| notmyname | what we do for cloud files at rax is deploy the code separate from running the code. ie apt-get install doesn't start any services. so install it and then start it | 00:00 |
|---|---|---|
| justinsb_ | notmyname: I think what you do at cloud files is in a totally different league to what we're talking about with a typical Ubuntu user's installation :-) | 00:01 |
| notmyname | heh. I'll buy that :-) | 00:01 |
| justinsb_ | notmyname: If they install with replicas=1, are they "stuck" at the moment if they want to go to e.g. replicas="2". They'd have to dump all their files out and reload them? | 00:02 |
| notmyname | there is no way to change the replica count right now. if you change it, you have to migrate the data so that it gets rehashed into the right location | 00:03 |
| notmyname | same with the partition count (set by the ring's partition power) | 00:04 |
| justinsb_ | notmyname: Does changing the replica count cause a rehash? I thought it just added another location? If it requires a rehash, my testing isn't likely to work! | 00:05 |
| notmyname | no, not directly. but since there is no way to add a replica now, it would mean building a new ring. and since volumes have a random component to their placement, things would need to get rehashed | 00:06 |
| notmyname | but if you are adding a replica (or removing) to the existing ring data structure, that could still work without rehashing all the data | 00:06 |
| justinsb_ | notmyname: OK! I think maybe I'll test it out, and see if it works (this may be academic anyway). It may be that we can keep this as an 'experimental, dangerous' program like those programs that resized your disk partitions back in the day | 00:07 |
| *** Mandell has quit IRC | 00:08 | |
| justinsb_ | notmyname: I've proposed a discussion of packaging issues for the design summit (like this one). Maybe we can tack this on or have a separate discussion just about growing swift from first install to happy cluster user | 00:08 |
| justinsb_ | notmyname: I'm pretty new to swift, so I'm not sure if there are other issues that people encounter along the way | 00:08 |
| justinsb_ | notmyname: And maybe we can have either replica count increment or as unique as possible in Folsom | 00:09 |
| notmyname | justinsb_: ya, I'd love to have a talk of 2 at the summit on it. please submit it :-) | 00:12 |
| justinsb_ | notmyname: "Of 2" = "Growing swift from first install to happy cluster" ? | 00:13 |
| notmyname | oops | 00:14 |
| *** btorch_ has joined #openstack-dev | 00:14 | |
| notmyname | s/a talk of 2/a talk or 2/ | 00:14 |
| justinsb_ | notmyname: OK! I'll submit a discussion topic along those lines and if you want to add to it feel free! | 00:15 |
| notmyname | thanks | 00:15 |
| *** thingee_zzz has quit IRC | 00:20 | |
| *** btorch_ has quit IRC | 00:24 | |
| *** ncode has joined #openstack-dev | 00:31 | |
| *** dolphm has joined #openstack-dev | 00:33 | |
| justinsb_ | notmyname: Just tried it using devstack and it does seem to work! Cool! | 00:44 |
| *** dolphm has quit IRC | 00:49 | |
| *** dolphm has joined #openstack-dev | 00:54 | |
| justinsb_ | notmyname: I think I'll try a real test with two machines, and then shut down the first one. But it's a credit to swift that something this evil does seem to work! | 00:54 |
| *** flaviamissi has quit IRC | 01:01 | |
| *** flaviamissi has joined #openstack-dev | 01:01 | |
| *** flaviamissi has quit IRC | 01:06 | |
