12:00:50 <ndevos> #startmeeting
12:00:50 <zodbot> Meeting started Tue Jan  6 12:00:50 2015 UTC.  The chair is ndevos. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:02 <ndevos> Welcome to todays Gluster Bug Triage meeting
12:01:10 <ndevos> The agenda is here: https://public.pad.fsfe.org/p/gluster-bug-triage
12:01:19 <ndevos> #topic Roll Call
12:01:27 * kkeithley_ is here
12:01:30 <ndevos> Who's attending today?
12:01:46 * hchiramm_ in .. but have to leave early ..
12:01:55 <ndevos> lalatenduM: ?
12:02:08 <hchiramm_> lalatenduM, ^^
12:02:09 <hchiramm_> :)
12:02:12 <lalatenduM> yesh
12:02:15 <lalatenduM> yeah :)
12:02:17 <ndevos> hehe
12:02:26 * lalatenduM is here
12:02:54 <ndevos> #topic Status of last years action items
12:03:00 <ndevos> #topic lalatenduM will send a reminder to the users- and devel- ML about (and how to) fixing Coverity defects
12:03:12 <lalatenduM> not done
12:03:14 * ndevos can not remember seeing an email
12:03:15 <ndevos> ah
12:03:43 <ndevos> lalatenduM: got time to send one this week?
12:03:52 <lalatenduM> ndevos, not sure what kind of info I should out thr
12:04:19 <lalatenduM> also I think hagrath wanted some diff kind of report
12:04:26 <ndevos> link to the coverity page/project and related wiki pages?
12:04:52 <lalatenduM> ndevos, that I anyway send everytime with the report
12:05:05 <ndevos> yeah, hagarth wanted to list the number of open issues - like the table on the coverity project site
12:05:39 <lalatenduM> ndevos, I did not a chance to talk abt these to hagarth
12:05:47 <lalatenduM> get a chance*
12:06:04 <ndevos> the report is one thing, the attracting of contributors that send a patch for a CID is different
12:06:42 <lalatenduM> ok, I will work with you guys on this , will compose a mail and send it to review to you folks
12:07:02 <ndevos> lets plan to send an email with detailed hints on how new contributors can find a CID and send a fix for it
12:07:10 <ndevos> yes, that
12:07:22 <ndevos> #topic hchiramm request NetBSD to be added to the hardware list in Bugzilla
12:07:30 <hchiramm_> ndevos, sorry .. missed it
12:07:37 <hchiramm_> will finish it by this week
12:07:48 <ndevos> okay, we'll check it next week again :)
12:07:52 <hchiramm_> sure :)
12:08:00 <ndevos> #topic lalatenduM initiate a discussion with the RH Gluster team to triage their own bugs when they report them
12:08:01 <kkeithley_> Hardware?
12:08:11 <hchiramm_> yeah, before we move on
12:08:13 <ndevos> Hardware like FreeBSD
12:08:21 <lalatenduM> ndevos, not done, lets carry it to next week
12:08:26 <ndevos> lalatenduM: ok :)
12:08:31 <kkeithley_> NetBSD isn't hardware
12:08:33 <lalatenduM> last two weeks were holidays
12:08:55 <ndevos> yeah, same here :)
12:09:04 <hchiramm_> it is in OS
12:09:11 <ndevos> oh, and FreeBSD is hardware?
12:09:32 <kkeithley_> no, FreeBSD isn't hardware either
12:09:40 <hchiramm_> freebsd is in "OS" list
12:10:10 <ndevos> well, I only have a Hardware field in my Bugzilla, one for the architecture, and one for the os
12:11:20 <ndevos> #topic pranithk to report how his team is assigning triaged bugs
12:11:45 <ndevos> 13:02 < pranithk> ndevos: I just checked the list of triaged bugs on replicate and re-assigned the bugs to kritika, anuradha, ravi. There are 4 bugs which I need to spend some time on before I do something with them.
12:11:59 <kkeithley_> I see what you're referring to. I withdraw my comment
12:12:41 <ndevos> I dont think it really matters how its called, NetBSD should just get added to the same list as Linux and FreeBSD and all
12:12:53 <kkeithley_> agreed
12:13:14 <ndevos> it seems that pranithk just assigns bugs to other developers... which can work
12:13:28 <hchiramm_> ndevos, I will do that
12:13:38 <ndevos> hchiramm_: yeah, thanks
12:13:41 <hchiramm_> np
12:14:42 <ndevos> I dont think having the maintainer assigning bugs to other developers without discussion would work... and I doubt all maintainers go through the Triaged bugs anyway
12:15:05 <ndevos> so, the AI "hagarth will look for somebody that can act like a bug assigner manager kind of person" is still very much valid :)
12:15:05 <hagarth> ndevos: right
12:15:13 <ndevos> ah, he's here!
12:15:16 <ndevos> #topic hagarth will look for somebody that can act like a bug assigner manager kind of person
12:15:18 <hagarth> ndevos: it still is on
12:15:23 <lalatenduM> ndevos, I am not sure of it will work, we have to wait and watch :)
12:15:26 <ndevos> I thought so :)
12:15:45 <hagarth> ndevos: however how about having a rotating role in the community till we find one such person?
