12:01:50 <ndevos> #startmeeting
12:01:50 <zodbot> Meeting started Tue Jun 16 12:01:50 2015 UTC.  The chair is ndevos. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:56 <ndevos> #info Agenda: https://public.pad.fsfe.org/p/gluster-bug-triage
12:02:06 <ndevos> #topic Roll Call
12:02:11 * atinm arrives
12:02:20 <ndevos> #chair atinm
12:02:20 <zodbot> Current chairs: atinm ndevos
12:03:04 <atinm> roll call please?
12:03:11 <ndevos> I think kkeithley_ and soumya have joined too
12:04:26 <atinm> so let's move on
12:04:38 <ndevos> well, I guess we can start the group triage :)
12:04:50 <ndevos> #unchair ndevos
12:04:50 <zodbot> Current chairs: atinm ndevos
12:04:56 <ndevos> hmm
12:05:15 <ndevos> #topic Group Triage
12:05:16 <atinm> #topic Group triage
12:05:20 <ndevos> :D
12:05:22 <atinm> ndevos, :)
12:05:35 <atinm> #info tbug triaging link : http://goo.gl/WuDQun
12:05:40 <ndevos> #info no bugs are waiting on bugs@gluster.org
12:05:59 <atinm> #info we have 22 bugs to triage today
12:06:12 * atinm locks 1231024
12:06:27 * ndevos locks 1232199
12:07:00 * kkeithley_ is here
12:07:08 * ndevos locks 1232172
12:08:31 * ndevos locks 1231686
12:09:23 * atinm locks 1231334
12:10:06 * ndevos locks 1231539
12:10:45 * atinm locks 1231195
12:11:48 * ndevos locks 1231195
12:11:52 <ndevos> ah, noooooo
12:11:59 <atinm> ndevos, :)
12:12:04 * ndevos locks 1231264 instead
12:13:07 * ndevos locks 1231202
12:14:32 * atinm locks 1229914
12:15:10 * ndevos locks 1231207
12:15:39 * atinm locks 1230169
12:17:34 * atinm locks 1230857
12:18:40 * ndevos locks 1231265
12:18:46 * kkeithley_ locks 1232155
12:19:35 * ndevos locks 1231983
12:20:13 * kkeithley_ locks 1231202
12:21:32 * ndevos locks 1230671
12:22:00 <atinm> ndevos, so many bugs on how to find total number of mounts
12:22:20 <ndevos> atinm: yes, quite
12:22:23 <atinm> ndevos, we could have had only one for all protocols
12:22:32 <ndevos> atinm: and I think we have tools for that already?
12:22:58 <atinm> ndevos, I believe so
12:23:17 <atinm> ndevos, you did mention in gluster-devel for NFS as far as I remember
12:23:39 <kkeithley_> how are glusterfs client mounts different from fuse client mounts?
12:23:45 <ndevos> atinm: yes, and Bipin who opened those bugs copied my anser in the bug for nfs :)
12:24:17 <ndevos> kkeithley_: glusterfs client can be gfapi/selfheal/nfs/... too?
12:24:52 * ndevos locks 1229808
12:25:14 <kkeithley_> those aren't really _mounts_ though.  Or is this everything's a file, everything's a mount, everything's a thing
12:25:38 <atinm> ndevos, kkeithley_ : I would suggest one person triaging all of these bugs ?
12:26:15 <ndevos> kkeithley_: yes, everything's a thing
12:26:19 <kkeithley_> pick one, close the rest as duplicates. I already set NeedInfo on the NFS one
12:26:51 <ndevos> I put needinfo for the nfs and an other one too...
12:27:17 * ndevos locks 1228785
12:27:47 <atinm> ndevos, that's a old one
12:27:54 <atinm> ndevos, I think we are done with the list
12:27:55 <ndevos> atinm: yes, looks like it
12:30:14 <ndevos> atinm: open floor?
