12:00:31 <atinm> #startmeeting Gluster Community weekly meeting
12:00:32 <zodbot> Meeting started Wed Dec  9 12:00:31 2015 UTC.  The chair is atinm. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:32 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:00:32 <zodbot> The meeting name has been set to 'gluster_community_weekly_meeting'
12:00:39 <atinm> Greetings folks!
12:00:47 <atinm> Its time to have our weekly meeting running
12:01:01 * partner greets the earthlings and stays at the background
12:01:07 * hgowtham is here
12:01:08 <atinm> #topic Roll Call
12:01:12 * jiffin is here
12:01:13 * obnox is here
12:01:49 <atinm> We'll wait for couple of minutes to get people settled :)
12:01:55 * rjoseph is here
12:02:10 * skoduri is here
12:02:39 <atinm> who else do we have here today?
12:02:53 * anoopcs is here
12:02:57 * rafi is here
12:03:02 <ggarg> ggarg is here :)
12:03:52 <atinm> All right, the wait is over, let's get into the business
12:03:58 <atinm> #topic AIs from last week
12:04:10 <atinm> ndevos send out a reminder to the maintainers about more actively enforcing backports of bugfixes
12:04:23 <atinm> do we have ndevos here?
12:04:33 <atinm> seems unlikely though
12:04:54 * Manikandan arrives bit late
12:05:14 <atinm> Will carry forward this AI for next week
12:05:23 <atinm> #action ndevos to send out a reminder to the maintainers about more actively enforcing backports of bugfixes
12:05:40 <atinm> Next one is raghu to call for volunteers and help from maintainers for doing backports listed by rwareing to 3.6.8
12:06:23 <atinm> Anyone recollects seeing this mail from raghu?
12:06:37 <jiffin> atinm: ndevos is on vacation
12:06:52 <atinm> #action raghu to call for volunteers and help from maintainers for doing backports listed by rwareing to 3.6.8
12:07:05 <atinm> jiffin, I know, was just trying my luck ;)
12:07:14 <jiffin> atinm: :)
12:07:17 <atinm> Moving on
12:07:18 <atinm> rafi1 to setup a doodle poll for bug triage meeting
12:07:29 <atinm> rafi, I believe its done
12:07:49 <rafi> atinm: yes, after  long time ;)
12:07:50 <atinm> rafi, any plan on when are you going to announce the result?
12:07:57 <hgowtham> atinm, i think only a few cast their vote
12:08:05 <rafi> atinm: after two weeks'
12:08:25 <atinm> rafi, hgowtham : an reminder after a week probably on this
12:08:33 <rafi> atinm: december 22
12:09:02 <hgowtham> yes okie atinm
12:09:10 <atinm> #info bug triage meeting doodle poll result to be announced on December 22, need more votes
12:09:24 <atinm> Next is rastar and msvbhat to publish a test exit criterion for major/minor releases on gluster.org
12:09:32 <atinm> I see bunch of updates in the etherpad
12:09:36 * rafi would expect to see people casting the vote after the meeting
12:09:49 <atinm> #info agenda is right here https://public.pad.fsfe.org/p/gluster-community-meetings
12:10:06 <atinm> rastar, msvbhat : you guys wanna talk about it?
12:10:34 <atinm> I'd take it as no, so carrying it forward for completion
12:10:36 <Manikandan> msvbhat is not here, in a meeting I guess
12:10:39 <atinm> #action rastar and msvbhat to publish a test exit criterion for major/minor releases on gluster.org
12:10:49 <atinm> kshlm & csim to set up faux/pseudo user email for gerrit, bugzilla,  github
12:11:03 <atinm> csim, any updates on this?
