12:01:24 <ndevos> #startmeeting
12:01:24 <zodbot> Meeting started Tue Dec  2 12:01:24 2014 UTC.  The chair is ndevos. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:01:24 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
12:01:33 <ndevos> Agenda: https://public.pad.fsfe.org/p/gluster-bug-triage
12:01:40 <ndevos> #topic Roll Call
12:01:48 <ndevos> Welcome today, who's there?
12:01:55 * jdarcy is
12:02:50 <ndevos> Lala and Humble mentioned that they probably won't join today
12:02:59 <ndevos> kkeithley_: is probably around?
12:03:15 * kkeithley_ is here
12:03:32 <ndevos> and hagarth would be joining later
12:04:50 <ndevos> #topic Status of last weeks action items
12:05:10 <ndevos> #topic Humble to remove pre-2.0 from versions
12:05:39 <ndevos> the versions are gone in bugzilla, users can not file new bugs with old versions
12:05:49 <ndevos> but the old bugs with old versions are still there
12:05:57 <ndevos> so, I think thats good
12:06:15 <ndevos> #topic ndevos will update the unsupported bugs with a message about being unsupported, and request for testing
12:06:47 <ndevos> I've updated all old bugs with a note, and will close them next week when they have not received an update or version change
12:07:32 <ndevos> none if the other items from last week can be discussed now, we're missing the people that are assigned :-/
12:07:48 <ndevos> #topic Group Triage
12:08:21 <ndevos> we dont have any bugs that NEEDINFO from  gluster-bugs@redhat.com
12:08:44 <ndevos> -> New bugs to triage (filed since last week)
12:09:03 <ndevos> 9 new bugs: http://goo.gl/0IqF2q
12:09:12 <ndevos> lets use IRC locking again :)
12:09:31 <jdarcy> 1168809
12:09:42 <ndevos> 1168167
12:10:26 <kkeithley_> 1169707
12:11:55 <jdarcy> 1168574
12:16:16 <hagarth> ndevos: which is the list that's being triaged now?
12:16:34 <ndevos> hagarth: http://goo.gl/0IqF2q
12:16:42 <jdarcy> 1168897
12:18:15 <jdarcy> We don't seem to have a "packaging" component.  What's the proper component for a bug in a package upgrade?
12:18:22 <hagarth> build
12:18:28 <hagarth> ?
12:18:43 <hagarth> 1169302
12:18:49 <jdarcy> WFM
12:19:10 <ndevos> yeah, most things got to build for that
12:20:38 * hagarth 1169707
12:20:46 <ndevos> 1169236
12:21:38 <hagarth> should we support hostnames that have an underscore in them?
12:22:01 <hagarth> bz 1169707 is about supporting underscores in hostnames
12:22:21 <ndevos> why not?
12:22:31 <jdarcy> Why is such support not automatic?  Do we use underscore for something else?
12:22:52 <hagarth> as per rfc 1034, underscore is not a valid character for a hostname
12:22:58 <jdarcy> IIRC, underscore is technically (i.e. per RFC) not allowed, but it's not that uncommon.
12:22:58 <kkeithley_> hmmm,   1169707 is about not being able to probe hostnames with '_' in them.  RFCs 952, 1035,  and 1123 say '_' is not a legal character in a hostname.
12:23:14 <kkeithley_> ah, hagarth didn't see my "lock"
12:23:44 <hagarth> underscore is not used for anything else in glusterd, our hostname validation fails if an underscore is present
12:23:55 <hagarth> kkeithley_: unfortunately I missed your advisory lock :)
12:24:01 <kkeithley_> no prob
12:24:07 <ndevos> isnt there an update for host/domain names? even many utf-8 characters should be allowed now?
12:25:14 <kkeithley_> I was looking for a superceding RFC
12:25:31 <hagarth> i think it is 3629
12:26:55 <jdarcy> Either way, seems like something we should fix.
12:26:56 <kkeithley_> the wikipedia article http://en.wikipedia.org/wiki/Hostname says _ is still not allowed
12:27:21 <jdarcy> Also, 1169236
12:27:29 <jdarcy> (that's a lock)
12:27:55 <jdarcy> Oops, ndevos beat me.  nvm
12:29:09 * hagarth 1169331
12:29:13 <jdarcy> Just saw a coyote cross our back yard.
