12:00:42 <rafi> #startmeeting Weekly community meeting 17-Aug-2016
12:00:42 <zodbot> Meeting started Wed Aug 17 12:00:42 2016 UTC.  The chair is rafi. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:42 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:00:42 <zodbot> The meeting name has been set to 'weekly_community_meeting_17-aug-2016'
12:01:24 <rafi> Welcome everyone to the community meeting
12:01:25 <rafi> #topic Roll call
12:01:29 <post-factum> o/
12:01:35 <rafi> please mark your presentation
12:01:40 <rafi> post-factum: thanks
12:02:17 * kshlm o/
12:02:22 <rafi> #info The agenda is available at https://public.pad.fsfe.org/p/gluster-community-meetings
12:02:28 * aravindavk is here
12:02:34 <rafi> #link https://public.pad.fsfe.org/p/gluster-community-meetings
12:02:45 * kkeithley is here
12:02:47 * ira_ still wants the bread's number.
12:02:48 * kotreshhr is here
12:02:53 * ndevos _o/
12:03:07 <rafi> we will wait some more time
12:03:27 * kkeithley doesn't see updated agenda for today
12:03:39 <ira> yeah... wait ~5m for people to get off hastur's call.
12:04:09 <rafi> kkeithley: it is there , right
12:04:29 <kkeithley> wednesday, 3 August
12:04:40 <rafi> kkeithley: i think the date was wrong
12:05:22 <rafi> let's get start
12:05:33 <rafi> #topic Next weeks meeting host
12:05:44 <rafi> who want to be the next week chair ?
12:05:45 <kkeithley> nm, refreshed the page and it's there now
12:06:13 <ndevos> kkeithley: oh, you really were living in the past!
12:06:13 <rafi> who want to be the next week chair ?
12:06:25 <rafi> ndevos: are you using IE :D
12:06:32 <rafi> kkeithley: ^
12:06:40 <ndevos> wasnt kshlm doing wvery 2nd meeting?
12:06:42 <post-factum> rafi: how rude
12:06:45 <kkeithley> What's IE?
12:06:56 <kkeithley> Internet Exploder?
12:07:02 <rafi> leave it
12:07:08 <kshlm> Yup.
12:07:19 <kshlm> I'll  be hosting next week
12:07:21 <rafi> kshlm: Can I mark your name
12:07:24 <rafi> kshlm: great
12:07:27 <rafi> kshlm++
12:07:29 <glusterbot> rafi: kshlm's karma is now 6
12:07:49 <rafi> #info kshlm will be hosting next week community meeting
12:07:52 * ndevos also isnt sure what IE is supposed to mean
12:08:08 <rafi> #topic GlusterFS-4.0
12:08:36 <rafi> ndevos: it is flexible for readers ;)
12:08:50 <rafi> kshlm: do you wanna start with glusterd-2.0
12:09:31 <kshlm> Not much.
12:09:50 <rafi> kshlm: okey
12:09:53 <kshlm> I've been stuck with a problem sending arbitrary data structs across the wire.
12:10:15 <kshlm> I'm looking at 2 ways to fix this.
12:10:22 <kshlm> That goes on this week.
12:10:28 <rafi> kshlm: okey
12:10:35 <ndevos> oh, you could always encode them in xdr!
12:10:37 <kshlm> And I just found out that etcd can be embedded into other projects.
12:10:47 <post-factum> noooooooo
12:10:50 <post-factum> etcd
12:11:12 <kshlm> post-factum, You don't want etcd?
12:11:22 <kshlm> Or was the nooooooo for xdr?
12:11:29 <ndevos> xdr++
12:11:30 <glusterbot> ndevos: xdr's karma is now 1
12:11:35 * ankitraj late join
12:11:40 <post-factum> etcd-- xdr++
12:12:00 <ndevos> post-factum: sorry, I think we lost the etcd battle a while ago :)
12:12:05 <kshlm> Well with gd2 its etcd++ and xdr--
12:12:19 <kshlm> post-factum, If we get it right, you don't need to worry about etcd.
12:12:32 * ndevos just hopes etcd does not break, and we dont have to fix it
12:12:34 <kshlm> Gluster should just work, as it always has.
