12:00:47 <hagarth> #startmeeting
12:00:47 <zodbot> Meeting started Wed Apr  1 12:00:47 2015 UTC.  The chair is hagarth. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:47 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:11 <hagarth> welcome to the All Fools' Day edition of the gluster community meeting
12:01:20 <overclk> :)
12:01:29 <hagarth> I assure you that we won't be fooling around much in the next 60 minutes :)
12:01:33 * spot resembles that remark
12:01:40 <saravanakumar> :)
12:01:41 <hagarth> #topic Roll Call
12:01:48 <hagarth> who do we have here today?
12:01:50 * kshlm o/
12:01:52 * jimjag is here
12:02:02 * overclk is here
12:02:06 * spot is here, no fooling
12:02:12 * hchiramm is here
12:02:39 <hagarth> ok, sounds good. Let's get rolling.
12:02:43 * raghu is here
12:02:43 * atinmu is here
12:02:45 <hagarth> #topic AI review from last week
12:02:59 <hagarth> AI - tigert to continue investigating the blog post duplication
12:03:00 * ndevos is present too _o/
12:03:39 <hagarth> I suspect we haven't had much action on this one. tigert if you are around later, can you please post an update to us?
12:03:55 <hagarth> AI - hchiramm will share the outcome of the non-mailinglist packagng discussions on the mailinglist (including the Board)
12:04:01 <hagarth> hchiramm: TBD right?
12:04:01 <hchiramm> hagarth, not yet..
12:04:04 <hchiramm> yep
12:04:04 <hagarth> ok
12:04:34 <hagarth> AI - hagarth and pranithk will complete the feature page for (k)vm hyperconvergence
12:04:57 <hagarth> still TBD, I am busy with 3.7 and pranithk is busy with erasure coding. We'll get there soon.
12:05:12 <hagarth> AI - spot to reach out to community about website messaging
12:05:20 <hagarth> spot: any updates here?
12:05:29 <spot> we should change this to "spot to work on new slogan ideas"
12:05:53 <hagarth> spot: consider that done :)
12:05:58 <spot> I have a meeting with the people in Red Hat's brand team to brainstorm new slogan ideas
12:06:06 <spot> unfortunately, that meeting is scheduled for April 30.
12:06:13 <spot> (earliest I could get with them)
12:06:28 <spot> so this item will be cold for a while
12:06:43 <hagarth> spot: OK, let us pick this up after then.
12:07:13 * spot will send a "got a cool slogan idea" email in the interim
12:07:26 <hagarth> spot: that would be great
12:07:30 <hagarth> next AI - hagarth to carry forward discussion on automated builds for various platforms in gluster-infra ML
12:07:46 <hagarth> still TBD, I will get this done over the week.
12:08:07 <hagarth> next AI - telmich will send an email to the gluster-users list about Gluster support in QEMU on Debian/Ubuntu
12:08:26 <hagarth> haven't heard from telmich, maybe we should drop this off.
12:09:06 <hagarth> next AI - hagarth will post more details about an overlay xlator (container targetted) when he has a bit more clarity (and post his hacks)
12:09:22 <hagarth> not found time for this, will keep it parked for next week.
12:09:34 <hagarth> next AI - firemanxbr to submit a patch for 3.6.3, change IP addr.
12:09:42 <hagarth> firemanxbr: any update on this one?
12:10:06 <firemanxbr> hagarth, sorry, no update :(
12:10:31 <ndevos> firemanxbr: thats a script you wanted to add, right?
12:10:43 <hagarth> firemanxbr: no problem, would you find some time in the coming week or do we need more time for that?
12:11:02 <ndevos> firemanxbr: if that is the case, we can probably wait and get it in 3.6.4
12:11:22 <firemanxbr> ndevos, exactly, I think one python solve this ti
12:11:23 <hagarth> ndevos: +1, raghu is almost done with 3.6.3.
12:11:27 * firemanxbr it
12:11:43 <hagarth> ok, let us track this for 3.6.4.
12:11:46 <hagarth> next AI -  jimjag will continue engaging with the board
12:11:58 <hagarth> jimjag: any updates for us here?