| *** dprince has joined #openstack-dev | 01:19 | |
| *** rkukura has joined #openstack-dev | 01:21 | |
| *** dprince has quit IRC | 01:25 | |
| *** dolphm has quit IRC | 01:58 | |
| *** dolphm_ has joined #openstack-dev | 02:00 | |
| *** dtroyer_zzz is now known as dtroyer | 02:14 | |
| *** ncode has quit IRC | 02:48 | |
| *** jakedahn_zz is now known as jakedahn | 03:06 | |
| *** dolphm has joined #openstack-dev | 03:27 | |
| *** dolphm_ has quit IRC | 03:30 | |
| *** deshantm has quit IRC | 03:34 | |
| *** deshantm has joined #openstack-dev | 03:41 | |
| *** ywu has quit IRC | 03:51 | |
| *** dtroyer is now known as dtroyer_zzz | 03:53 | |
| *** Mandell has joined #openstack-dev | 03:53 | |
| *** mikal has quit IRC | 03:59 | |
| *** dolphm has quit IRC | 04:19 | |
| *** mdomsch has joined #openstack-dev | 04:31 | |
| *** anotherjesse is now known as anotherjesse_zz | 04:40 | |
| *** ghanima has joined #openstack-dev | 04:45 | |
| *** anotherjesse_zz is now known as anotherjesse | 04:46 | |
| *** Mandell has quit IRC | 04:50 | |
| *** ghanima has left #openstack-dev | 04:52 | |
| *** dolphm has joined #openstack-dev | 04:52 | |
| *** dolphm has quit IRC | 04:52 | |
| *** Mandell has joined #openstack-dev | 04:55 | |
| *** hattwick has quit IRC | 05:02 | |
| *** mikal has joined #openstack-dev | 05:16 | |
| *** mdomsch has quit IRC | 05:19 | |
| *** Mandell has quit IRC | 05:19 | |
| *** sniperd has quit IRC | 05:22 | |
| *** Mandell has joined #openstack-dev | 05:28 | |
| *** Mandell has quit IRC | 05:36 | |
| *** deshantm has quit IRC | 05:40 | |
| *** Mandell has joined #openstack-dev | 05:51 | |
| *** neodosh has joined #openstack-dev | 06:04 | |
| *** jakedahn is now known as jakedahn_zz | 06:12 | |
| *** dachary has quit IRC | 06:15 | |
| *** halfss has joined #openstack-dev | 06:22 | |
| vishy | tr3buchet: we have a potentially very nasty performance issue in the dhcp code in linux_net. We are doing 2 queries for every running instance. I think we may have to add in a nasty join to get reasoable performance out of it | 06:22 |
| vishy | tr3buchet: I know we want to keep the stuff separate, but doing 2000 queries for every instance launch or terminate if there are 1000 instances running is not acceptable | 06:23 |
| *** anotherjesse is now known as anotherjesse_zz | 06:34 | |
| *** jakedahn_zz is now known as jakedahn | 06:45 | |
| *** journeeman has joined #openstack-dev | 06:53 | |
| journeeman | Hi all. Didn't Openstack make it to GSoC 2012? | 06:53 |
| journeeman | I was really looking forward to participate :/ | 06:55 |
| journeeman | Well, the Ideas page is still there so I'll continue looking into the ideas to work on this summer. | 06:57 |
| *** Mandell has quit IRC | 07:10 | |
| *** jakedahn is now known as jakedahn_zz | 07:14 | |
| *** halfss has quit IRC | 07:29 | |
| *** apevec has joined #openstack-dev | 07:44 | |
| *** danwent has quit IRC | 07:48 | |
| *** hattwick has joined #openstack-dev | 07:59 | |
| *** popux has joined #openstack-dev | 08:10 | |
| *** neodosh has quit IRC | 08:18 | |
| *** anotherjesse_zz is now known as anotherjesse | 08:25 | |
| *** popux has quit IRC | 08:30 | |
| *** anotherjesse is now known as anotherjesse_zz | 08:33 | |
| *** dneary has joined #openstack-dev | 08:38 | |
| *** dneary has quit IRC | 08:38 | |
| *** dneary has joined #openstack-dev | 08:38 | |
| *** rbasak has joined #openstack-dev | 08:39 | |
| *** halfss has joined #openstack-dev | 08:40 | |
