12:01:06 <hagarth> #startmeeting
12:01:07 <zodbot> Meeting started Wed Oct 14 12:01:06 2015 UTC.  The chair is hagarth. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:07 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:07 * JustinClift1 waves
12:01:12 <hagarth> agenda: https://public.pad.fsfe.org/p/gluster-community-meetings
12:01:22 <hagarth> #topic roll call
12:01:27 <lpabon> o/
12:01:27 <hagarth> who do we have here today?
12:01:34 * skoduri is here
12:01:37 * kkeithley is here
12:01:39 * xavih is here
12:02:00 * raghu is here
12:02:10 * anoopcs is here
12:02:16 <hagarth> ok, let us get started
12:02:19 <hagarth> #topic AIs from last week
12:02:27 <hagarth> kshlm to check back with misc on the new jenkins slaves
12:02:45 <hagarth> I don't see either of them here, carrying this AI further
12:02:49 <hagarth> #action kshlm to check back with misc on the new jenkins slaves
12:03:02 <hagarth> AI - krishnan_p and atinmu will remind developers to not work in personal repositories, but request one for github.com/gluster
12:03:15 <hagarth> atinmu: any update on this?
12:03:37 * poornimag arrives late
12:03:43 * jiffin is here
12:03:43 * rastar is here
12:03:50 <hagarth> moving on as I don't see atinmu yet
12:03:54 <hagarth> #action krishnan_p and atinmu will remind developers to not work in personal repositories, but request one for github.com/gluster
12:04:07 <hagarth> next AI - krishnan_p will send an email to the -devel list about merging the glusterd-2.0 work into the main glusterfs repo
12:04:26 <hagarth> this was partially done, we decided to keep it as a separate repo till glusterd-2.0 is more mature
12:04:55 <hagarth> all new xlators will get into xlators/expiremental of the main repo
12:05:03 <hagarth> next AI - kkeithley will close all the EOL'd bugs with a note
12:05:07 <kkeithley> done
12:05:16 <hagarth> cool
12:05:23 <hagarth> next AI - overclk will send another patch to enhance latest dht-scalability design doc in glusterfs-specs
12:05:44 <hagarth> I don't see overclk, did anybody see this in glusterfs-specs?
12:06:00 * msvbhat arrives late to the meeting
12:06:05 <hagarth> figure not
12:06:11 <hagarth> carrying the AI for next week
12:06:15 <hagarth> #action overclk will send another patch to enhance latest dht-scalability design doc in glusterfs-specs
12:06:24 <hagarth> next AI - jdarcy and avra  will post version of the NSR spec "pretty soon"
12:06:35 <hagarth> this is now done, thanks jdarcy and asengupt!
12:06:50 <hagarth> next AI - kshlm to clean up 3.7.4 tracker bug
12:06:59 <hagarth> I am not sure if this happened. anybody aware of this?
12:07:21 * ira is here... kinda
12:07:28 <hagarth> maybe we will take this up in the 3.7.x topic
12:07:35 <hagarth> next AI - ndevos send out a reminder to the maintainers about more actively enforcing backports of bugfixes
12:07:43 <hagarth> don't remember seeing this
12:07:48 <hagarth> #action ndevos send out a reminder to the maintainers about more actively enforcing backports of bugfixes
12:07:55 <hagarth> next AI - poornimag and skoduri will write a trip report for SDC
12:08:03 <hagarth> poornimag, skoduri: has this been done?
12:08:03 <skoduri> I think obnox already sent it
12:08:08 <skoduri> to storage-eng
12:08:24 <kkeithley> IIRC I saw it
12:08:30 <hagarth> skoduri: I think posting it on gluster-devel would be more useful
12:08:44 <hagarth> so let us merge both your AIs to one :)
12:08:50 <skoduri> hagarth, sure..will check with obnox_ and get it done :)
12:08:53 <poornimag> yeah, obnox set it on storage-eng, and that covers a great deal of info,
12:09:02 <poornimag> yes ok
12:09:13 <hagarth> #action skoduri, poornimag and obnox_ to post SDC trip report on gluster-devel
12:09:17 <kkeithley> community gluster doesn't know what storage-eng is. ;-)
12:09:25 <skoduri> :)
12:09:34 <hagarth> I forget about storage-eng when I wear my community hat ;)
12:09:43 <hagarth> next AI - pranithk will send out a mail to the community on 3.7.5 release and updates
12:09:51 <hagarth> this has been done, thanks pranithk!
12:10:03 <hagarth> next AI - raghu to call for volunteers and help from maintainers for doing backports listed by rwareing to 3.6.7
12:10:09 <hagarth> raghu: any updates on this?