12:12:09 <jimjag> Contacted each board member indivdually to try to
12:12:16 <firemanxbr> ndevos, I believe in 3.6.4 I can solve
12:12:24 <jimjag> arrange a f2f in April. response has been lackluster.
12:12:44 <jimjag> Also asked how many would be attending the summit in May, and have not heard back yet.
12:12:56 <hagarth> jimjag: ok, lackluster without gluster :)
12:13:08 <jimjag> hagarth: :)
12:13:18 <hagarth> jimjag: what are our next steps here? should we retain this AI or drop it from this meeting?
12:13:37 <jimjag> I'd like to continue reporting this issue
12:13:48 <spot> jimjag: i have a few board members confirmed as wanting to attend the summit, will send you names via email
12:13:52 <hagarth> jimjag: ok, noted.
12:14:04 <jimjag> I think it's important for the community to know what's happening from the board perspective
12:14:10 <hagarth> jimjag: +1
12:14:16 <jimjag> end-of-report
12:14:26 <hagarth> jimjag: thanks, moving on to next AI
12:14:29 <firemanxbr> ndevos, I'm making some tests with my gluster in CentOS 7.1, I believe my patch to be for this environment
12:14:29 <hagarth> AI - ndevos to poll the community about continued el5 support
12:15:00 <hagarth> ndevos: do we have any results/preferences from this exercise?
12:15:00 <ndevos> oh, I did send an email, but not a lot of responses yet
12:15:20 <kkeithley> there was one today
12:15:50 <ndevos> 2 responses in total, I think - neither made me very eager to provide support for old distributions
12:16:27 <hagarth> ndevos: maybe wait and watch for a little more?
12:16:50 <ndevos> from today: http://www.gluster.org/pipermail/gluster-devel/2015-April/044287.html
12:17:05 <kkeithley> for Ubuntu 12.4 LTS we only need to disable RDMA.  Other than old python in RHEL/CentOS 5, what other "old" distros?
12:17:07 <ndevos> hagarth: yes, wait a little longer and see if there are more responses
12:17:20 <hagarth> ndevos: ok
12:17:41 <atinmu> Folks, I am looking for response on the security hardening flags, I've got couple of responses from ndevos & kshlm, but will appreciate if more people can chip in with their thoughts
12:17:42 <hagarth> kkeithley: do we have userspace-rcu in CentOS 5 and older Ubuntu?
12:17:58 <ndevos> and from last week: http://www.gluster.org/pipermail/gluster-devel/2015-March/044205.html about combining community gluster with rhs...
12:18:01 <kkeithley> good point, good question
12:18:25 <kkeithley> userspace-rcu that is
12:18:35 <hagarth> without userspace-rcu, 3.7 cannot work.
12:18:56 <ndevos> indeed, that is one of the drivers for the whole questioning
12:19:20 <ndevos> userspace-rcu is (currently) only needed server-side, the client bits can do without it
12:19:31 <hagarth> ndevos: right
12:20:05 <hagarth> availability of userspace-rcu will definitely have a bearing on our server side supportability stance.
12:20:40 <ndevos> well, we can get userspace-rcu in EPEL-5 if we want to..
12:21:03 <hagarth> ndevos: good point, let us discuss more next week by when we will hopefully have more responses.
12:21:08 <hagarth> next AI - jdarcy to schedule a 4.0 catch up meeting next week
12:21:09 <kshlm> userspace-rcu should be available in the testing repos for epel5 currently.
12:21:25 <jdarcy> Not done.  Then again, nobody's been working on 4.0 this week.
12:21:38 <hagarth> kshlm: have we done any basic sanity checks with userspace-rcu on centos5?
12:21:49 <hagarth> jdarcy: ok, defer by a week or two?
12:21:56 <jdarcy> Will try again next week.
12:22:13 <kshlm> hagarth, O
12:22:29 <hagarth> jdarcy: ok, thanks.
12:22:30 <kshlm> I'd built my own userspace-rcu package and had tested it out.
12:22:35 <kshlm> And it had worked.
