19:02:55 <notmyname> #startmeeting swift
19:02:58 <openstack> Meeting started Wed Apr 16 19:02:55 2014 UTC and is due to finish in 60 minutes.  The chair is notmyname. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:02:59 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:03:02 <openstack> The meeting name has been set to 'swift'
19:03:36 <notmyname> welcome. I'm on conference wifi, so I've got a lot of lag
19:03:42 <notmyname> agenda this week is
19:03:43 <creiht> howdy
19:03:48 <notmyname> #link https://wiki.openstack.org/wiki/Meetings/Swift
19:04:25 <notmyname> #topic status of in-process functional tests
19:04:35 <notmyname> portante this is your topic
19:04:51 <portante> sorry for joining late
19:04:54 <notmyname> portante: you're up :-)
19:04:59 <portante> :)
19:05:11 <portante> so in-process func tests
19:05:20 <notmyname> what's going on? what do you need?
19:05:39 <portante> in-process functional tests have been posted for review
19:06:02 <portante> they enable us to run the full suite of functional tests without using an external swift cluster
19:06:09 <portante> for example: tox -e func
19:06:27 <portante> or in a development environment, ./.functests, will just work
19:06:47 <portante> with this we can generate coverage of core swift code from just the functional tests
19:06:57 <portante> and, if one is so inclined,
19:07:10 <peluse> portante:  dumb question but how does this differ from running ./.functests in my SAIO?
19:07:11 <portante> write new functional tests easier, since debugging can be done much easier
19:07:31 <portante> in your SAIO, you use: startmain; ./.functests
19:07:43 <cschwede> peluse: you don’t need a SAIO
19:07:47 <portante> the startmain starts up all the daemon processes seaprately
19:07:56 <peluse> ahhhh...
19:08:01 <portante> here, with in-process, there are no daemon processes
19:08:06 <peluse> OK, I'll take it for a spin
19:08:07 <creiht> portante: were you able to hook up the coverage report stuff with the inprocess func tests?
19:08:09 <portante> it is all run within the one process
19:08:16 <portante> yes, it just works
19:08:31 <notmyname> portante: how so? we get a coverage report at the end?
19:08:39 <portante> I believe so
19:08:56 <portante> but that may be from tox, I might have to check .functests to see
19:09:02 <creiht> I think that is one of the biggest benefits of having the in-process functests
19:09:08 <portante> agreed
19:09:37 <portante> to me, the motivation was to step through functional tests to review code to see if it was doing what it said
19:10:17 <portante> what I am hoping we can do with this is make sure the storage policies is thoroughly tested
19:10:37 <portante> I'd like to think we can use the coverage reports to help catch what needs shoring up
19:10:37 <cschwede> i really like the approach, because it speeds up this whole „python … install ; swift-init restart ; .functests“ a lot
19:10:48 <portante> yes, that is another benefit
19:11:03 <portante> there would no longer be any execuse for a dev not to first run func tests
19:11:22 * peluse wasn't aware there was an excuse today
19:11:27 <portante> in theory, one could do this with probe tests as well, but that will be another effort
19:11:30 <notmyname> portante: what do you need from the rest of us?
19:11:37 <portante> review and testing
19:12:01 <portante> during this development, it was clear that there are *many* ways folks use the functional tests and in different setups
19:12:13 <portante> so need to make sure folks are comfortable with the behaviors
19:12:30 <portante> there is an environment variable to explicitly turn them on/off
19:12:51 <portante> but if you don't have a /etc/swift/test.conf, it will now run them in-process instead of skipping them
19:13:12 <portante> additionally, there are two patches left to get in for this
19:13:18 <portante> https://review.openstack.org/#/c/86710/
19:13:34 <portante> which is a base dependency to shore up constraint usage in the functional tests
19:13:40 <portante> https://review.openstack.org/#/c/66108/
19:13:53 <portante> which is the actual in-process patch set
19:14:44 <portante> there have been a number of very useful go-rounds of these pathes, and want to make sure the various sub-groups of devs (rax, eno, hp, ss) get eyes on them
19:15:03 <portante> once this lands, I'd like to see them as part of the check/gate jobs as well
19:15:35 <portante> any feedback, questions, concerns, or throw-portante-off-a-bus things that need to be said?