12:10:43 <raghu> hagarth: I spoke pranith about it and as per my discussion with him, I am going to identify the backports in separate components and ask for volunteers
12:10:56 <hagarth> raghu: ok thanks, will carry forward this AI
12:11:00 <hagarth> #action raghu to call for volunteers and help from maintainers for doing backports listed by rwareing to 3.6.7
12:11:12 <hagarth> next AI - All the leads to update community on the outcomes of the design discussions.
12:11:29 <hagarth> I saw atinmu's report on this one. can we consider this as done?
12:12:09 <hagarth> I take that as a yes :).
12:12:31 <hagarth> moving on to topic.next
12:12:38 <hagarth> #topic Gluster 3.7
12:12:56 <hagarth> 3.7.5 has been released with all the usual packages except Debian
12:13:04 <hagarth> kkeithley: is that right?
12:13:19 <kkeithley> Debian dpkgs are there
12:13:37 <kkeithley> along with all the other usual ones
12:13:53 <hagarth> kkeithley: cool, thanks!
12:14:09 <hagarth> we need a release maintainer for 3.7.6.. any volunteers?
12:14:34 <rastar> I volunteer
12:14:38 <hagarth> rastar: thanks!
12:14:46 <hagarth> #info rastar will be the release maintainer for 3.7.6
12:14:49 <kkeithley> Debian dpkgs for jessie and stretch
12:15:25 <hagarth> 3.7.6 is expected to land towards the end of this month as usual
12:15:40 <hagarth> any other questions, thoughts on 3.7.x?
12:16:01 <hagarth> raghu: do you happen to know if the 3.7.4 tracker was cleaned up?
12:16:14 <raghu> hagarth: dont know
12:16:22 <hagarth> ok
12:16:38 <hagarth> #action kshlm to clean up 3.7.4 tracker bug
12:16:40 <hagarth> moving on
12:16:50 <hagarth> #topic Gluster 3.6
12:17:00 <hagarth> raghu: do you want to provide an update?
12:17:19 <raghu> hagarth: not much from me. I wm waiting for backports
12:17:30 <raghu> as of now there are no patches that can be merged
12:18:13 <hagarth> raghu: let us target those backports for the next release
12:18:31 <raghu> hagarth: sure
12:18:39 <hagarth> let us work with the respective maintainers to have those backports land in 3.6
12:18:54 <hagarth> raghu: are you still shooting for 20th to push this release out?
12:19:19 <raghu> hagarth: yeah. If enough patches goes in I can release on 20th
12:19:30 <raghu> otherwise I will have to consider some other date.
12:19:34 <hagarth> raghu: ok
12:19:56 <hagarth> any more questions on 3.6?
12:20:00 <raghu> hagarth: nope
12:20:17 <hagarth> ok, moving on
12:20:22 <hagarth> #topic Gluster 3.5
12:20:36 <hagarth> since ndevos is not here today, let us skip this topic.
12:20:37 <hagarth> moving on
12:20:45 <hagarth> #topic Gluster 3.8
12:21:03 <hagarth> please add your planned inclusions to 3.8 in this etherpad
12:21:11 <hagarth> https://public.pad.fsfe.org/p/gluster-3.8-features
12:21:28 <hagarth> i will post a tracking page on gluster.org for 3.8 by next week's meeting
12:21:44 <hagarth> #action hagarth to post a tracking page on gluster.org for 3.8 by next week's meeting
12:21:59 <hagarth> any questions on 3.8?
12:21:59 <kkeithley> I'm concerned about when the converged HA for SMB+Ganesha is going to land!
12:22:18 <hagarth> kkeithley: anytime in the next 3 months? :)
12:22:34 <kkeithley> Hopefully at the beginning of that three months, not at the end
12:22:40 <kkeithley> nearer the beginning
12:22:54 <hagarth> kkeithley: yes, that would be helpful
12:22:57 <kkeithley> Are we still three months from GA?
12:23:12 <kkeithley> I was anticipating early December.
12:23:55 <hagarth> kkeithley: I am thinking of sometime around then. Have shelved the December plan given that we are working on multiple things.
12:24:10 <kkeithley> okay, that gives us a bit of breathing room
12:24:10 <hagarth> will have a rough schedule in the project page for 3.8 by next week
12:24:38 <hagarth> anything else on 3.8?
12:25:31 <hagarth> figure not, let us move on
12:25:35 <hagarth> #topic GlusterFS 4.0
12:25:52 <hagarth> jdarcy: anything to note this week on 4.0?
12:26:32 <jdarcy> Not much.  Still have one more spec (for the journal part) to post.  Working on the etcd support in glusterd right ow.
12:27:00 <jdarcy> Other people are still chugging away on DHT2, GlusterD2, etc.
