12:01:29 <ndevos> #startmeeting Weekly Gluster Community Meeting
12:01:29 <zodbot> Meeting started Wed Sep 23 12:01:29 2015 UTC.  The chair is ndevos. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:29 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:38 <ndevos> #info Agenda: https://public.pad.fsfe.org/p/gluster-community-meetings
12:01:43 <ndevos> #topic Roll Call
12:01:50 * partner hiding
12:01:53 * kdhananjay is here
12:01:54 * itisravi is here
12:01:57 * raghu is here
12:02:02 * amye is here
12:02:24 * kkeithley is present
12:02:32 <ndevos> #chair partner kdhananjay itisravi raghu amye kkeithley hagarth
12:02:32 <zodbot> Current chairs: amye hagarth itisravi kdhananjay kkeithley ndevos partner raghu
12:02:42 <kdhananjay> :)
12:02:50 <amye> errybody's chair!
12:03:00 <hagarth1> yay!
12:03:13 <ndevos> #chair hagarth_
12:03:13 <zodbot> Current chairs: amye hagarth hagarth_ itisravi kdhananjay kkeithley ndevos partner raghu
12:03:21 <partner> \o/
12:03:25 <hagarth_> ok, let's get rolling
12:03:33 * overclk is here
12:03:36 <ndevos> some people need to feel more important than others, and want to be the chair 2x :)
12:03:40 <ndevos> #chair overclk
12:03:40 <zodbot> Current chairs: amye hagarth hagarth_ itisravi kdhananjay kkeithley ndevos overclk partner raghu
12:03:55 <itisravi> ndevos: that's chairitable of you ;)
12:04:22 <ndevos> #topic Action Items of last week
12:04:26 <ndevos> #topic kshlm to check back with misc on the new jenkins slaves
12:04:58 <ndevos> did you hear anything from kshlm, csim?
12:05:08 <hagarth> no, I have not
12:05:32 * tigert is here oops
12:05:48 <ndevos> #topic krishnan_p to update Gluster News about Gluster.next progress
12:05:58 * krishnan_p is here
12:06:06 <hagarth> krishnan_p: right on time to talk about your AI
12:06:12 <ndevos> krishnan_p: welcome! we're just at "krishnan_p to update Gluster News about Gluster.next progress"
12:06:46 <ndevos> #chair tigert krishnan_p
12:06:46 <zodbot> Current chairs: amye hagarth hagarth_ itisravi kdhananjay kkeithley krishnan_p ndevos overclk partner raghu tigert
12:07:06 <kkeithley> too many cooks
12:07:08 * hchiramm_ is here
12:07:13 <ndevos> #unchair kkeithley
12:07:13 <zodbot> Current chairs: amye hagarth hagarth_ itisravi kdhananjay krishnan_p ndevos overclk partner raghu tigert
12:07:22 <hagarth> kkeithley: +1
12:07:27 <tigert> ;)
12:07:45 <krishnan_p> I haven't gotten a chance to do that. Are there any ideas on how we should provide progress of ongoing efforts?
12:08:43 <ndevos> krishnan_p: I would say, regular status updates, which specifications/docs have been updated/improved
12:09:08 <krishnan_p> ndevos, OK.
12:09:27 <hagarth> I think we need a volunteer for sending out gluster news this week.. anybody up for it?
12:09:27 <atinmu> I propose Individual feature owners to update the etherpad with the progress and then at the end it can be collated as form of a blog, IMO it shouldn't be an individual's task
12:09:48 <krishnan_p> ndevos, I could provide that information on GlusterD 2.0 for gluster-weekly-news
12:09:51 <ndevos> krishnan_p: also, we need to make sure all the development is happening under github.com/gluster/some-repo so that everyone knows where to find things
12:10:01 <krishnan_p> ndevos, I will add that to the etherpad
12:10:14 <krishnan_p> ndevos, Is that a hard requirement?
12:10:30 <tigert> flightgear flight simulator project does that for their monthly newsletter in a wiki
12:10:36 <tigert> could work as a etherpad too
12:10:55 <krishnan_p> ndevos, how do we raise requests to gluster project in github to include our repos?
12:10:56 <ndevos> krishnan_p: "hard" maybe not, but for "official" Gluster Community work we would like to see that happening under the Gluster GitHub org
12:11:11 <tigert> they have a template they fill up
12:11:12 <atinmu> ndevos, the plan is to integrate Gluster.Next changes in the existing repo itself
12:11:39 <krishnan_p> ndevos, makes sense. So what is the procedure to add repos maintained elsewhere to be integrated into gluster project in github?