12:22:58 <hagarth> kshlm: ok, maybe we should just do a sanity check with the package provided in epel-testing
12:23:16 <kkeithley> Ubuntu Precise (12.4 LTS) has liburcu(-dev) 0.6.7.  Is that new enough?
12:23:23 <kshlm> The package was only enabled after I'd given a +1.
12:23:29 <hagarth> kshlm: ok
12:23:29 <kshlm> I'll check it again anyway.
12:23:41 <kshlm> kkeithley, Nope. Need atleast 0.7
12:24:05 <kkeithley> userspace-rcu-0.7.7 is already built for epel-5, not sure if it's been pushed to stable yet
12:24:16 <hagarth> ok, seems like a good discussion for next week. kshlm - your results will also be helpful for that.
12:24:47 <hagarth> next AI - hagarth and JustinClift to send a list of spurious failures to gluster-devel.
12:25:01 <ndevos> kkeithley: still in testing - https://admin.fedoraproject.org/updates/FEDORA-EPEL-2015-1134/userspace-rcu-0.7.7-1.el5
12:25:25 <hagarth> the list of spurious regression failures has been sent out, we need a little more attention on these failures.
12:25:56 <hagarth> having said that, the number of spurious failures seems to have come down by a bit as compared to last week.
12:26:09 <hagarth> #link - https://public.pad.fsfe.org/p/gluster-spurious-failures
12:27:20 <hagarth> I plan to send a note to the original developers of these tests and then determine whether we should carry these tests or drop them from the repo till they are fixed.
12:27:38 <hagarth> next AI - kshlm to propose gluster winter of code ideas on blog/mailing lists.
12:27:45 <hagarth> kshlm: any progress here? :)
12:28:24 <hagarth> guess not, kshlm - please provide an update when you get a chance.
12:28:27 <hagarth> moving on
12:28:30 <hagarth> #topic Gluster 3.6
12:28:51 <hagarth> raghu: do you want to provide an update here?
12:29:43 <raghu> I have made 3.6.3beta2 and sent out a mail to both users and devel
12:30:01 <hagarth> raghu: any ETA for GA of 3.6.3?
12:30:08 <ndevos> raghu: got the release notes ready for review too?
12:30:12 <hchiramm> the packages are available at download.g.org http://download.gluster.org/pub/gluster/glusterfs/qa-releases/3.6.3beta2/
12:30:39 <hagarth> hchiramm: great, thanks!
12:30:48 <hchiramm> yw!
12:31:08 <raghu> ndevos: release notes I have just added the list of bugs fixed as part of the mail. I have not added it in docs/release_notes.md. I will add it when I do the GA
12:31:29 <hchiramm> do we need a release_notes.md for minor releases ?
12:31:41 <ndevos> hchiramm: I tend to do that, yes
12:31:55 <kkeithley> as do I for 3.4.x
12:32:02 <raghu> hagarth: I want to make it early next week. There are still 3 patches for release-3.6 branch which need to be reviewed. I dont know whether they must be included for 3.6.3 or is it ok to defer it to 3.6.4
12:32:35 <kkeithley> we would have to do another beta if you're going to add more patches.  I'd push them off to 3.6.4
12:32:53 <raghu> kkeithley: Agree. Makes sense.
12:32:54 <hagarth> raghu: it is your call, unless very critical I think we can move them to 3.6.4
12:32:54 <hagarth> #info 3.6.3 GA to happen early next week
12:32:58 <kshlm> ON the topic of release notes, can they be made more ...public?
12:33:01 <ndevos> raghu: push them to 3.6.4, and add them to a glusters-3.6.4 tracker in bugzilla?
12:33:04 <kshlm> I didn't know we had release notes in docs/
12:33:14 <hchiramm> kshlm, there should be a blog on the same anyway
12:33:21 <raghu> yeah. Will defer them to 3.6.4 then.
12:33:44 <ndevos> yes, release notes normally get added to the repo on release, included in the release email and a blog post
12:33:53 <ndevos> practically all just copy/paste
12:34:07 <kshlm> Blogs/emails are good,
12:34:11 <ndevos> kshlm: how could we make it more public?