| *** garyk has joined #openstack-dev | 09:05 | |
| *** Ryan_Lane has quit IRC | 09:07 | |
| *** koolhead11 has joined #openstack-dev | 09:08 | |
| *** Gordonz has joined #openstack-dev | 09:10 | |
| *** Gordonz has quit IRC | 09:10 | |
| *** Gordonz has joined #openstack-dev | 09:10 | |
| *** Gordonz has quit IRC | 09:23 | |
| *** LinuxJedi has joined #openstack-dev | 09:34 | |
| *** reidrac has joined #openstack-dev | 09:58 | |
| *** zigo has joined #openstack-dev | 10:02 | |
| *** oneiroi has joined #openstack-dev | 10:06 | |
| *** shevek_ has quit IRC | 10:26 | |
| *** kyriakos has left #openstack-dev | 10:31 | |
| *** hattwick has quit IRC | 10:32 | |
| *** paulormg has joined #openstack-dev | 10:46 | |
| *** rkukura has quit IRC | 10:58 | |
| *** milner has joined #openstack-dev | 11:10 | |
| *** crobinso has joined #openstack-dev | 11:17 | |
| *** dachary has joined #openstack-dev | 11:20 | |
| *** zigo has quit IRC | 11:21 | |
| *** zigo has joined #openstack-dev | 11:21 | |
| *** zul has joined #openstack-dev | 11:24 | |
| *** dneary has quit IRC | 11:26 | |
| *** zigo has quit IRC | 11:26 | |
| *** Shrews has joined #openstack-dev | 11:28 | |
| *** bsza has joined #openstack-dev | 11:38 | |
| *** journeeman has quit IRC | 11:39 | |
| *** vizsla has joined #openstack-dev | 11:46 | |
| *** vizsla has left #openstack-dev | 11:47 | |
| *** vizsla has joined #openstack-dev | 11:47 | |
| *** vizsla has quit IRC | 11:57 | |
| *** vizsla has joined #openstack-dev | 11:57 | |
| *** asalkeld has quit IRC | 12:14 | |
| *** oneiroi has quit IRC | 12:20 | |
| *** oneiroi has joined #openstack-dev | 12:20 | |
| *** bsza has quit IRC | 12:24 | |
| *** maploin has joined #openstack-dev | 12:38 | |
| *** maploin has quit IRC | 12:38 | |
| *** maploin has joined #openstack-dev | 12:38 | |
| *** andrewsmedina has quit IRC | 12:43 | |
| *** refactor has joined #openstack-dev | 13:01 | |
| *** sandywalsh has quit IRC | 13:01 | |
| *** berendt has joined #openstack-dev | 13:03 | |
| *** dprince has joined #openstack-dev | 13:06 | |
| *** mattray has joined #openstack-dev | 13:08 | |
| *** davidkranz has joined #openstack-dev | 13:08 | |
| *** andrewsmedina has joined #openstack-dev | 13:14 | |
| *** mattray has quit IRC | 13:14 | |
| *** sandywalsh has joined #openstack-dev | 13:15 | |
| *** mattray has joined #openstack-dev | 13:16 | |
| *** kbringard has joined #openstack-dev | 13:18 | |
| *** zigo has joined #openstack-dev | 13:23 | |
| *** blamar has joined #openstack-dev | 13:25 | |
| *** deshantm has joined #openstack-dev | 13:25 | |
| *** berendt has quit IRC | 13:28 | |
| *** kbringard has quit IRC | 13:29 | |
| *** kbringard has joined #openstack-dev | 13:30 | |
| *** kbringard has quit IRC | 13:31 | |
| *** rkukura has joined #openstack-dev | 13:40 | |
| *** kbringard has joined #openstack-dev | 13:40 | |
| jaypipes | blamar: mornin. | 13:40 |
| blamar | jaypipes: greetings. | 13:41 |
| jaypipes | blamar: I'm good with your KeyError patch. just get eglynn__ or bcwaldon for another +1\ | 13:41 |
| jaypipes | blamar: and thanks! | 13:41 |
| blamar | jaypipes: many many thanks to you sir | 13:41 |
| blamar | will do | 13:41 |
| *** dneary has joined #openstack-dev | 13:43 | |
| *** dneary has quit IRC | 13:43 | |
| *** dneary has joined #openstack-dev | 13:43 | |
| *** gpernot has joined #openstack-dev | 13:47 | |
| jaypipes | dprince: around? | 13:47 |
