but at this point I think it's unlikely everything on that list will be achievable.
12:35:05 <ndevos> #idea update gluster.org with planned roadmaps for all gluster.next releases
12:35:34 <hagarth> ndevos: maybe convert that to an AI?
12:35:53 <ndevos> hagarth: for you or jdarcy?
12:36:10 <hagarth> ndevos: happy to take that and run it through you folks before going live
12:36:23 <jdarcy> Sounds good.
12:36:49 <ndevos> #action hagarth will start with planned roadmaps for all gluster.next releases
12:37:19 <ndevos> and for completeness, proposed features are on http://www.gluster.org/community/documentation/index.php/Planning40
12:37:34 <ndevos> hagarth: you have more thoughts?
12:38:07 <ndevos> I guess he's lost in his thoughts now...
12:38:10 <ndevos> #topic Open Floor / Bring Your Own Topic
12:38:22 <spot> I've got one.
12:38:26 <ndevos> #info Weekly reminder to announce Gluster attendance of events: https://public.pad.fsfe.org/p/gluster-events
12:38:48 <ndevos> spot: a thought or a topic? go for it!
12:38:50 <spot> Red Hat is now actively recruiting for a full-time Gluster Community Manager.
12:38:57 <spot> http://jobs.redhat.com/jobs/descriptions/gluster-community-manager-job-1-5457472
12:39:12 <ndevos> #info Red Hat is actively recruiting for a full-time Gluster Community Manager
12:39:15 <spot> If you (or someone you know) is interested, have them go ahead and apply.
12:39:54 <ndevos> spot: quick description of what is expected from a Community Manager?
12:40:03 <spot> ndevos: sure.
12:40:07 <jdarcy> If I have to travel to Dustin's house and draft him, will Red Hat reimburse me for the expense?
12:40:27 <spot> Basically, the role involves helping out with all of the things that aren't writing code. :)
12:40:47 <spot> Marketing, promotion, events, materials, goal tracking, growth.
12:41:03 <hagarth> jdarcy: maybe the gluster board will reimburse :D
12:41:34 <spot> Comes with a budget and control over how it's spent.
12:42:13 <spot> That's the super high level description. :)
12:42:43 <spot> Make Gluster super successful, meet interesting people, recruit new ants to Gluster. :)
12:42:45 <ndevos> I guess travelling to events is involved, does it matter what the office location is?
12:43:00 <spot> ndevos: nope.
12:43:53 <ndevos> okay, thanks spot!
12:44:24 <ndevos> #topic Infra issues: Gerrit terribly slow
12:44:48 <hagarth> I think we need more volunteers to help with infra issues like:
12:44:49 <ndevos> hagarth: you added that after overclk, msvbhat and atinmu complained
12:44:54 <hagarth> 1. gerrit being terribly slow
12:45:05 <hagarth> 2. netbsd regression infra acting up
12:45:59 <hagarth> how do we want to debug these problems?
12:46:10 <hagarth> I don't think we have RCA for any of these problems.
12:46:28 <ndevos> those "problems" are also not really well defined
12:46:30 <hagarth> even today we lost a bunch of netbsd regression runs due to network issues IIUC
12:46:44 <hagarth> ndevos: yes, we need to start by defining them better.
12:47:03 <hagarth> am looking for ideas on what we can do in the near term
12:47:16 <ndevos> hagarth: should we not move gerrit to a rackspace system, or on a server that we host ourselves?
12:47:32 <hagarth> ndevos: do we think the problem is with iweb?
12:47:45 <ndevos> hagarth: at least their dns acts up often enough
12:47:46 <hagarth> ndevos: I wonder if we should just move to gerrithub
12:48:03 <jdarcy> I thought somebody had already created a functional clone of r.g.o elsewhere.
12:48:22 <ndevos> hagarth: gerrithub would be an other option, less sysadmin work for us would probably be good
12:48:28 <hagarth> jdarcy: in all probability it was JustinClift
12:48:45 <hagarth> ndevos: the only hitch there could be about migrating our existing CR patches.
12:49:03 <ndevos> csim: would you mind if we drop our Gerrit instance and use gerrithub.io instead?
12:49:06 <hagarth> i am not sure if gerrithub allows us to import patches from gerrit
12:49:39 <csim> ndevos: then that mean having 2 different user db/accounts system, no ?
12:49:54 <ndevos> hagarth: the gerrithub people are very helpful, they even are subscribed to the nfs-ganesha-devel list and jump on gerrit issues/questions :_
12:50:09 <ndevos> csim: it integrates with the GitHub accounts
12:50:16 <hagarth> ndevos: interesting, can we open a discussion with them and see if they can help with migration?
12:50:31 <ndevos> hagarth: we can surely ask them if they can help with that
12:50:52 <csim> ndevos: so my 2nd question is "can we be sure that gerrithub is not going the way of gitorious later" :)
12:51:10 <csim> ie, if we can resaonably think they are going well financially
12:51:24 <csim> and can we migrate away if there is a problem
12:51:29 <csim> (and how)
12:51:30 <ndevos> csim: gerrithub is the free (but also commercial) product of the people that develop Gerrit, I think
12:51:53 <csim> ndevos: so was gitorious at that time
12:52:44 <ndevos> csim: I dont know, but would we not have the same issue when we just stick with Gerrit?