12:11:50 <ndevos> atinmu: integrate *what*? we can have a glusterd-2.0 repo or something while you guys are hashing things out
12:12:32 <ndevos> krishnan_p: the owner of the repo can pass ownership to one of the Gluster admins, and they can move it in the Gluster org
12:12:40 * anoopcs arrives late
12:12:41 <atinmu> ndevos, for eg - have glusterd2.0 as a separate folder in glusterfs codebase itself
12:12:51 <ndevos> krishnan_p: or, ask before creating a repo, and have it started at the right place :)
12:13:05 <krishnan_p> ndevos, OK. I will provide ownership to one of the gluster project admins in github
12:13:24 <hagarth> krishnan_p: move everything to gluster/glusterfs ?
12:13:26 <atinmu> ndevos,  That would ensure that we don't have contribution going on in multiple repos
12:13:59 <ndevos> atinmu: well, I understand development is currently happening in different repositories already?
12:14:16 <krishnan_p> hagarth, or under gluster retaining the repo name we currently have
12:14:23 <atinmu> ndevos, yes, currently its like that, but I am talking about the plan :)
12:14:28 <hagarth> ndevos: that needs to be stopped soon
12:14:41 <krishnan_p> hagarth, I am not sure if someone who is cloning gluster/glusterfs would like to pull all of it.
12:14:42 <ndevos> atinmu: and I'm talking about fixing things that are broken *now* :)
12:15:17 * ndevos likes github.com/gluster to be the single place for development
12:15:45 <krishnan_p> ndevos, but we don't want to use gerrit for sending patches for GlusterD 2.0
12:15:52 <ndevos> #action krishnan_p will add information about GlusterD-2.0 to the weekly news
12:15:53 <krishnan_p> ndevos, pull request model works like a charm
12:15:54 <hagarth> krishnan_p: that should not be a problem
12:16:22 <ndevos> krishnan_p: yeah, I would suggest to keep it in a separate repo for now, but under the Gluster org in GitHub
12:16:44 <hagarth> krishnan_p: we can evaluate github pull v/s gerrit, but I have a strong preference to host everything soon in gluster/glusterfs
12:16:56 <krishnan_p> hagarth, OK. I need some details before moving them into gluster/glusterfs. We could figure that on -devel ML. Does that make sense?
12:17:04 <krishnan_p> hagarth, Sure.
12:17:06 <hagarth> krishnan_p: sounds good
12:17:24 <ndevos> #action krishnan_p and atinmu will remind developers to not work in personal repositories, but request one for github.com/gluster
12:18:07 <atinmu> ndevos, sure
12:18:09 <ndevos> #action krishnan_p will send an email to the -devel list about merging the glusterd-2.0 work into the main glusterfs repo
12:18:21 <ndevos> #topic poornimag to send a mail on gluster-devel asking for volunteers to backport glfs_fini patches to release-3.5
12:18:29 <krishnan_p> ndevos, how does one remind developers about it? Do you want me to send an email to -devel?
12:18:34 <ndevos> did anyone see an email from Poornima about this?
12:18:44 <kdhananjay> ndevos: Nope.
12:18:55 <ndevos> krishnan_p: maybe you know who is working on Gluster.Next features, and tell them?
12:19:12 <ndevos> kdhananjay: ok, thanks, me neither :-.
12:19:13 <krishnan_p> ndevos, I'd prefer sending an email on -devel.
12:19:22 <ndevos> krishnan_p: sure, works as well
12:19:35 <ndevos> #topic kkeithley will send an email to the list to get opinions on how to close/move EOL'd bugs
12:19:47 <kkeithley> I did
12:20:11 <ndevos> was there a consensus about the proposal(s)?
12:20:13 <kkeithley> and we had a whole two responses; from atin and sankarshan
12:20:23 * ndevos counted 3
12:20:38 * hagarth counted 4
12:21:05 * sankarshan merely pointed out the specific text used by Fedora
12:21:12 <kkeithley> really? I only see from Atin and Sankarshan.  Who else.
12:21:15 <kkeithley> Who else?
12:21:34 <ndevos> oh, yes, 4, one on Monday
12:21:41 <kkeithley> Yes, sankarshan's was just as you said
12:21:45 <ndevos> Dave Warren and Pat Riehecky
12:22:11 <kkeithley> oh, they only replied on -users
12:22:14 <kkeithley> I see them now
12:23:44 <kkeithley> three votes for option B (close EOL)
12:24:04 <ndevos> #action kkeithley will reply to his previous email, confirming that End-Of-Life bugs will be closed
12:24:34 <ndevos> kkeithley: will you also close all the EOL bugs with a note?