12:27:22 <hagarth> ok and as noted earlier the NSR design document is available for review. Please post your comments on the ML thread.
12:28:02 <hagarth> any other comments, questions on 4.0?
12:28:05 <atinmu> as far as GlusterD 2.0, we are working on ReST APIs for peer commands and also working on the volume create
12:28:35 <atinmu> I'd like to see other features coming up with their thoughts on their integration with GD 2.0
12:28:51 <atinmu> I had sent a mail asking for participation for it
12:28:56 <hagarth> atinmu: cool, it would be nice to have the APIs posted on both -devel and -users for review.
12:29:13 <lpabon> atinmu: will you have the apis on a wiki or a document?
12:29:14 <atinmu> as a start people can start defining the APIs for their respective features
12:29:30 <atinmu> lpabon, I think we will have it in a document
12:29:54 <lpabon> atinmu: ok, thanks
12:29:57 <hagarth> atinmu: I think we should create api.gluster.org and host the definitions there :)
12:30:42 <hagarth> thanks for the updates, atinmu and jdarcy!
12:30:50 <hagarth> moving on
12:30:55 <hagarth> #topic Open Floor
12:30:56 <lpabon> you may need to edit many times to get it right.  FYI, document for heketi has over 100 revisions
12:31:52 <hagarth> lpabon: agree. the more eyes that we get to review the APIs, it would be better.
12:32:19 <hagarth> If you are attending a conference, meetup or event in the near future, please list that here - https://public.pad.fsfe.org/p/gluster-events
12:32:36 <hagarth> next topic - change the time for the bug triage meeting (now each Tuesday at 12:00 UTC)
12:32:44 <hagarth> rafi: is this your entry?
12:33:06 <rafi> hagarth: that was added by Niels
12:33:13 <hagarth> I wonder if we should just have a poll for that on doodle/google
12:33:24 <hagarth> rafi: ok
12:33:25 <kkeithley> earlier or later?
12:34:11 <hagarth> kkeithley: not sure, maybe a poll with a range of options will help
12:34:37 <hagarth> rafi: would it be possible for you to post this poll?
12:35:01 <rafi> hagarth: in gluster-devel ?
12:35:25 <hagarth> rafi: both on -devel and -users. I can help you setup the poll on doodle.
12:35:39 <rafi> hagarth: I can do that
12:35:56 <hagarth> rafi: thanks, noting an AI on this one.
12:35:56 <rafi> hagarth: you can put an AI on me
12:36:04 <rafi> hagarth: sure
12:36:19 <hagarth> #action rafi to setup a doodle poll for bug triage meeting
12:36:41 <hagarth> next topic - Release testing strategy
12:36:45 <kkeithley> just keep in mind that Daylight Savings is ending soon in US, UK, and Europe in a few weeks.
12:37:23 <hagarth> kkeithley: yes, thanks for bringing that up.
12:38:08 <hagarth> I feel that we can have our pre-release testing be more deterministic
12:38:25 <hagarth> there are also requests from the community about the nature of tests being run before every release
12:38:37 <JustinClift1> eg: https://www.gluster.org/pipermail/gluster-devel/2015-October/046906.html
12:38:46 <JustinClift1> (just came in)
12:38:58 <hagarth> JustinClift1: thanks for the pointer
12:39:25 <JustinClift1> yw
12:39:51 <hagarth> I am of the opinion that we publish a plan so that there is more transparency about the nature of tests done and also have a lightweight exit criteria for the release
12:39:52 <hagarth> thoughts?
12:40:12 <sankarshan> would be nice
12:40:13 <JustinClift1> Sounds good
12:40:34 <kkeithley> A plan other than "passes smoke and regression tests"?
12:40:53 <hagarth> I also would like this to be driven by distaf in the near term, will send a mail on -devel about populating tests in distaf soon
12:40:57 <hagarth> kkeithley: of course, yes!
12:40:59 <kkeithley> And soon maybe DistAF?
12:41:19 <hagarth> kkeithley: great minds think alike ;)
12:41:34 <hagarth> rastar, msvbhat: can we work on publishing a plan in gluster.org?
12:42:19 <rastar> hagarth: yes I will take that up
12:43:04 <hagarth> #action rastar and msvbhat to publish a test exit criterion for major/minor releases on gluster.org
12:43:08 <hagarth> rastar: thanks!
12:43:17 <hagarth> moving on to next topic
12:43:26 <hagarth> close bugs filed against mainline and pre-release that are >1 year old? (except for RFEs?)
12:43:31 <hagarth> kkeithley: is this your topic?
12:43:47 <msvbhat> hagarth: When you expect this? Is it Okay for first week of Nov?
12:44:12 <kkeithley> yes. We have hundreds of open bugs against mainline and pre-release. Some are several years old.  Should we close the really old ones?
