18:59:17 <jeblair> #startmeeting infra
18:59:17 <openstack> Meeting started Tue Apr 29 18:59:17 2014 UTC and is due to finish in 60 minutes.  The chair is jeblair. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:59:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:59:20 <openstack> The meeting name has been set to 'infra'
18:59:23 <jeblair> Agenda:
18:59:23 <jeblair> #link https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting#Agenda_for_next_meeting
18:59:23 <jeblair> Last meeting:
18:59:23 <jeblair> #link http://eavesdrop.openstack.org/meetings/infra/2014/infra.2014-04-22-19.01.html
18:59:39 <jeblair> #topic Gerrit upgrade
18:59:42 <lifeless> o/ ping if you need me otherwise I'm in -> -alt
19:00:07 <fungi> heyo
19:00:17 <fungi> we're early?
19:00:25 <aburaschi1> hi
19:00:26 <jeblair> i think this is the only thing we really have on the agenda...
19:00:28 <jeblair> fungi: are we?
19:00:33 <sdague> not any more :)
19:00:46 <fungi> heh
19:01:00 <jeblair> so, we upgraded!
19:01:05 <sdague> yay!
19:01:10 <jeblair> i have a few things on the punch list
19:01:15 <aburaschi1> \o/
19:01:18 <jeblair> * something's weird with tag jobs
19:01:30 <anteaya> o/
19:01:48 <jeblair> * turbo hipster's vote looks wrong on this patch: https://review.openstack.org/#/c/86443/
19:02:07 <jeblair> * and of course we need to restart for the cosmetic changes
19:02:16 <jeblair> that's all i have right now; am i missing anything?
19:02:17 <SergeyLukjanov> yeah
19:02:23 <SergeyLukjanov> it was re restart ;)
19:02:34 <anteaya> searching for strings doesn't seem to work
19:02:38 <anteaya> regex's work
19:02:43 <dims> jeblair, and "recheck migrations" did not work on that same review
19:02:53 <fungi> i'm as of yet unconvinced turbo-hipster is operable. it ignored dims's recheck
19:02:57 <fungi> that
19:03:17 <jeblair> yeah, i'm not really concerned about that.  turbo-hipster inc has their own crack team of ops to deal with that...
19:03:19 <sdague> jeblair: I noticed merge collision not -2ing a patch
19:03:21 <clarkb> they probably need a regex update
19:03:30 <anteaya> clarkb: hey
19:03:37 <anteaya> clarkb: you are on holidays
19:03:44 <jeblair> clarkb: heh, you're terrible at vacations
19:03:49 <jeblair> sdague: link?
19:04:00 <sdague> let me go dig it up
19:04:12 <sdague> https://review.openstack.org/#/c/44714/
19:04:14 <jeblair> (re turbo-hipster -- it's the fact that it was -1 then +1 but the approval summary still said -1 that's troubling)
19:04:36 <fungi> agreed
19:04:39 <jeblair> oh good grief, what's the deal with the bird?
19:04:49 <clarkb> not really here >_>
19:04:56 <jeblair> we should get rid of that
19:05:08 <zaro> anteaya: what do you mean about searching for strings?
19:05:21 <sdague> hah, I totally missed the bird when I looked at it this morning
19:05:37 <sdague> I noticed that change because it was in my review list as 'submitted'
19:05:43 <jesusaurus> well, turbo-hipster +1'd patch 21 after -1'ing patch 23, so the -1 is accurate
19:05:51 <anteaya> zaro: if you look at teh backscroll with tjones and myself today, she was looking for message:"spawn refactor - phase 1" to return results
19:05:55 <fungi> jesusaurus: good eye
19:05:56 <anteaya> it didn't
19:06:06 <sdague> "Submitted, Merge Pending" that is
19:06:34 <ianw> re gerrit upgrage : is it a known issue that the regex matching for jenkins results isn't doing the formatting any more?
19:06:34 <jeblair> sdague: oh, that's actually a missing dependency, not a conflict -- that's something we should fix in zuul
19:06:43 <anteaya> zaro: I got her going again with topic:^bp/vmware-spawn.* but taht doesn't solve the string search issue
19:06:44 <sdague> ok
19:06:46 <jeblair> sdague: but i think it's consistent with old behavior
19:06:48 <fungi> jeblair: i think jesusaurus's insight explains the turbo-hipster -1 rather well
19:07:11 <jeblair> fungi, jesusaurus: thanks for catching that; i'm striking it from the punch list
19:07:11 <zaro> anteaya: hmm, i remember that working for me.
19:07:11 <anteaya> ianw: what regex?
19:07:14 <sdague> ok, then it can wait until later
19:07:52 <jeblair> also, wow that change was started a long time ago.  :)
19:07:59 <fungi> ianw: the formatting is what we've got a pending fix merged for, awaiting the next reasonable timme to restart gerrit
19:08:22 <anteaya> fungi: would that explain my outcome with tjones?
19:08:33 <sdague> yep
19:08:36 <ianw> fungi: ok, thanks
19:08:39 <fungi> anteaya: unrelated
19:08:49 <anteaya> hmmmm
19:09:02 <sdague> jeblair: so question on dashboards, where do those go? (i.e. how do I submit them)
19:09:11 <fungi> anteaya: i think ianw was asking about the regular expressions used in commentlink configuration within gerrit, not having to do with search queries
19:09:25 <anteaya> oh
19:10:22 <jeblair> sdague: i think mordred was looking into how to do that.  they are special branches in the repo -- my guess is we might want to have jeepyb do it, or we might want to set up those branches for review.