12:16:34 <ndevos> hmm, that is possible, but we need someone that knows a little about the workload of different developers
12:16:47 <ndevos> and, well, know all the areas all the developers are working on
12:17:22 <hagarth> ndevos: right, let me think a little more about it. How does it work in Fedora, are you aware?
12:17:54 <lalatenduM> they track blockers I think
12:17:58 <lalatenduM> not all bugs
12:18:11 <ndevos> hagarth: in Fedora all the package maintainers work on their own bugs, and forward the bugs to the upstream communities (hopefully with patch_
12:18:28 <hagarth> ndevos: ok
12:18:30 <kkeithley_> maybe BZs can be assigned to a "lead" in each area who will either keep them him- or herself, or reassign them to someone else in their team?
12:18:40 <hchiramm_> r  v planning to rotate this task among component maintainers ?
12:18:59 <hagarth> kkeithley_: the leads would naturally be the sub-maintainers by default I think
12:19:19 <hagarth> unless the sub-maintainer delegates it to somebody else
12:19:46 <ndevos> kkeithley_: I'm not very comfortable to list all nfs+Triaged bugs and assign them to any of the developers that know nfs
12:21:09 <kkeithley_> but in the case of NFS you (you=ndevos in the case of NFS) could have a conversation with the particular developers first before you assign the BZ to them.
12:22:09 <kkeithley_> anyway, it was just a thought
12:22:32 <ndevos> yes, it is possible, but not something I very much look forward to do :)\
12:22:52 <hagarth> ndevos: I suspect that having a single person do it for the entire project is not scalable either
12:23:05 <ndevos> oh, and that is only 11 bugs btw - nfs isnt that 'busy'
12:23:43 <ndevos> hagarth: maybe not... its difficult to say
12:24:17 <hagarth> ndevos: I think it would be optimal if sub-maintainers triage bugs in their respective components & decide the next steps
12:24:29 <ndevos> and another 9 nfs bugs NEW/not-Triaged
12:25:22 <ndevos> hagarth: yes, I'm trying to do that for nfs - but it makes me feel uncomfortable to assign bugs to others
12:26:11 <hagarth> ndevos: maybe send a list of triaged bugs on gluster-devel & let nfs developers pick up voluntarily?
12:27:24 <ndevos> hagarth: yeah, something like that would have my preference - maybe have a query with triaged bugs to be taken by developers too
12:28:01 <ndevos> and each group working on certain components would need to keep the list of triaged+NEW bugs as small as possible
12:28:01 <hagarth> ndevos: yes, sending it on gluster-devel might encourage new developers to contribute too
12:29:19 <ndevos> #action ndevos will send an email to gluster-devel with some standard bugzilla queries/links to encourage developers to take NEW+Triaged bugs
12:29:36 <ndevos> I guess we'll see how that works out
12:30:10 <ndevos> but getting the bugs assigned is the next step after we Triaged them, and thats quite an important step to get things fixed :)
12:31:10 <ndevos> Hmm, lets move the topics around
12:31:15 <ndevos> #topic Open Floor
12:31:40 <ndevos> we got one topic for this: http://bugs.cloud.gluster.org/
12:31:42 <hagarth> lalatenduM: wanted to discuss about coverity emails that we send
12:32:31 <lalatenduM> ndevos++ , http://bugs.cloud.gluster.org/
12:33:02 <lalatenduM> it looks nice, however I have not gone in to details
12:33:06 <ndevos> so, I've taken https://github.com/tracyajones/nova_bugs and munged it to work with Bugzilla and our Gerrit instance
12:33:15 <hagarth> ndevos: fantastic!
12:33:42 <ndevos> it can be used to see what bugs are not assigned (assigned to bugs@gluster.org) and all
12:33:55 <jdarcy> What do the fields in "Review Status" mean?
12:34:14 <ndevos> M=Modified, A=Abandonde, N=New
12:34:36 <ndevos> and it is a counter of patches
12:34:55 <jdarcy> So M=Merged would work too.
12:35:03 <ndevos> yes
12:35:10 <ndevos> :)
12:35:51 <jdarcy> Looks nice.
12:35:54 <ndevos> the labels in the webui are mainly based on how Launchpad calls things, Bugzilla is a little different
12:36:37 <ndevos> and, I think some bugs throw parser errors and get skipped - but I'm not sure why/when that happens
12:36:53 <jdarcy> Priority gets sorted as: high, low, medium.  Should be easy to fix.
12:37:08 <ndevos> the details are from a snapshot of Bugzilla and Gerrit, gathered by cron.daily (UTC)
12:38:04 <ndevos> yeah, I think it sorts it alphabetically - "easy to fix" for people knowing JavaScript
12:38:45 <jdarcy> Is the script posted somewhere?  I could take a look.
12:39:05 <ndevos> https://forge.gluster.org/bugzappers/bug-status-webui
12:39:46 <lalatenduM> ndevos, have you cloned it to github also?