12:25:22 <ndevos> kkeithley: you can easily automate that with as script based on https://github.com/gluster/release-tools/blob/master/close-bugs.sh
12:25:23 <kkeithley> ack
12:25:36 <ndevos> #action kkeithley will close all the EOL'd bugs with a note
12:25:38 <ndevos> thanks!
12:25:51 <ndevos> #topic jdarcy (and/or others) will post version of the NSR spec "pretty soon"
12:26:22 <atinmu> overclk might be aware of the status if jdarcy is not around
12:26:29 <ndevos> jdarcy is missing today, I guess "pretty soon" might be a week later
12:26:32 <overclk> ndevos, there are some things that need to be sorted out
12:26:46 <overclk> ndevos, and that will happen today, so count it next week.
12:26:47 <ndevos> oh, wait, skipped one :-/
12:26:49 <ndevos> #topic overclk will get the dht-scalability doc in glusterfs-specs update to the latest design
12:27:04 <ndevos> overclk: next week too?
12:27:14 <overclk> ndevos, that EINPROGRESS. need to churn out few things regarding dht2 to
12:27:28 <ndevos> overclk: ok, thanks :)
12:27:28 <overclk> ndevos, yeh, next week.
12:27:40 * krishnan_p needs to go now, apologies.
12:28:01 <ndevos> #topic GlusterFS 3.7
12:28:11 <ndevos> #link https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&id=glusterfs-3.7.4&hide_resolved=1
12:28:39 <ndevos> hmm, pranithk isnt here, he committed to do the next minor release
12:28:47 <hagarth> pranithk will join in now
12:28:58 <hagarth> there you go
12:29:08 <hagarth> pranithk: hmm, pranithk isnt here, he committed to do the next minor release
12:29:20 <pranithk> hagarth: ndevos: Yes I will do
12:29:22 <ndevos> pranithk: we're at the 3.7 topic now :)
12:29:24 <hagarth> pranithk: can you provide an update on 3.7.5?
12:29:38 <ndevos> pranithk: how are things looking? will the release make it on the 30th?
12:30:08 <pranithk> hagarth: I am still in the process of merging patches. I will send out a mail to get the regressions pass on their patches asap for people who sent patches on 3.7
12:30:17 <hagarth> pranithk: thanks!
12:30:18 <ndevos> 3.7.5 or 3.7.4?
12:30:24 <pranithk> ndevos: 3.7.5
12:30:36 <ndevos> the 3.7.4 tracker is still open...
12:30:40 <pranithk> ndevos: kaushal did 3.7.4
12:30:55 <hagarth> ndevos: yes, that needs to be closed
12:31:00 <hagarth> let us add an AI on kshlm
12:31:17 <ndevos> hagarth: yeah, please do :)
12:31:53 <hagarth> #action kshlm to clean up 3.7.4 tracker bug
12:32:07 <ndevos> #topic GlusterFS 3.6
12:32:28 <raghu> I have made 3.6.6 and the tar ball can be found here
12:32:29 <raghu> http://download.gluster.org/pub/gluster/glusterfs/3.6/LATEST/
12:32:50 <ndevos> thanks raghu, and all the packagers too!
12:32:52 <raghu> packages have to be built. Once the packages are ready, I will announce in the mailing list
12:33:22 <kkeithley> Fedora and EPEL rpms are on d.g.o now.  msvbhat is building Debian .debs.
12:33:42 <kkeithley> hgowtham is building Ubuntu .debs in Launchpad
12:33:45 <raghu> kkeithley: cool. Will announce once msvbhat is done with debian build
12:33:52 <hchiramm_> http://download.gluster.org/pub/gluster/glusterfs/3.6/LATEST/
12:33:55 <hchiramm_> raghu, ^^^
12:34:15 * ndevos does not see a difference in those URLs, hchiramm_
12:35:00 <ndevos> thanks raghu, anything more?