19:10:28 <ianw> anteaya: yeah, the regex's in https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/manifests/gerrit.pp
19:10:55 <jeblair> sdague: i think in the interim, we can manually push some in when you're ready
19:10:56 <zaro> anteaya: i don't think the spaces work, you need to do "message:spawn message:refactor message:phase .."
19:10:57 <sdague> jeblair: ok. I can probably convert most of these custom queries to dashboards, but having a way to review / test them would be good
19:11:25 <anteaya> zaro: yeah, which is not how I would expect to have to do a string search query
19:11:35 <sdague> jeblair: ok, I'll see about converting the basic no negative feedback one, that will be the most generic one people will want I think
19:11:53 <zaro> anteaya: gerrit doc doesn't say it supports what you want :)
19:11:55 <anteaya> zaro: searching for one word at a time and chaining them with AND worked but was contrary to my expectations
19:12:10 <anteaya> zaro: you used to be able to search for strings like that
19:12:10 <jeblair> sdague: yeah, we certainly want to review them; testing them might be hard without just pushing them in to see what happens.  though the query language being the same as searches helps.
19:12:25 <anteaya> and I can't search for message:- or message:"-"
19:12:36 <anteaya> which was a character in the original string
19:13:14 <zaro> interesting, i guess i never tried to search like that before.
19:13:35 <sdague> jeblair: yeh
19:13:51 <jeblair> any other new-gerrit issues?
19:14:07 <sdague> anteaya: my guess is secondary indexes are doing stop words, so that's dropped
19:14:16 <sdague> it's a lucene index on the secondary, right?
19:14:25 <jeblair> sdague: yes
19:14:39 <fungi> links including target="_blank" by default could be considered a new gerrit issue i suppose, though not sure whether it's one we care enough to solve/investigate
19:14:50 <sdague> yeh, so think less "grep" and more google
19:15:34 <jeblair> fungi: i wonder if there's an option for that
19:16:09 <fungi> i don't see a personal preferences option anyway
19:16:35 <fungi> could be a global setting somewhere, i suppose
19:16:44 <fungi> i'll dig in the config ref a little
19:17:43 <jeblair> we _could_ write a commentlink parser that strips them out
19:18:20 <jeblair> (which, i believe technically we have inadvertently done for the zuul links -- it's just inconsistent now)
19:18:36 <jeblair> (or will be after the restart)
19:18:53 <fungi> true
19:19:10 <fungi> apparently commentlinks are now configurable per-project as well
19:20:09 <sdague> jeblair: I think this is a reasonable inbox zero kind of query -  status:open NOT label:Code-Review>=0,self label:Verified>=1,jenkins NOT label:Code-Review<=-1 NOT label:Workflow<=-1
19:20:58 * sdague <3 queries that can mix label and id
19:21:08 <jeblair> yeah that's nice
19:21:32 <jeblair> sdague: what if you -2 a change?
19:21:58 <fungi> label:Code-Review<=-1 would include -2
19:22:10 <jeblair> derp
19:22:17 <sdague> :)
19:23:27 <sdague> I think my only question is if NOT label:Code-Review>=0,self applies to all patches, or current patch only. If #1, then it probably has to change to >=1
19:23:29 <anteaya> my connection had dropped
19:23:36 <anteaya> what are stop words?
19:23:38 <sdague> if #2 then it's probably ok
19:24:16 <fungi> i believe labels only ever apply to the latest patchset
19:24:22 <jeblair> sdague: i would guess #2 based on my usage...
19:24:22 <sdague> fungi: ok, cool
19:24:38 <sdague> yeh, I'm trying to find an instance where I have an old 0 comment
19:26:45 <jeblair> #topic Upcoming project renames
19:26:53 <jeblair> we have some now
19:27:04 <jeblair> #link https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting#Upcoming_project_renames
19:27:58 <jeblair> i have some things to do friday and will be out next week
19:28:07 <fungi> ahh, i missed blazar was finally okayed. awesome
19:28:10 <jeblair> so i may not be particularly useful in that department for a bit
19:29:00 <jeblair> we could just do it on, say, thursday -- or we can see if the crew that's around next week is interested
19:29:31 <fungi> i could rename things this weekend, but with both you and clarkb travelling that seems unwise. thursday's fine with me though, or roping people into it next week
19:29:52 <fungi> i have no real preference and can do it whenever
19:30:26 <jeblair> fungi: i lean toward deferring it right now, we have quite a bit on our plate while being short-staffed
19:31:03 <jeblair> (i think i'm not eager to make _more_ work for us right now)
19:31:19 <fungi> fair enough
19:31:42 <jeblair> it being the first 2.8 rename, we could have... surprises.
19:31:45 <fungi> i'll bring it up next thursday and see how those who are around feel about viability
19:31:58 <fungi> er, next tuesday (in the meeting)
19:32:05 <jeblair> *nod*
19:32:23 <anteaya> can you create a test repo and rename it prior to next tuesday
19:32:30 <anteaya> would that have any value?
19:32:53 <anteaya> guess not, since you can never delete anything in gerrit
19:32:57 <jeblair> anteaya: on review-dev, sure, if we have time.
19:33:05 <anteaya> oh okay, yeah there
19:33:42 <jeblair> #topic open discussion
19:33:56 <jeblair> anything else?
19:34:01 <ianw> just wanted to chase up on https://review.openstack.org/#/c/86842/
19:35:09 <anteaya> zaro: you were going to change the db schema on review-dev to match the db schema on review, were you not?
19:35:25 <fungi> ianw: i'm cool with merging that later today, after i confirm i have no other gotchas in nodepool
19:35:40 <jeblair> fungi, ianw: ++
19:36:25 <ianw> thanks
19:37:35 <jeblair> well, thanks everyone, and enjoy the rest of the "off week"  :)
19:37:40 <jeblair> #endmeeting