12:44:36 <hagarth> msvbhat: sounds good to me
12:44:49 <msvbhat> hagarth: Great, Will do. Thanks
12:45:31 <hagarth> kkeithley: I would vote to close them, especially if they happen to be logged from a period where mainline maps to 3.4.x or earlier
12:46:15 <hagarth> we can possibly publish a note to re-open if relevant today and bulk close such bugs
12:46:52 <hagarth> any other opinions?
12:46:53 <kkeithley> sure.  should we try to clean up the newer ones and reassign them to 3.[567].0?
12:47:12 <hagarth> kkeithley: that sounds good to me
12:48:11 <hagarth> kkeithley: let us push a note on the MLs with this plan and see other responses to this plan
12:48:33 <kkeithley> and add a note to our "release" process that reassigns {mainline,pre-release} bugs to the 3.X.0 when we release 3.X?
12:48:49 <hagarth> kkeithley: +1
12:49:03 <kkeithley> Okay. Do you want to give me that action item then?
12:49:47 <hagarth> #action kkeithley to post details of the plan on closing bugs filed against mainline and pre-release
12:50:03 <hagarth> is there any semantic difference between mainline and pre-release any more?
12:50:17 <hagarth> i doubt that, maybe we should drop pre-release ?
12:50:50 <kkeithley> that was my thought.
12:51:00 <hagarth> yeah, let us do that.
12:51:11 <hagarth> I can ask hchiramm to follow up on this one
12:51:16 <kkeithley> but first we have to close or reassign those bugs
12:51:34 <kkeithley> I can do that too, but ^^^
12:51:45 <hagarth> yes, reassign to mainline/3.x and then drop the version
12:52:17 <hagarth> kkeithley: maybe include a list of such bugs in the mail that you send? we could have the respective assignees/maintainers look into such bugs.
12:52:27 <kkeithley> sure
12:53:15 <hagarth> next topic - What's our position on the syscall wrappers? Are we supposed to be using them?
12:53:30 <hagarth> I figure the answer is yes.
12:53:52 <kkeithley> okay, good
12:53:53 <kkeithley> ;-)
12:53:54 <rastar> hagarth: syscall wrappers as in what posix xlator should call when it is calling down?
12:54:22 <kkeithley> syscall wrappers in .../libglusterfs/src/syscall.[ch]
12:54:22 <hagarth> rastar: yes and in other places where we directly invoke a syscall
12:54:42 <hagarth> are there many violations to this now?
12:54:59 <rastar> i see very few places where we use these wrappers
12:55:05 <rastar> very very few
12:55:28 <hagarth> rastar: we need to change them or else we'll break portability
12:55:30 <kkeithley> I have four patches in review to fix it
12:56:12 <hagarth> kkeithley: cool, let us get them in and have a more cleaner code base :)
12:56:14 <kkeithley> and ideas on a smoke or regression  test to catch it
12:56:16 <rastar> ok :) will review
12:56:35 <hagarth> kkeithley: makes sense.
12:57:04 <hagarth> 3 minutes away from the end of this meeting
12:57:09 <hagarth> any other topics for today?
12:57:53 <hagarth> figure not
12:57:57 <JustinClift1> Hmmm
12:58:01 <JustinClift1> Actually...
12:58:09 <hagarth> JustinClift1: go ahead :)
12:58:18 <JustinClift1> Does someone have time to answer this guy directly?
12:58:21 <JustinClift1> https://www.gluster.org/pipermail/gluster-devel/2015-October/046906.html
12:58:35 <JustinClift1> He emailed me directly, and I don't know the latest state of the regressions stuff.
12:58:45 <JustinClift1> So, looking for someone to respond to him. ;)
12:58:48 <kkeithley> It's all in the source
12:58:55 <kkeithley> what else does he want to know?
12:58:57 <sankarshan> the docs page for the regressions is dated. but he could look at the /tests
12:59:07 <JustinClift1> On a different note, sorry for the delay getting the new Gluster Forge up and running
12:59:17 <hagarth> and possibly our new plan that we will publish
12:59:23 <JustinClift1> kkeithley: He's not a dev
12:59:38 <hagarth> JustinClift1: no worries, we are almost set to nuke f.g.o :)
12:59:40 <sankarshan> essentially looking to avoid running duplicate tests
13:00:09 <JustinClift1> hagarth: I'm getting back into Python, so will be working on the forge again reasonably soon
13:00:13 <JustinClift1> Just saying. ;)
13:00:32 <hagarth> JustinClift1: cool, thanks for that update!
13:01:05 <JustinClift1> np
13:01:08 <hagarth> ok, thanks everyone for being here today. talk to you all next week.
13:01:16 <hagarth> #endmeeting