12:11:24 <atinm> #action kshlm & csim to set up faux/pseudo user email for gerrit, bugzilla,  github
12:12:05 <atinm> For the next AI on NetBSD regression failure on 3.6, hagarth and rastar volunteered to try backporting the regression framework changes
12:12:13 <atinm> So I am just ignoring it
12:12:41 <atinm> Next action was on rastar to call for volunteers for 3.7.7 release-manager
12:12:58 <atinm> I see its done however do we have a 3.7.7 manager by now?
12:13:02 <pranithk> atinm: talking to rastar
12:13:29 <atinm> pranithk, do you want to volunteer for it?
12:13:45 <pranithk> atinm: rastar says, vijay is deciding based on that random spin to decide the volunteer
12:13:59 <pranithk> atinm: I mean hagarth
12:14:00 <atinm> pranithk, ok
12:14:12 <atinm> #action hagarth to decide on 3.7.7 release manager
12:14:25 <atinm> Moving on
12:14:27 <atinm> amye to get on top of disucssion on long-term releases.
12:15:06 <atinm> I see some progress made on this as per the etherpad
12:15:17 <atinm> Carrying it forward for next week
12:15:28 <atinm> #action amye to get on top of disucssion on long-term releases.
12:15:44 <atinm> the next AI is a redundant one, so skipping it
12:16:05 <atinm> hagarth_ to     start a thread on review backlog
12:16:19 <atinm> I see it has been initiated
12:16:30 <atinm> Moving to next one
12:16:32 <atinm> atinm, kshlm to review IPv6 patchset
12:17:01 <atinm> kshlm has reviewed it and once it passes the regressions should be ready for merge
12:17:25 <atinm> The last but not the least is hagarth to post Gluster Monthly News this week
12:17:39 <atinm> Since we don't have hagarth for today carrying it forward
12:17:47 <atinm> #action hagarth to post Gluster Monthly News this week
12:18:05 <anoopcs> atinm, I think IPv6 bug fix patch has passed the regression.
12:18:21 <atinm> anoopcs, we will merge it then :)
12:18:36 <atinm> #topic GlusterFS 3.7
12:18:51 <anoopcs> atinm, need to wait for NetBSD regression.
12:19:00 <atinm> Anyone has anything to say on here?
12:19:37 <atinm> I take it as 'No' so moving forward
12:19:49 <atinm> #topic GlusterFS 3.6
12:19:55 <atinm> raghu and others ?
12:20:11 <raghu> I have to create 3.6.8 bug id.
12:20:15 <atinm> probably backports are welcome would be a nice statement here :)
12:20:27 <atinm> #info raghu to create 3.6.8 tracker
12:20:32 <raghu> but there is no 3.6.8 field in bugzilla
12:20:50 <raghu> I am waiting for hagarth to create the field
12:20:51 <atinm> raghu, can you get in touch with hagarth for this?
12:21:03 <raghu> I have already done it. I will follow up with him
12:21:15 <raghu> And yes, backports are welcome ;)
12:21:24 <atinm> #action hagarth to create 3.6.8 for bugzilla version
12:21:31 <atinm> Thanks raghu
12:21:32 <kkeithley_> until we release 3.6.8 nobody can file bugs against it anyway.  Usually I ask for things like 3.6.x right around the release time.
12:22:05 <atinm> kkeithley_, makes sense
12:22:20 <atinm> Anything else on 3.6 ?
12:22:29 <raghu> kkeithley_: then probably I will create the tracker bug right away and set some fields such as target milestone to 3.6.8 when it becomes available
12:23:00 <raghu> atinm: Probably I might need a release manager for 3.6.8 as I would not be around during 20th week
12:23:09 <kkeithley_> indeed, you may create the tracker now.  There is no more target release or target milestone fields
12:23:23 <raghu> I can send a mail to mailing list asking for the same
12:23:26 <kkeithley_> s/is no/are no/
12:23:32 <raghu> kkeithley_: sure
12:23:38 <atinm> #action community needs to find out 3.6.8 release manager
12:23:55 <atinm> raghu, that would be great
12:24:18 <raghu> atinm: sure. Will ask in the mailing list
12:24:22 <atinm> #action raghu to ask for volunteers for release manager for 3.6.8
12:24:27 <atinm> Moving on
12:24:48 <atinm> I intend to skip GlusterFS 3.5 as ndevos is not around, any objection?