12:30:25 <atinm> ndevos, yes
12:30:32 <atinm> #topic Open Floor Discussion
12:31:05 <ndevos> #info reminder, bugs and the status of their patches in Gerrit: http://bugs.cloud.gluster.org/
12:31:43 <atinm> Kaushal has proposed a change for the criteria to trigger regression run
12:31:49 <ndevos> I was wondering if it makes sense to extend this meeting to not only do bug triage, but also discuss on testing
12:32:07 <atinm> probably he will talk about it on tomorrow's community meeting
12:32:38 <ndevos> yeah, I did read his email quickly, but have to think about formulating an answer
12:33:09 <atinm> and NetBSD regression is definitely a pain, it takes ages to finish and if you are unlucky you wouldn't realize that it even didn't run (probably crashed in between)
12:33:32 <atinm> ndevos, any progress on automating the bugzilla status movements?
12:33:48 <ndevos> yes, but with the recent dns changes on build.gluster.org, things should be more stable?
12:34:04 <ndevos> no, no progress on the automated bugzilla changes
12:34:24 <ndevos> we need to find someone that can spend a day or so to work that out
12:34:52 <atinm> ndevos, I think Kaushal's proposal + launching docker containers from the vms for each run should bring things stable?
12:35:15 <atinm> ndevos, hmm
12:35:16 <ndevos> atinm: it does not help with netbsd tests, no docker for netbsd for all I know
12:35:58 <ndevos> atinm: I would prefer a solution that works on all the OS's that we support, with no or minimal changes
12:35:59 <atinm> ndevos, Ahh!! I didn
12:36:02 <atinm> oops
12:36:08 <atinm> I didn't know that
12:36:57 <ndevos> I could be wrong, but a few weeks/months ago there was nothing I could find to make docker work on netbsd
12:37:25 <ndevos> so, my preference to speed up tests, is by splitting the test-cases into smaller chunks, and run those concurrent
12:38:21 <ndevos> we should have dynamic poweron/poweoff of VMs, so that we can have more VMs, but they would not be running (=costing us money) 24x7
12:39:12 <atinm> ndevos, running concurrent test is hard, specially for glusterds
12:39:32 <ndevos> atinm: not really, they do not need to run on the same VM
12:39:57 <ndevos> atinm: also woth docker containers, the tests would need to get split up somehow
12:40:07 <atinm> ndevos, we are already vm short, how will that help?
12:40:20 <atinm> ndevos, for docker containers it makes sense
12:40:46 <ndevos> atinm: a lot of the time, not all VMs are used, we can turn them off and save our budget, only turn them on when needed
12:41:00 <atinm> ndevos, ahh I didn't see your last statement
12:41:13 <atinm> ndevos, if we can do it, that will be great
12:41:23 <ndevos> that way, we could even have more VMs running during peak hours, and none during the weekends
12:42:04 <atinm> ndevos, right
12:42:06 <ndevos> atinm: G_Garg started to look into a Jenkins plugin that can poweron/poweroff VMs when needed, I need to follow up with him on that
12:43:02 <ndevos> and VMs are a well understood technology, we all know what to do if something fails with them, with containers thats different...
12:43:40 <atinm> ndevos, ok, I can talk to G_Garg on this
12:44:10 <ndevos> atinm: yeah, please check, he's our Java programmer for this :D
12:44:20 <atinm> :)
12:44:37 <atinm> ndevos, the moment you say this he quits :D :D
12:44:49 <ndevos> atinm: yeah, he's hiding!
12:44:55 <ndevos> ... maybe in shame?
12:45:56 * ndevos does not have any other topics for the open floor
12:46:26 <atinm> so that's all about it I believe
12:46:32 <ndevos> I think so
12:46:35 <atinm> all right
12:46:41 <atinm> thanks ndevos & kkeithley_
12:46:44 <atinm> ndevos++
12:46:47 <ndevos> thanks atinm++ !
12:46:48 <atinm> kkeithley++
12:46:51 <atinm> #endmeeting