12:34:26 <kkeithley> I've been putting the release notes for 3.4.x on blog.gluster.org and they're in the source tree. I'm not sure they can be any more public than that.
12:34:31 <kshlm> but the when looking at just the source, they don't stand out.
12:34:45 <kshlm> I never thought to look into the docs directory to find them.
12:35:06 <kkeithley> and blogs get tweeted
12:35:07 <ndevos> kshlm: would a mentioning of release notes help in the main README file?
12:35:15 <kshlm> Other projects generally have a CHANGELOG file in source root detailing the same.
12:35:38 <kshlm> ndevos, that would help.
12:36:07 <hagarth> kshlm: care to send a patch?
12:36:19 <kshlm> Let's discuss this further later, and continue with the meeting for now.
12:36:21 <ndevos> I do not think release notes should be mixed with changelogs too much, and we do have a CHANGELOG in the tarball, its generated from the 'git log'
12:36:35 <kkeithley> Our ChangeLog in the source root says "ChangeLog is maintained by "git log"."  ;-)
12:36:41 <hchiramm> :)
12:36:50 <hagarth> yeah, let us discuss this more over email.
12:37:00 <hagarth> moving on
12:37:05 <hagarth> #topic Gluster 3.5
12:37:10 <ndevos> \o/
12:37:32 <ndevos> I've been looking into the current status, and was planning to do a beta today
12:37:40 <ndevos> that would be 3.5.4beta1
12:37:49 <hagarth> cool
12:38:02 <hchiramm> :(
12:38:14 <ndevos> sorry, hchiramm
12:38:15 <hagarth> hchiramm: what disappoints you?
12:38:29 <hchiramm> hagarth, have to build packages
12:38:48 <hchiramm> we already did 2 betas
12:38:49 <ndevos> there are currently 23 bugs fixed in the release-3.5 branch after the last release, so it should be an improvement
12:38:52 <hchiramm> and going with one more
12:38:58 <hagarth> hchiramm: are we reaching the phase of automate or perish?
12:39:12 <kkeithley> since Lala is gone (not really, but....)  we should find someone else to help with packaging
12:39:18 <hchiramm> indeed.
12:39:33 * ndevos notes that it just started to hail outside
12:39:54 <jdarcy> All hail ndevos.
12:39:58 <hagarth> hchiramm: more reasons for me to send that build automation email :)
12:40:10 <hchiramm> and one more coming ( 3.6.3  ) in next week
12:40:27 <hagarth> hchiramm: I plan to do an alpha for 3.7 also soon ;)
12:40:37 * hchiramm hiding
12:40:46 <hagarth> anything more on 3.5?
12:40:53 <ndevos> no, not from me
12:40:57 <kshlm> I can help with the packaging, if help is needed.
12:41:12 <hchiramm> as kkeithley said, we need some more hands on packaging
12:41:16 <hchiramm> kshlm, thanks !!
12:41:18 <hagarth> kshlm: great, let us look to automate the entire workflow.
12:41:25 <hagarth> moving on
12:41:30 <hagarth> #topic Gluster 3.4
12:41:47 <ndevos> hchiramm: maybe also send a request for help to the mailinglists? some community members might be interested too
12:42:11 <atinmu> I would like to learn that too from kshlm :)
12:42:15 <hagarth> kkeithley: any updates here?
12:42:17 <hchiramm> ndevos, yeah, lets discuss it..
12:42:28 <kkeithley> beta4 is out, GA on Monday
12:42:35 <hagarth> kkeithley: fantastic
12:42:42 <ndevos> packages!
12:42:48 <hagarth> #info 3.4.7 GA to be released early next week
12:42:49 <hchiramm> !! \o
12:42:54 <kkeithley> More packages for hchiramm (and me)
12:43:03 <hagarth> next week seems to be a busy one from a release perspective :)
12:43:06 <kkeithley> any testing feedback?
12:43:27 <hagarth> I haven't heard any so far
12:43:29 <hchiramm> kkeithley, Isnt it Easter ?  :)
12:43:48 <kkeithley> Monday?
12:44:03 <hagarth> anything more on 3.4?