12:53:47 <csim> ndevos: well, if we stick with gerrit, we have access to the data to migrate away :)
12:53:53 <csim> and we would still have the code
12:54:25 <csim> while, in the case the company crash and we have 1 month to move ( hypothetical ), we would need data and need a backup plan
12:54:34 <hagarth> ndevos: maybe we should ask the gerrithub folks these questions?
12:54:52 <hagarth> i think these might be FAQs for them
12:55:02 <ndevos> hagarth: I guess so, maybe csim wants to contact them>
12:55:03 <ndevos> ?
12:55:05 <csim> I do not think there is a FAQ "are you sustainable" :)
12:55:33 <overclk> csim, :)
12:55:35 <ndevos> csim: no, but "is there a way to get our posted patches?" and the like?
12:55:35 <hagarth> csim: put it differently, it indeed would be :)
12:55:46 <csim> ndevos: yep
12:55:55 <ndevos> csim: care to ask them?
12:56:02 <csim> let's discuss on the ML, as maybe there is more questions coming to mind ?
12:56:06 <csim> ndevos: I could, yeah
12:56:48 <ndevos> #action csim will send an email about the idea to migrate to Gerrithub.io, and collect questions for relaying
12:57:06 <hagarth> we still have to investigate the netbsd connectivity problem
12:57:09 <msvbhat> I'm not entirely sure it fits our requirement. But anyone heard of Travis CI? http://docs.travis-ci.com/user/languages/c/
12:57:22 <csim> msvbhat: travis run stuff on ubuntu, afaik
12:57:37 <csim> and I guess they may not appreciate us meddling with their kernel
12:57:41 <ndevos> msvbhat: yeah, that would be a Jenkins replacement - but I dont think it can do multi-system tests
12:58:01 <csim> ( like meddling, I mean "crashing" )
12:58:21 <ndevos> hagarth: what NetBSD issues are left?
12:58:28 <msvbhat> Okay, I haven't done any RD on it anyway.
12:58:30 <csim> also, we have bought servers for jenkins, so if someone want to move away from jenkins, please take care of reimbursing the servers since they would be unused
12:58:50 <hagarth> ndevos: connectivity to NetBSD VMs gets bad often and all regression tests bail out
12:58:58 <ndevos> csim: we need to know how/when we can start using those servers :)
12:59:26 <ndevos> hagarth: connectivity between iweb and rackspace, right?
12:59:27 <csim> ndevos: they are plugged and stuff, I need to ask for public ip and this kind of things
12:59:36 <hagarth> csim: +1 to get that done
12:59:40 <hagarth> ndevos: right
12:59:50 <csim> but after 1 month of time off, stuff did accumulate
13:00:08 <csim> so what kind of connectivity issue are we speaking off ?
13:00:22 <ndevos> #action csim to pickup the requesting of public IPs and those kind of things for our new servers
13:00:37 <hagarth> csim: there is some kind of network heartbeat messages that get exchanged between jenkins master and slaves
13:01:01 <ndevos> csim: I blame awkward dns behaviour for many problems from build.gluster.org towards our Jenkins slaves ;-)
13:01:14 <csim> ndevos: what about adding a local cache ?
13:01:16 <hagarth> when there are some nw connectivity problems, the heartbeat excahnges fail and all regression tests scheduled to run get dequeued
13:01:35 <hagarth> ndevos: certainly I have seen dns problems too
13:01:51 <ndevos> csim: well, how do we know a local dns caches the correct IP?
13:02:40 <csim> ndevos: does the ip change so much ?
13:03:32 <ndevos> csim: I dont know, the slaves are rackspace VMs
13:03:33 <csim> now, dns problem might just be a symptom of network problem
13:03:40 <csim> ndevos: they don't change much
13:04:03 <ndevos> well, dns sometimes gets a wrong IP, and that breaks things
13:04:12 <csim> oh
13:04:16 <csim> W T F ?
13:04:51 <csim> let's try to set a local cache
13:05:24 <csim> Could not retrieve mirrorlist http://mirrorlist.centos.org/?release=6&arch=x86_64&repo=os error was
13:05:27 <csim> 14: PYCURL ERROR 6 - "Couldn't resolve host 'mirrorlist.centos.org'"
13:05:53 <ndevos> csim: we should follow up on that on the infra list
13:05:57 <csim> ndevos: yep
13:06:04 <ndevos> #topic The Gluster Board
13:06:10 <ndevos> hagarth: yours again?
13:06:28 <hagarth> ndevos: yes
13:06:43 <hagarth> JoeJulian is looking for some help to re-constitute an interim board
13:07:13 <hagarth> If you know of some Gluster ecosystem users/developers that can be part of the board, please encourage them to reach out to JoeJulian
13:07:30 <hagarth> #link https://twitter.com/JoeCyberGuru/status/600731358067994625
13:07:43 <hagarth> ndevos: that's about it
13:08:01 <ndevos> #info If you know of some Gluster ecosystem users/developers that can be part of the board, please encourage them to reach out to JoeJulian
13:08:10 <ndevos> hagarth: ok, thanks!
13:08:41 <ndevos> if there are no other topics from anyone else, we'll finish up for today
13:09:14 <hagarth> ndevos: thanks!
13:09:17 <ndevos> #endmeeting