12:25:13 <atinm> I'd take it as 'No'
12:25:16 <atinm> #topic GlusterFS 3.8
12:25:17 <hgowtham> no objection from me
12:26:04 <atinm> I see bunch of design ideas landing in devel which is a good news for the progress made
12:26:20 <atinm> Any specific things you guys want to bring up here?
12:26:51 <atinm> again a 'NO' ?
12:27:10 <atinm> Moving on
12:27:13 <atinm> #topic GlusterFS 4.0
12:27:44 <atinm> Let me talk about GlusterD 2.0 quickly here
12:28:16 <atinm> We have been focusing on coming up with a simple txn framework both design and prototype and we were working on that and will continue to do the same for this month
12:28:42 <atinm> From NSR side work is going on Journaling
12:29:15 <atinm> DHT2 team is focusing on making directory operations work in their PoC
12:29:51 <atinm> and from Gluster Eventing side a bunch of discussion happened over the mailing list and samikshan is exploring all these ideas and evaluating them
12:30:00 <atinm> That's all what I have from 4.0 side
12:30:06 <jiffin> atinm: on 3.8 ,last week  hagarth organized meeting with feature owner and decide to have feature freeze on mid-last  Jan 2016
12:30:12 <kkeithley_> two reminders for 4.0.  first is the new symbol-check regression test is in now to check that the sys_foo system call wrappers are used. Second is that we want new features to be run through something like cppcheck, clang analyze, or coverity to make sure we don't introduce bugs
12:30:19 <kkeithley_> And valgrind to check for leaks.
12:30:39 <kkeithley_> Sometime towards the end of the feature development cycle
12:31:14 <atinm> kkeithley_, point noted, however would it be possible for you to share these warnings in devel ML as well?
12:31:24 <kkeithley_> sure
12:31:36 <atinm> #info 3.8 feature freeze to happen on mid-last  Jan 2016
12:32:35 <jiffin> atinm: and also hagarth will organize another meeting on Jan first week
12:32:47 <atinm> #action kkeithley_ to send a mail about using sanity checker tools in the codebase
12:32:48 <jiffin> to know the status of each feature
12:32:55 <atinm> jiffin, cool
12:33:23 <atinm> #info Another follow up meeting on 3.8 to take place on first week of January, 2016
12:33:37 <atinm> Anything else from 3.8/4.0 side?
12:33:56 <atinm> Allright, lets move to our last topic
12:34:02 <atinm> #topic Open Floor
12:34:43 <atinm> I wanted to bring one here
12:34:45 <atinm> Any conclusion on http://www.gluster.org/pipermail/gluster-devel/2015-November/thread.html#47125
12:35:29 <kkeithley_> there's a lot there, which one?
12:35:46 <jiffin> atinm: yes, which one??
12:36:04 <obnox> atinm: afaic - fast forward only is the way to go
12:36:16 <atinm> Today I merged couple of independent patches and then it broke the build, this is because our submit type is cherry pick
12:36:49 <atinm> kkeithley_, jiffin :  Any conclusion on http://www.gluster.org/pipermail/gluster-devel/2015-November/thread.html#47125
12:37:07 <kkeithley_> that shows me a page with 100 mails on it
12:37:18 <atinm> obnox, yes that's what the proposal says
12:37:24 <obnox> kkeithley_: 47125
12:37:25 <atinm> http://www.gluster.org/pipermail/gluster-devel/2015-November/047125.html
12:37:35 <obnox> that is the top post with a few replies
12:37:49 <obnox> atinm: sry, i mistook it as one of the alternatives
12:38:20 <atinm> Probably I will follow up with rastar to continue the discussion
12:38:28 <rastar> yes, we need to decide on that
12:38:28 <obnox> atinm: what I can tell from Samba: we always have a last rebase and test run before things get pushed. that can't be prevented.