12:44:13 <kkeithley> no, not from me
12:44:19 <hagarth> ok, moving on.
12:44:20 <hagarth> #topic Gluster Next releases
12:44:23 <kkeithley> 3.4.7 will probably be the last 3.4.x release
12:44:25 <hagarth> 3.7
12:44:30 <hagarth> kkeithley: right
12:44:57 <hagarth> on the topic of 3.7, we seem to be doing ok. I had a few nightlies pass my regular sanity tests.
12:45:16 <hagarth> this is the default configuration without enabling any of our shiny new features in 3.7 :)
12:45:36 <hagarth> I plan to run more tests and gather some performance numbers over this week.
12:46:04 <hagarth> I intend branching 3.7 after we merge more coverity, logging fixes and critical bugs.
12:46:36 <hagarth> If you have any critical bugs to be fixed in 3.7, please add them to the tracker.
12:46:43 <jdarcy> And spurious regression-test failures?
12:46:51 <hagarth> jdarcy: thanks for reminding, that too.
12:47:21 <hagarth> I plan to do an alpha release next week and plan an open test week for new features the week after.
12:48:06 <hagarth> I intend sending out 3.7 status emails on a fairly regular basis on gluster-devel to help co-ordinate our efforts.
12:48:22 <hagarth> that's all I have from 3.7. any questions?
12:48:50 <hagarth> figure not, jdarcy - would you like to provide any updates on 4.0?
12:48:52 <ndevos> is https://bugzilla.redhat.com/showdependencytree.cgi?id=glusterfs-3.7.0 still up to date?
12:49:03 <hagarth> ndevos: yes, that is what I am using for tracking.
12:49:21 <jdarcy> Nothing there.  Been spending all my time on regression failures, and others have been busy elsewhere too.
12:49:38 <hagarth> jdarcy: ok
12:49:40 <hagarth> moving on
12:49:57 <hagarth> #topic docs
12:50:05 <hagarth> hchiramm: any updates here?
12:50:19 <hchiramm> hagarth, its progressing
12:50:29 <hchiramm> the admin guide in source is getting updated to the latest
12:50:46 <hagarth> hchiramm: fantastic, so more consumable docs by 3.7.0?
12:50:58 <hchiramm> expecting.. :)
12:51:06 <hagarth> hchiramm: that would be great! :)
12:51:08 <hchiramm> and lala sent a mail to gluster-devel
12:51:15 <hchiramm> requesting docs for all new features
12:51:26 <hchiramm> we have to make sure its in before code freeze
12:51:47 <hagarth> hchiramm: sure, I will capture that in my summary email for 3.7 status.
12:51:53 <hchiramm> thanks !
12:52:01 <hagarth> moving on to next topic
12:52:04 <hagarth> #topic Awesum Web Presence
12:52:23 <hagarth> debloper is working on a refresh of the website
12:52:39 <hchiramm> http://code.debs.io/glusterweb/
12:52:45 <hagarth> an early mock up is here - http://code.debs.io/glusterweb/
12:52:54 <hagarth> hchiramm: thanks :)
12:53:11 <hchiramm> in draft stage , however welcome community feedback on the same.. :)
12:53:39 <hagarth> it would be *really* nice to have a functional website by the time we release 3.7.0
12:53:49 <hchiramm> yep :)
12:54:05 <hagarth> anything more on this topic?
12:54:20 <hagarth> guess not, moving on
12:54:24 <hagarth> #topic Gluster Summit Barcelona, second week in May
12:54:31 <hagarth> spot: any updates on this one?
12:54:42 <spot> not yet, sorry, still working with EMEA marketing.
12:55:04 <ndevos> so, this has not been confirmed yet?
12:55:47 <spot> it is happening, i promise, but we haven't locked in venue yet.
12:56:09 <hagarth> spot: ok, maybe we can discuss more about agenda etc. next week?
12:56:12 <ndevos> sure, but the date and location are still open, right?
12:56:29 <spot> location is Barcelona unless we cannot find an affordable venue.
12:56:31 * ndevos just wonders if he needs to put the dates in his calendar
12:56:41 <spot> dates are likely May 13-14
12:56:49 <ndevos> okay, thanks!