12:29:19 <hagarth> overclk: do you want to take a look into 1169331?
12:29:43 <ndevos> jdarcy: glad to know you're paying attention :)
12:30:04 * ndevos locks 1169701
12:30:17 <overclk> hagarth, yeh, I'll take care of that.
12:30:30 <hagarth> overclk: thx
12:30:31 <jdarcy> ndevos: Predators large enough to eat cats tend to distract me.
12:32:17 * hagarth has never seen a coyote
12:34:03 * ndevos would love to see those from his window too
12:34:15 <kkeithley_> even RFC 2181 still references 1035
12:34:51 <kkeithley_> I get coyotes, deer, and turkeys in my garden (back yard)
12:35:31 <hagarth> kkeithley_: nice
12:36:42 <kkeithley_> so we are agreed that we ought to relax hostname checking for 1169707?
12:36:45 <ndevos> well, once I had to add support for numeric hostnames, see http://review.gluster.org/4017
12:36:56 <jdarcy> Also foxes, skunks, raccoons.  Even thought I saw a fisher (large weasel relative) once.
12:36:56 <kkeithley_> numeric hostnames are legal
12:37:33 <jdarcy> kkeithley_: Yes, I think it's not our business to enforce RFC restrictions except where they remove ambiguity, and AFAICT this one doesn't.
12:38:09 <hagarth> valid_host_name() in libglusterfs is the one which does hostname validations. We can relax this restriction there.
12:41:00 <hagarth> ndevos: reg 1169236, I wonder if fio works fine with fuse
12:41:44 <hagarth> maybe we could ask the bug reporter to check that?
12:41:49 <ndevos> hagarth: yeah, would be interesting
12:42:04 <ndevos> hagarth: sure, Kiran is very helpful in testing, he'll surely do that
12:44:49 <ndevos> okay, so all new bugs have now been triaged, thanks!
12:45:47 <ndevos> #topic hagarts action items from last week
12:46:02 <ndevos> hagarth will look for somebody that can act like a bug assigner manager kind of person
12:46:17 <ndevos> hagarth: any update on that front?
12:46:36 <ndevos> or, on hagarth should update the MAINTAINERS file, add current maintainers and new components like Snapshot
12:47:09 <hagarth> ndevos: bug assigner manager in progress. I hope to provide an update soon.
12:47:17 <ndevos> hagarth: cool
12:47:51 <hagarth> ndevos: I will be updating the maintainers after Dec 8th.
12:49:10 <ndevos> hagarth: okay, so the sub-maintainers and the MAINTAINERS file should match?
12:49:35 <ndevos> hagarth: in that case, we need a sub-maintainer for snapshot an possibly other features too?
12:50:14 <ndevos> when triaging bugs, we sometimes check the MAINTAINERS file so that we can put certain developers on CC to get their attention
12:50:19 <hagarth> ndevos: yes, ec and other features too
12:50:46 <ndevos> hagarth: ok, we'll verify it next week
12:50:55 <hagarth> ndevos: maybe we need to draw a distinction between MAINTAINERS and feature owners
12:51:26 <jdarcy> It would be nice if the owner got changed automagically when the component is changed, but it doesn't seem to happen currently.
12:51:29 <ndevos> hagarth: yes, I think MAINTAINERS in the git sources does not need to be the same as the sub-maintainers in Gerrir
12:51:33 <ndevos> *Gerrit
12:51:55 <hagarth> ndevos: yeah
12:52:06 <ndevos> jdarcy: you mean in Bugzilla? bugs@gluster.org is the default owner for all components
12:53:03 <ndevos> jdarcy: and maintainers or component owners are expected to use the bugzilla notifications (mailinglist, rss, ...) to get informed about new bugs/updates
12:54:11 <ndevos> also, there is no 1:1 match between Bugzilla components and our MAINTAINERS file :-/
12:54:32 <ndevos> #topic Open Floor
12:54:45 * jdarcy (don't fall in)
12:54:46 <ndevos> we got 5 minutes to discuss other business
12:55:37 <ndevos> nothing to add?
12:55:47 <jdarcy> Nope.
12:56:06 <kkeithley_> not me
12:56:09 <ndevos> okay, I guess we'll talk to eachother next week again
12:56:13 <ndevos> if not before
12:56:21 <ndevos> #endmeeting