12:38:49 <atinm> as a maintainer I feel guilty when something is broken in my component but these are almost impossible to catch at the same time
12:38:51 <rastar> I would opt for rebase+fast-forward only
12:39:06 <atinm> rastar, you are here
12:39:17 <obnox> rastar: +1
12:39:18 <rastar> let us continue on the same mail thread, I wasn't here for a week
12:39:21 * obnox waves at rastar
12:39:22 <kkeithley_> firefox is showing me all emails from November.  What topic are we discussing?
12:39:22 <atinm> rastar, we had a classic example of the problem today what you explained in your mail
12:39:38 <atinm> kkeithley_, http://www.gluster.org/pipermail/gluster-devel/2015-November/047125.html
12:39:57 <rastar> obnox: Hi
12:39:59 <kkeithley_> that's better
12:40:01 <rastar> atinm: I just sae
12:40:03 <rastar> *saw
12:40:11 <atinm> rastar, thanks, shall I put AI on you to follow up on this?
12:40:16 <rastar> atinm: sure
12:40:25 * msvbhat joins meeting very late and apologises for it
12:40:49 <atinm> #action rastar to continue the discussion on rebase+fast forward as an option to gerrit submit type
12:41:25 <atinm> Anything else on the open floor?
12:41:36 <atinm> or we are done for the day?
12:42:02 <obnox> seems very hard to get regression tests completed. is this only now?
12:42:05 <obnox> is it getting worse?
12:42:31 <atinm> I'd say it is far far better than we had few months back
12:42:41 <obnox> ok.. :-)
12:43:00 <rastar> obnox: considering the mails we have got , I would still say we are at local maxima
12:43:01 <atinm> people used to wake up at midnight to check the regression vote and retrigger such that their patches can be taken in ;)
12:43:12 <rastar> more patches coming in last week and more failures
12:43:33 <kkeithley_> it's better, but it's not perfect.  Several of us in Westford are looking at the worst, i.e. most frequent, spurious failures. We're going to nag the maintainers to fix theirs.
12:43:47 <obnox> I am just submitting the odd patch here and there , but seems to be close to impossible to get a regression run through
12:43:57 <kkeithley_> And hagarth kicked jenkins again to get the triggered tests to run automatically.
12:43:57 <obnox> just talking last week or so
12:44:23 <obnox> right. hagarth asked me to report any unrelated/intermittent failures to the -dev list
12:44:26 <kkeithley_> yeah, I frequently have to retrigger tests to get them to pass
12:44:33 <obnox> which is why you see these mails
12:44:57 <obnox> ok. so let's say retriggering 5-10 times for a single patch is normal?
12:44:57 <rastar> yes, I propose that every maintainer sends a patch now which make run-tests.sh to run only those tests that belong to their component
12:45:03 <atinm> obnox, the more ideal thing would be to raise a bug itself
12:45:17 <obnox> atinm: ok, got it. will do so!
12:45:21 <kkeithley_> 5-10 seems unusually high, but maybe it's not so uncommon
12:45:53 <obnox> thanks, just wanted to know the bigger picture. will continue watching
12:46:25 <atinm> obnox, and if you happen to see a test failing multiple times do mark it as bad along with raising bug
12:46:40 <obnox> atinm: ok, in the run-tests.sh bad tests list?
12:46:46 <atinm> obnox, yes
12:46:49 <obnox> will do
12:47:18 <atinm> obnox, thanks
12:48:02 <atinm> so anything else?
12:48:52 <atinm> Thanks everyone for attending, we will see you next week at the same time, have a great day/evening ahead!
12:49:06 <obnox> cheers
12:49:26 <atinm> #endmeeting