12:56:53 <hagarth> spot: thanks
12:57:06 <hagarth> moving on to next topic
12:57:16 <hagarth> #topic Gluster Winter of Code
12:57:29 <hagarth> kshlm: ?
12:57:51 <kshlm> I had a conversation with hagarth on this
12:57:58 <kkeithley> Richard III says now is the Winter of our disCode Extents
12:58:15 <hagarth> lol
12:58:25 <kshlm> and we decided to have at least one more conversation to iron out the details.
12:58:38 <kshlm> That one more conversation is yet to happen.
12:59:07 <hagarth> kshlm: ok, let us post some details by next week.
12:59:17 <kshlm> Yes.
12:59:24 <hagarth> ok, moving on.
12:59:26 <hagarth> #topic Static Analysis results
12:59:45 <hagarth> how many of us are aware of static analysis results on download.gluster.org?
12:59:56 * ndevos _o/
13:00:16 <hagarth> #link - http://download.gluster.org/pub/gluster/glusterfs/static-analysis/
13:00:17 <hchiramm> \o
13:00:34 <hagarth> I wonder how we can make these results more presentable?
13:00:57 <kkeithley> yeah, it could stand some improvement
13:01:13 <hagarth> these are nightly results and are more relevant than what scan provides
13:01:24 <kshlm> AFAIK clang-analyzer produces html output, so those could be hosted directly.
13:01:45 <hagarth> kshlm: right, these are also html reports
13:02:19 <kshlm> But they were huge the last time I tried. Somewhere in the hundreds of MBs.
13:02:27 <hagarth> let us figure out how we can make it better
13:02:32 <kkeithley> more relevant that what scan provides?  scan as in coverity scan?
13:02:32 <ndevos> 116MB on http://download.gluster.org/pub/gluster/glusterfs/static-analysis/master/glusterfs-clang-analyze/2015-03-31-f5e4c943/html/
13:02:37 <ndevos> as a tarball...
13:02:51 <hagarth> so that we fix issues as they are found
13:03:00 <hagarth> kkeithley: yes, since we can schedule coverity scan upstream runs only twice per fortnight.
13:03:13 <hagarth> these results would be more relevant on an ongoing basis.
13:03:41 <hagarth> I think I'll start a mailing list discussion on this to get more thoughts.
13:03:43 <kkeithley> yes, definitely.
13:03:44 <ndevos> we need something that compares the latest result with the previous run, and send out an email if something was added
13:03:58 <kshlm> I remember hearing about some gerrit/jenkins plugin that could do static analysis.
13:04:02 <hagarth> ndevos: yeah, something like that would be great.
13:04:05 <hagarth> kshlm: sonar
13:04:12 <kshlm> Yeah, that.
13:04:13 <kkeithley> there are nightly coverity scans there too, but it's not integrated with workbench so no false positive elimination
13:04:21 <hagarth> kkeithley: right
13:04:31 <kshlm> Why haven't we moved onto that?
13:04:34 <hagarth> any more last topics for discussion today?
13:04:41 <hagarth> kshlm: it needs a gerrit upgrade :-/
13:04:49 <kkeithley> I think we're waiting for jenkins and gerrit to be upgraded
13:05:07 <kshlm> Aren't we getting one soon?
13:05:17 <hagarth> kshlm: that's the plan. let us see.
13:05:18 <ndevos> yeah, a jenkins update and then install a plugin like https://wiki.jenkins-ci.org/display/JENKINS/Clang+Scan-Build+Plugin
13:05:18 <kkeithley> JustinClift is working on it
13:05:32 <hagarth> guess we don't have any more topics for today. I need to run to a different appointment.
13:05:38 <overclk> just an FYI: There's a bitrot detection hangout scheduled for tomorrow.
13:05:46 <hagarth> thanks everyone for attending today, tty all next week!
13:05:54 <hagarth> overclk: look forward to that!
13:05:59 <kshlm> overclk, It'll be recorded right?
13:06:02 <overclk> hagarth, thanks!
13:06:08 <hagarth> #endmeeting