12:13:05 <kshlm> ndevos, This feature was driven by requests from openshift and kubernetes
12:13:11 <rafi> kshlm: thanks for the update
12:13:13 <kshlm> So we should be covered.
12:13:19 <kshlm> rafi, I'm done.
12:13:27 <rafi> kshlm: cool
12:13:29 <rafi> Do we have any one to talk about DHT2 and/or JBR ?
12:14:07 * ndevos does not see jdarcy or shyam
12:14:37 * rafi looks for asengupta
12:15:07 <rafi> we can postpone the discussion for next week
12:15:19 <rafi> #topic GlusterFS-3.9
12:15:29 <rafi> aravindavk: Your turn ;)
12:15:38 <aravindavk> not much update
12:16:15 <aravindavk> events api feature is making good progress, didn't get chance to followup with other feature owners
12:16:16 <rafi> aravindavk: okey
12:17:53 <rafi> will move on to the next topic
12:18:10 <rafi> #topic GlusterFS-3.8
12:18:30 <ndevos> 3.8.2 was released last week, and packages seem to be available in Fedora and CentOS Storage SIG
12:18:31 <rafi> ndevos: please use the chair
12:18:41 <rafi> ndevos: great
12:19:02 <kkeithley> Debian, Ubuntu (Launchpad), and SuSE (SBS) packages also are available
12:19:09 * post-factum will post another bunch of backports for 3.8 a little bit later
12:19:13 <ndevos> in three weeks its time for 3.8.3, some patches have been merged already, and more are waiting for review
12:19:30 * msvbhat arrives bit late
12:19:40 <rafi> I see lindsay had some problem with 3.8.2
12:20:24 <ndevos> I think krutika was looking into that
12:20:39 <rafi> ndevos: okey
12:21:26 <ndevos> ... nothing more to 3.8 from me, and nobody asks questions?
12:21:44 <rafi> #info GlusterFS-3.8.3 is scheduled in first week of Sept
12:22:32 * rafi assumes that everyone are happy with 3.8
12:22:43 * ndevos hopes so too!
12:22:55 <rafi> #topic GlusterFS-3.7
12:23:07 <post-factum> we haven't make an upgrade yet to 3.8
12:23:30 <post-factum> there are 29 commits on to of 3.7.14, so .15 should be wonderful release
12:23:34 <post-factum> *on top
12:23:58 <kshlm> post-factum, Yup.
12:24:06 <rafi> post-factum: that is right, we will have a wonderful release on 30th Aug
12:24:09 <kshlm> .15 is still on target for 30th August.
12:24:16 <kshlm> 2 more weeks to go.
12:24:24 <kshlm> I'll send out a reminder at the end of this week.
12:24:49 <rafi> kshlm: when is the last date to push the code ?
12:25:23 <rafi> #action kshlm will send out a reminder for 3.7.15 time lines
12:25:42 <kshlm> That would be about half a day before the tagging time.
12:25:46 <kshlm> Make it a day.
12:26:01 <rafi> kshlm: thanks for the update
12:26:36 <rafi> moving in to the next topic
12:26:42 <rafi> #topic GlusterFS-3.6
12:27:26 <rafi> Yesterday we had bug screening section
12:27:37 <rafi> along with gluster bug triage meeting
12:27:52 <ndevos> .... and most maintainers were missing :-(
12:28:11 <rafi> ndevos++ kshlm++
12:28:11 <zodbot> rafi: Karma for devos changed to 2 (for the f24 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
12:28:55 * rafi thinks we might have got more attention if we would have sent mail to maintainers list
12:28:56 <kshlm> We'll be having one more next week.
12:29:10 <kshlm> rafi, I had sent a mail.
12:29:15 <kshlm> Didn't work.
12:29:34 <rafi> kshlm: Ooops, I didn't know that
12:29:39 <rafi> kshlm: that is sad
12:30:03 <ndevos> #link http://www.gluster.org/pipermail/maintainers/2016-August/001227.html
12:30:35 <rafi> letus hope that we will have more participation in next meeting
12:30:55 <kshlm> I'm keeping my fingers crossed.