19:15:51 <portante> bueller, bueller
19:16:02 <creiht> lol
19:16:18 <peluse> I'll bang on it with SP and Yuanz's new SP functional tests
19:16:18 <notmyname> they are today, aren't they?
19:16:22 <clayg> there's patches, they need reviews - got it
19:16:31 <notmyname> clayg: do you have any concerns on it for this meeting?
19:16:33 <torgomatic> did folks ever figure out what they wanted as far as constraint precedence? there was a lot of back-and-forth on that
19:16:55 <portante> I believe that has settled out in the first patch above, 86710
19:16:59 <torgomatic> ok, thanks
19:17:41 <clayg> notmyname: i don't think so, i'll checkup on how the test.conf stuff shook out
19:18:26 <portante> ss, rh and hp looked at things so far, but would love reps from the other orgs to consider it as well
19:18:43 <notmyname> portante: anything else on that topic?
19:18:51 <portante> nothing on my end
19:18:55 <notmyname> ok, thanks
19:18:58 <creiht> portante: yeah it is on my list
19:18:59 <notmyname> #topic final Icehouse (Swift 1.13.1) check
19:19:19 <notmyname> anything that has come up that warrants a backport to 1.13.1? this is the final call
19:19:38 <creiht> notmyname: not that I am aware of
19:19:50 <notmyname> creiht: thanks
19:19:53 <notmyname> clayg: portante: peluse: anythign from your end?
19:20:28 <portante> nothing here
19:20:33 <clayg> yawn
19:20:38 <notmyname> :-)
19:21:37 <clayg> notmyname: there's a few open bugs that are like weird, account-reaper, container-updater, object tombstones, I don't think any of them are regressions - but like... idk there out there
19:24:16 <clayg> yeah we need to do some bug triage
19:24:28 <clayg> but I don't know of anything that seems to have broken as of late
19:25:08 <notmyname> ok, I'll pass on to ttx that 1.13.1-rc2 is good to go for icehouse
19:25:08 <notmyname> thanks
19:25:08 <notmyname> #topic Oslo liason
19:25:08 <notmyname> the oslo project is asking each project to nominate a core dev to be a liason to oslo
19:25:08 <notmyname> #link https://wiki.openstack.org/wiki/Oslo/ProjectLiaisons
19:25:10 <notmyname> see that link
19:25:10 <notmyname> and I'll be asking for a volunteer :-)
19:25:11 <notmyname> #topic storage policy status update
19:25:12 <notmyname> clayg: peluse: how's it going?
19:25:21 <ttx> notmyname: received
19:25:22 <peluse_> working on docs mainly but doing so I discovered 3 gaps in current policy work on feature/ec
19:25:24 <creiht> lol
19:25:31 <peluse_> 2 of them are here: SP support for recon middleware and container sync are ready for eyes at https://review.openstack.org/#/c/87387/ https://review.openstack.org/#/c/86469/
19:25:33 <creiht> probably shouldn't be me :)
19:25:44 <peluse_> third is SP support for acct reaper which I'll tackle next
19:25:45 <clayg> creiht: should *totally* be you!
19:25:51 <creiht> hah
19:26:13 <creiht> notmyname: I can do it if you need
19:26:21 <portante> I second creiht nomination of himself
19:26:27 <creiht> I'll try to be nice :)
19:27:08 <peluse_> thanks everyone who has been reviewing the doc patch BTW, appreciate all the edits/feedback
19:27:09 <clayg> peluse_: notmyname: making progress on the reconciler, moved on to some more stuff in the container-replicator (where there's more cleanup and tests falling out) - I'm starting converge on a hybridization of torgomatic approach and my original thoughts on the subject that is starting feel pretty good
19:27:28 <torgomatic> clayg: i'll definitely be interested to see that
19:28:09 <clayg> last night I started everything up and when all chaos monkey on my saio putting objects and containers all over the place, and the damn thing is to a point now where I'm gunna have think pretty hard to come up with something that will break it
19:28:23 <peluse_> nice!