12:35:07 <raghu> I might be missing something, but I am not able to see the RPMs for fedora 22
12:35:10 <kkeithley> I think raghu meant http://bits.gluster.org/pub/gluster/glusterfs/src/glusterfs-3.6.5.tar.gz
12:35:22 <hchiramm_> raghu, F22 is in official build of fedora
12:35:34 <kkeithley> 3.6.x RPMs for F22 are in Fedora updates-testing
12:35:35 <hchiramm_> F22/ f22 rpms are
12:35:52 <raghu> kkeithley: cool
12:35:57 <kkeithley> get them with dnf (or yum)
12:36:30 <kkeithley> In a few days they'll move to Fedora updates
12:36:41 <kkeithley> No reason for them to be on d.g.o
12:36:41 <ndevos> #info if you are on Fedora 22, you can do 'dnf --enablerepo=udpates-testing update glusterfs'
12:37:03 <raghu> kkeithley: ok
12:37:05 <ndevos> well, if you can type udpates correctly
12:37:27 <kkeithley> s/--enablerepo=udpates-testing/--enablerepo=updates-testing/
12:37:45 <raghu> more backports are welcome for 3.6.7. 3.6.6 had only 5-6 patches. I hope 3.6.7 gets more backports
12:38:07 <ndevos> raghu: the fewer backports needed, the stabler the release ;-)
12:38:51 <ndevos> #topic GlusterFS 3.5
12:38:56 <ndevos> #link https://bugzilla.redhat.com/showdependencytree.cgi?maxdepth=2&hide_resolved=1&id=glusterfs-3.5.7
12:39:03 <raghu> ndevos: yeah. More backorts are needed than the number of backports that are being sent
12:39:44 <ndevos> raghu: hmm, yeah, 3.5 could use some fixes too, but nobody seems to care enough to send patches :-/
12:40:18 <hagarth> ndevos: how can we better that?
12:40:46 <ndevos> hagarth: I dont think we have anyone that is responsible for doing backports
12:41:10 <hagarth> ndevos: wouldn't it be a joint responsibility of component maintainers & release maintainers?
12:41:40 <kkeithley> It is nearing EOL after all. I personally wouldn't work too hard to make even more work for myself.
12:42:02 <ndevos> hagarth: possibly, but there is always $dayjob that is taking time too
12:42:24 <kkeithley> I was going to mention $dayjob too, but then decided not to. ;-)
12:42:24 <ndevos> kkeithley: for 3.5 yes, but 3.6 has the same problem
12:42:30 <hagarth> ndevos: right, I think we can be very reactive for 3.5 and 3.6
12:42:34 <kkeithley> But 3.6 isn't nearly EOL
12:42:52 <kkeithley> I was speaking specifically to 3.5.
12:42:59 <hagarth> if nobody reports a problem, should we fix it? unless of course the bug is a severe one.
12:43:15 <ndevos> sure, but we still need to convince develeopers and maintainers to do backports
12:43:21 <ndevos> *developers even
12:43:23 <kkeithley> Yes, we certainly should fix them.
12:43:26 <hagarth> ndevos: agree
12:43:28 <kkeithley> If they're real bugs
12:43:35 <raghu> ndevos: agree
12:44:21 <hagarth> I think maintainers can influence developers to contribute more backports (that's what I intended in the previous responsibility statement)
12:45:22 <hagarth> we need not have maintainers alone doing all the backports. maintainers can always request developers to backport when they notice a backport worthy fix that is getting into master.
12:45:32 <ndevos> hagarth: yes, but somehow the developers need to get the time to do backports for a non-productized version, and that seems difficult
12:45:57 <kkeithley> I don't know why it isn't automatic to do a backport when you fix a bug on master or release-3.7. I think it should be automatic.
12:46:12 <ndevos> kkeithley: yes, I think so too
12:46:35 <hagarth> ndevos: I think we should try calling out RFBs explicitly and over a period of time our behavioral pattern could improve.
12:46:42 <kkeithley> That should be factored into the time estimate for making the fix
12:46:44 <hagarth> RFB = request for backport ;)
12:47:28 <ndevos> hagarth: I'm thinking of sending a reminder to all maintainers, when they merge a change, they should check for a backport-bug, or create one
12:47:49 <kkeithley> although I can remember a time, not too long ago, when I felt like I was the only one doing it, and the backports would sit in gerrit forever.
12:47:53 <hagarth> ndevos: or atleast ask the concerned developer to clone one
12:48:17 <ndevos> hagarth: yes, how they get it done is up to them :)
12:48:40 <ndevos> #action ndevos send out a reminder to the maintainers about more actively enforcing backports of bugfixes
12:48:51 <ndevos> #topic GlusterFS 3.8
12:49:03 <ndevos> hagarth: you sent an email about 3.8, right?