12:31:13 <ndevos> I always hope that, for months if not years already...
12:32:11 <rafi> Just an information
12:32:11 <rafi> we have screened all the bugs in post, modified and on-qa state
12:32:31 <rafi> and we need help from all mainteners
12:32:55 <ndevos> #info 84 bugs for 3.6 still need to be screened
12:33:48 <rafi> 84 is a small number provided we have 15+ mainteners :)
12:34:02 <rafi> and other volunteers like me ;)
12:35:05 <rafi> cool, moving into the next topic #topic Related projects
12:35:06 <rafi> #topic Infra
12:35:26 <rafi> nigelb: Do you wanna give an update ?
12:36:09 <rafi> nigelb is not around
12:36:19 <rafi> #topic NFS ganesha
12:36:37 <rafi> skoduri, kkeithley, ndevos: ^
12:36:46 <kkeithley> so some prelimary testing of 2.4dev26 looks good from stability and perf standpoint
12:36:56 <kkeithley> (we're up to dev29 in the mean time)
12:37:18 <kkeithley> rc1 will probably get tagged this week and we should GA pretty soon
12:38:20 <rafi> kkeithley: great any  major points to keep in mind for gluster vfs
12:38:25 <kkeithley> most of the FSALs have been refactored to x-api, the new, extended api for FSALs
12:38:33 <kkeithley> no, not really
12:38:55 <kkeithley> the remaining fsals that would be nice to have refactored aren't blockers
12:39:14 <kkeithley> e.g. gpfs (IBM's clustered fs)
12:39:19 <kkeithley> gpfs2
12:39:33 <rafi> kkeithley: i hope gluster vfs doesn't require a re factor
12:39:42 <ndevos> gpfs is correct, it is not gfs2 ;-)
12:39:57 <kkeithley> the gluster FSAL has already been refactored.
12:40:06 <ndevos> rafi: skoduri did that already
12:40:18 <rafi> kkeithley: ohh wonderful
12:40:29 <rafi> skoduri++
12:40:29 <zodbot> rafi: Karma for skoduri changed to 1 (for the f24 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
12:40:29 <glusterbot> rafi: skoduri's karma is now 3
12:40:31 <kkeithley> that was a prereq for upcall
12:40:38 <kkeithley> IIRC
12:40:51 <rafi> skoduri: come forward take our appreciation ;)
12:40:56 <rafi> *and
12:41:02 <kkeithley> yes, skoduri++
12:41:06 <skoduri> rafi, thanks :)
12:41:20 <kkeithley> skoduri++
12:41:21 <glusterbot> kkeithley: skoduri's karma is now 4
12:41:33 * rafi gives a virtual gift
12:41:43 <rafi> cool
12:41:49 <ndevos> skoduri++ have a cookie!
12:41:52 <zodbot> ndevos: Karma for skoduri changed to 2 (for the f24 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
12:41:56 <rafi> #topic Gluster samba
12:42:28 <ira> Just got off a long call about md-cache.  Expect more updates.
12:42:28 <skoduri> wohoo thanks :D
12:42:46 <rafi> ira: that is right
12:42:58 * ira nods.
12:43:07 <kkeithley> skoduri++ cookie
12:43:13 <rafi> ira: poornima was mentioning about a 300% increase in performance for readdirp
12:43:17 <ira> You asked.  That's the biggest news on the Samba front.  Samba 4.5 is going along.
12:43:28 <ira> rafi, we're not sure what the numbers will be yet.
12:43:43 <rafi> ira: ohh okey
12:43:46 <ira> There's been all types of problems in testing...
12:43:54 <rafi> excited to see the numbers
12:44:08 <ira> We all are... but understand the caution being shown ;).
12:44:39 <ira> I think we may get more next week?  Not sure.
12:44:41 <kkeithley> Where did poornima get the 300% readdirp increase, in the vfs? Or in gfapi?
12:45:06 <skoduri> I think iin SMB/gfapi
12:45:22 <ira> basically, I've heard all types of numbers.
12:45:29 <ira> Please wait for the finals...
12:45:29 <kkeithley> does that mean ganesha will or won't benefit?