19:28:37 <clayg> it's starting to feel "like swift" - turn one thing of and something else picks up the slack - break it over here - and when it starts back up it puts thing back in order
19:29:10 <portante> cool, perhaps encode that in probe tests / functional tests?
19:29:17 <clayg> portante: yeah I have a few
19:32:20 <portante> clayg: great, I am out on vacation next week, but when I come back I can help review and/or write more probe tests if you have ideas for some
19:32:48 <notmyname> Moving to cell phone irc client
19:32:55 <clayg> notmyname: lol
19:32:56 <peluse_> wow
19:33:16 <peluse_> portante:  not related to SP, but can you take another quick look at https://review.openstack.org/#/c/87101/ before you head out for R&R?
19:33:25 <portante> yes
19:33:31 <peluse_> gratzi
19:33:39 <notmyname> I have no idea what's been said in here :)
19:34:00 <portante> not much
19:34:03 <chmouel> R&R?
19:34:06 <peluse_> is it Fri yet?
19:34:09 <portante> rest and relaxation
19:34:26 <notmyname> Anything said about when policy patches might be proposed to master?
19:34:27 <peluse_> US military term, sorry
19:34:49 <clayg> notmyname: peluse_ and I already both already have some up?
19:35:03 <notmyname> Ok :)
19:35:26 <peluse_> notmyname:  as I understand it we still want to go with the 'complete' picture staged approach right?  So we need the couple I mentioned before reviewd, clayg's stuff and the acct repaer before we can propose anything SP specifuc
19:35:32 <clayg> or are you asking when the reconciler and the reaper and docs and all the other stuff is going to be done?
19:35:41 <notmyname> Yes that :)
19:36:06 <notmyname> What peluse mentioned is good
19:36:33 <peluse_> I don't think repaer will take long, docs are coming along nicely I think so when clayg is ready I'll be ready :)
19:37:33 * clayg has total faith in peluse_
19:37:56 <peluse_> ha!
19:37:58 <notmyname> What else needs to be brought up on storage policies in this meeting?
19:38:24 <peluse_> I think we need to rebase feature/ec
19:38:36 <clayg> rebase - or just merge master down again?
19:38:41 <peluse_> ya
19:38:45 <notmyname> Merge master in?
19:38:50 <clayg> it was an either or question :P
19:38:51 <notmyname> Ah ok
19:38:59 <peluse_> whatever we've been doing :)
19:39:12 <notmyname> That thing. Do more of it
19:39:12 <clayg> nice - yeah never hurts to get those upstream changes on feature/ec
19:39:37 <peluse_> will save ys pain later if we keep doing it weekly
19:39:50 <notmyname> Yes
19:40:25 <notmyname> I can work on that
19:40:29 <peluse_> cool
19:40:51 <notmyname> Anything else?
19:41:06 <peluse_> dont think so
19:41:21 <notmyname> #topic open discussion
19:41:38 <peluse_> well, one more quick plug for docs.  Please read if you can:  https://review.openstack.org/#/c/85824/
19:42:01 <notmyname> Yes. The docs patch is an essential starting point
19:42:42 <notmyname> What else would you like to discuss in the meeting right now?
19:42:43 <cschwede> peluse_: actual, it’s a great way to get into SP! nice work
19:43:16 <peluse_> thanks!
19:44:23 <notmyname> If there is nothing else, let's adjourn this meeting
19:44:45 <notmyname> Thanks for coming
19:44:47 <peluse_> rock on man... thanks
19:44:56 <notmyname> #endmeeting