12:49:23 <hagarth> ndevos: yes, I created an etherpad for tracking features aimed for inclusion in 3.8
12:49:35 <ndevos> #link http://www.gluster.org/pipermail/gluster-devel/2015-September/046791.html
12:49:41 <hagarth> #link https://public.pad.fsfe.org/p/gluster-3.8-features
12:49:56 <hagarth> if you are planning to add a new feature to 3.8, please list it in the etherpad
12:50:19 <hagarth> I will translate that to a project tracking page on gluster.org by 2nd week of october
12:50:37 <ndevos> hagarth: ok, thanks!
12:50:56 <ndevos> #topic Gluster 4.0
12:51:14 <ndevos> #link https://bugzilla.redhat.com/showdependencytree.cgi?id=glusterfs-4.0
12:51:33 <hagarth> ndevos: disable nfs by default?
12:51:52 <ndevos> hmm, yeah, that still seems to be the only "feature" for 4.0
12:52:19 <kkeithley> Don't we want/need that in 3.8?
12:52:33 <ndevos> others were planning to file bugs for their features, but that seems to be getting delayed
12:52:56 <ndevos> kkeithley: I dont think we can, it'll break a lot of expectations from users
12:53:28 <hagarth> ndevos: I am still debating if we should do that
12:54:35 <kkeithley> Well, if we don't do it, we should at least document it really well that it's in the pipeline.
12:55:02 <kkeithley> And start promoting nfs-ganesha as our preferred NFS solution
12:55:31 <ndevos> yeah, we need to have a more modular glusterd design for it, and glusterd should fir the bill
12:55:49 <ndevos> glusterd-2.0 that is
12:55:56 <hagarth> ndevos: +1, I am inclined to try modular daemon start/stop in 3.8 itself
12:56:30 <ndevos> hagarth: if that makes it in 3.8, splitting off Gluster/NFS would not be such a huge change
12:56:53 <hagarth> ndevos: yeah, let us discuss more about this offline
12:57:01 <ndevos> ... but I think thats all we have about Gluster 4.0 for now?
12:57:15 <atinmu> I have a quick update on 4.0, we are going to have a design discussion about GlusterD 2.0, DHT 2.0, NSR etc from 28th Sep to 29th Sep, I will send a hang out invite to the community asking for participation
12:57:37 * atinmu needs to jump into another meeting
12:57:42 <hagarth> ndevos: looks like, I expect that we will have more concrete updates in about 2 weeks from now.
12:58:03 <ndevos> atinmu: oh, interesting, please keep us informed! :)
12:58:11 <ndevos> #topic Open Floow
12:58:18 <ndevos> #topic Testing for various releases
12:58:20 <atinmu> ndevos, sure
12:58:38 <hagarth> I would like all maintainers to think about the nature of tests to be done for various releases
12:59:06 <hagarth> maybe we should discuss a list of tests that we absolutely want a release candidate to pass in next week's meeting
12:59:26 <hagarth> this is to make our pre-release testing more deterministic
13:00:02 <ndevos> #action hagarth will add a topic to the agenda for next weeks meeting about release testing
13:00:13 <ndevos> #topic posting http://blog.gluster.org/2015/08/welcome-to-the-new-gluster-community-lead/ to the mailing lists and general introduction for Amye
13:00:15 <hagarth> sounds good
13:00:36 <hagarth> amye: an AI for you this week?
13:00:47 <amye> hagarth: So ordered.
13:01:08 <amye> Should be easy, now that I have access to all of that ;)
13:01:26 <hagarth> #action amye will post http://blog.gluster.org/2015/08/welcome-to-the-new-gluster-community-lead/ to the mailing lists and provide a general intro
13:01:29 <hagarth> amye: cool, thanks!
13:01:42 <amye> hagarth: or maybe something even better. ;)
13:01:58 <ndevos> #topic Moderator for the next meeting(s)
13:02:01 <hagarth> amye: take over the community by surprise ;)
13:02:18 * hagarth spins the bottle for next week
13:02:36 * hagarth and it points at himself
13:02:44 <hagarth> so I will run the next meeting :)
13:02:51 * ndevos \o/
13:03:08 <ndevos> #info hagarth will host next weeks meeting
13:03:18 <ndevos> #info Weekly reminder to announce Gluster attendance of events: https://public.pad.fsfe.org/p/gluster-events
13:03:31 <ndevos> #info REMINDER to put (even minor) interesting topics on https://public.pad.fsfe.org/p/gluster-weekly-news
13:03:50 <ndevos> okay, anyone wants to mention something else?
13:04:35 <ndevos> I guess not
13:04:40 <ndevos> thanks all for joining!
13:04:45 <ndevos> #endmeeting