12:39:48 <ndevos> other feedback and patches are very welcome :)
12:40:01 <ndevos> lalatenduM: no, only on the forge
12:40:18 <lalatenduM> ndevos, github plz :)
12:40:49 <ndevos> lalatenduM: in that case I would like to have it in the gluster organisation - hagarth?
12:40:57 <hagarth> ndevos: sure
12:41:26 <ndevos> hagarth: can you clone https://github.com/tracyajones/nova_bugs into the org and give me commit access? I'll update the repo then
12:41:38 <hagarth> ndevos: will do
12:41:42 <ndevos> hagarth: thanks!
12:42:18 <ndevos> hagarth: oh, and 'nixpanic' is my github login
12:42:34 <hagarth> ndevos: noted
12:42:41 <ndevos> hagarth, lalatenduM: you want to discuss about the Coverity email?
12:43:18 <lalatenduM> yes, let me give me a link to an existing mail
12:44:03 <hagarth> lalatenduM: I think it would be effective if we send a component-wise breakdown of bugs
12:44:23 <hagarth> something like what you get in the dashboard of https://scan.coverity.com/projects/987 when you login
12:45:55 <lalatenduM> here is a mail example http://www.gluster.org/pipermail/gluster-devel/2014-November/043059.html
12:46:03 <lalatenduM> I mean old mail
12:46:08 <lalatenduM> hagarth, sure
12:46:26 <lalatenduM> hagarth, will try to get similar info in to the mail
12:46:48 <hagarth> lalatenduM: cool, thanks!
12:47:29 <lalatenduM> as ndevos mentioned the mail should invite new contributors , need some help on that
12:48:28 <lalatenduM> also the how to part
12:48:30 <hagarth> lalatenduM: yeah
12:48:30 <jdarcy> Coverity bugs are often pretty accessible.  Just need to understand code, not the deep details of what GlusterFS does.
12:48:54 <lalatenduM> right
12:49:08 <ndevos> yes, they are good candidates for new contributors - just like our EasyFix bugs
12:49:10 <lalatenduM> new contributors should pick it
12:49:45 <lalatenduM> ndevos, yeah
12:49:56 <ndevos> lalatenduM: got all you need, or want more?
12:50:06 <lalatenduM> ndevos, should be ok
12:50:21 <lalatenduM> will send you the draft mail for review
12:50:40 <ndevos> okay, cool
12:50:53 <ndevos> no other topics for the open floor?
12:51:09 <ndevos> #topic Group Triage
12:51:21 <ndevos> there are no bugs waiting for bugs@gluster.org
12:51:29 <ndevos> and none have been added to the etherpad
12:51:42 <ndevos> so, here is a list of bugs from the last two weeks
12:51:44 <ndevos> https://bugzilla.redhat.com/buglist.cgi?bug_status=NEW&chfield=[Bug creation]&chfieldfrom=-2w&chfieldto=Now&f1=keywords&o1=notsubstring&product=GlusterFS&v1=Triaged
12:52:20 <ndevos> 12 bugs in total, please issue IRC locks when triaging them
12:52:40 * ndevos locks 1179076
12:53:19 <lalatenduM> 1177167
12:53:57 <ndevos> kkeithley_: 1177411 is for you :)
12:54:21 * ndevos 1177928
12:54:39 <kkeithley_> what git submodule?
12:54:58 <jdarcy> 1179208 is clearly mine.
12:55:18 * hagarth drops off now..
12:55:34 <ndevos> kkeithley_: wait, we dont use that here - nfs-ganesha has the libntirp as submodule
12:56:30 <kkeithley_> no, it's $%^&* cmockery cruft
12:56:47 <kkeithley_> fsck
12:56:59 <ndevos> yeah, cmockery2 was a git submodule and that broke *much*
12:57:30 <ndevos> but it is not a submodule anymore... not sure what traces are left
12:59:09 <jdarcy> 1178619
12:59:31 * lalatenduM locks 1178619
13:00:22 * kkeithley_ things sqrt(-1) is the number of unit tests we actually have
13:00:30 <lalatenduM> ahh 1178619 is done
13:00:53 <jdarcy> Who's working on [HC] stuff?
13:01:10 <lalatenduM> jdarcy, for you too https://bugzilla.redhat.com/show_bug.cgi?id=1178590
13:01:34 <ndevos> what is HC ?
13:01:47 <jdarcy> Hyperconverged, I think.
13:01:55 <ndevos> ah, yes, probably
13:01:56 <jdarcy> Looks like someone started up a project without telling the rest of us.
13:02:00 <lalatenduM> yeah
13:02:21 * jdarcy looks for the feature page.
13:02:50 <jdarcy> Not sure how 1178590 is particularly mine, but I was going to triage it anyway.
13:02:59 <ndevos> okay, time for this meeting is up - but please triage the last few bugs :)
13:03:31 <ndevos> thanks all for attending, we'll repeat this meeting again next week!
13:03:37 <ndevos> #endmeeting