12:45:34 <ira> Should.
12:45:41 <ira> Or could.
12:46:01 <ndevos> should, there is a reason I'm intersted in it ;-)
12:46:02 <ira> It may not benefit as much as Samba because it didn't have as far to go ;).
12:46:30 <rafi> the whole gluster will get the fruits of the work as there are some changes in md-cache
12:46:33 <kkeithley> we'll take anything/everything we can get ;-)
12:46:39 <skoduri> yeah may or may not  ganesha does caching too
12:46:47 * ira agrees with rafi.  That was the original design.
12:46:52 <kkeithley> poornima++ have a cookie!
12:47:11 <ira> kkeithley: I suggest you mail her a small crate of 'em.
12:47:34 <rafi> cool
12:47:46 <rafi> will move on to the next
12:47:55 <rafi> #topic Last weeks AIs
12:47:58 <ira> Yeah... that's the big news... Samba release seems to be on schedule.
12:47:58 * kkeithley wonders what the zodbot cookie magic is
12:48:00 <ira> :)
12:48:22 <rafi> #topic kshlm to setup a time to go through the 3.6 buglist one last time (everyone should attend).
12:48:22 <ira> Give it your magic cookie and find out kkeithley? ;)
12:48:37 <rafi> i think this is partially done
12:48:42 <kshlm> rafi, I did, and you were there.
12:48:50 <ndevos> kkeithley: one cookie per fedora release cycle, and only for Fedora contributors
12:49:09 <rafi> kshlm: should we keep the AI , as we are continuing the excise for next week
12:49:18 <kshlm> Keep it.
12:49:28 <kshlm> I'll do it again.
12:49:35 <rafi> kshlm: cool, will close this next week
12:49:36 <ndevos> rafi: and call it out in the meeting minutes, remind everyone to attend next Tuesday
12:49:47 <rafi> ndevos: I can do that
12:50:23 <rafi> #action kshlm to send reminder to go through the 3.6 buglist one last time (everyone should attend).
12:50:34 <rafi> moving on
12:50:50 <rafi> #topic open floor
12:51:05 <rafi> #topic Glusto - libraries have been ported by the QE Automation Team and just need your +1s on Glusto to begin configuring upstream and make available.
12:51:34 <kkeithley> that's last weeks open floor
12:51:52 <rafi> kkeithley: sorry my bad
12:52:01 <kshlm> Oops, I guess I forgot to clean that out.
12:52:03 <rafi> #topic * umbrella versioning for glusterfs in bugzilla (i.e. 3.9, not 3.9.0, 3.9.1, etc.  starting with 3.9 release)
12:52:18 <kshlm> +1 from me.
12:52:22 <post-factum> +1
12:52:22 <kkeithley> anyone have any objections?
12:52:40 <rafi> kkeithley++ for suggesting
12:52:43 <ndevos> I like it a lot, we ask for the exact version in the comment anyway
12:52:56 <kkeithley> it's really ndevos who has been suggesting it for a long time now
12:53:11 <rafi> ndevos++ kkeithley++
12:53:42 * rafi thinks that glusterbot is dead with too much cookie
12:53:59 <kkeithley> I don't hear any objections. I'll send email to gluster-devel mailing list
12:54:17 <rafi> kkeithley: great
12:54:25 <rafi> i will mark this for reference
12:55:32 <rafi> #action kkeithley will send more information to gluster ML's about changing the bugzilla versioning  to umbrella
12:56:02 <rafi> coool
12:56:16 <rafi> any thing else to discuss
12:56:18 <rafi> ?
12:57:31 <ndevos> atinm: something on restructuring bugzilla components/subcomponents?
12:58:26 <rafi> ndevos: since atinm is not around , do you mind moving this discussion to next week
12:58:46 <ndevos> rafi: thats fine, he wanted to start an email thread anyway
12:58:48 <rafi> ndevos: if so, I will add this topic for next week open floor
12:59:19 <rafi> Does anybody has anything to bring up for one last time ?
12:59:37 <rafi> if not I will end this meeting
13:00:03 * rafi looks on clock and it is on time :D
13:00:12 <rafi> #endmeeting