| kbringard | hmmm, what is the status of the nova2ools stuff? | 13:49 |
| *** mattray has quit IRC | 14:01 | |
| *** mattray has joined #openstack-dev | 14:02 | |
| *** berendt has joined #openstack-dev | 14:02 | |
| *** hattwick has joined #openstack-dev | 14:02 | |
| *** berendt has quit IRC | 14:02 | |
| *** AlanClark has joined #openstack-dev | 14:02 | |
| *** blamar has quit IRC | 14:06 | |
| *** ayoung is now known as ayoung-afk | 14:07 | |
| *** deshantm_ has joined #openstack-dev | 14:10 | |
| *** sdake has quit IRC | 14:10 | |
| *** deshantm has quit IRC | 14:11 | |
| *** deshantm_ is now known as deshantm | 14:11 | |
| *** dolphm has joined #openstack-dev | 14:12 | |
| *** sdake has joined #openstack-dev | 14:12 | |
| *** ayoung-afk is now known as ayoung | 14:22 | |
| *** zzed has joined #openstack-dev | 14:23 | |
| *** mdomsch has joined #openstack-dev | 14:23 | |
| *** lts has joined #openstack-dev | 14:25 | |
| *** dolphm has quit IRC | 14:31 | |
| *** tryggvil_ has joined #openstack-dev | 14:33 | |
| *** andrewbogott has joined #openstack-dev | 14:34 | |
| *** andrewbogott has joined #openstack-dev | 14:34 | |
| eglynn__ | jeblair: ping | 14:34 |
| andrewbogott | I'm having a rotten time trying to get essex installed via apt. Does essex pretty much require a Precise beta in order to install? (On oneiric I see quite a few unmet and un-meetable dependencies.) | 14:35 |
| davidkranz | Is cloud-publish-tarball supposed to work in essex with ubuntu-11.10-server-cloudimg-amd64.tar.gz? I am doing the same as I did with oneiric/diablo (no keystone) but now get trace back in compute log about DecryptionFailure. | 14:37 |
| davidkranz | Other euca commands work. | 14:37 |
| davidkranz | I deployed essex with precise packages. | 14:38 |
| davidkranz | Using credentials from novarc generated by nova-manage | 14:41 |
| *** dtroyer_zzz is now known as dtroyer | 14:44 | |
| *** rnirmal has joined #openstack-dev | 14:45 | |
| ttx | andrewbogott: do you have an example of those unmet/unmeetable deps in Oneiric ? | 14:55 |
| andrewbogott | ttx: I had to melt down my test machine, but I should be back to that point shortly. | 14:56 |
| andrewbogott | ttx: You would expect it to work on oneiric? | 14:57 |
| ttx | andrewbogott: my understanding is that it's mostly usable under Oneiric, as evidenced by Devstack | 14:57 |
| andrewbogott | doesn't devstack install essex as source though, rather than using apt? | 14:57 |
| andrewbogott | devstack works great for me; it's the dependency-checking in 'apt-get install glance' that's failing. | 14:58 |
| bcwaldon | andrewbogott: (yes, it uses source) | 14:58 |
| ttx | andrewbogott: there may have been some runtime-deps that crept in | 14:58 |
| ttx | andrewbogott: hence my question :) | 14:58 |
| andrewbogott | ok -- lemme catch up and I'll let you know when I hit a wall. | 14:58 |
| ttx | andrewbogott: what packages are you using ? | 14:59 |
| jeblair | eglynn__: pong | 14:59 |
| jeblair | jeblair: could you re-enable GLANCE_TEST_S3_CONF & GLANCE_TEST_SWIFT_CONF for the glance builds in Jenkins? | 14:59 |
| jeblair | eglynn__: ^ i'll do that now | 14:59 |
| andrewbogott | ttx: nova-core/trunk | 15:01 |
| jeblair | eglynn__: done | 15:02 |
| ttx | andrewbogott: <jeditrick>that's not the PPA you're looking for.</jeditrick> | 15:03 |
| andrewbogott | are there stable essex packages someplace? | 15:03 |
| ttx | andrewbogott: try http://wiki.openstack.org/PPAs | 15:03 |
| ttx | "stable essex" haha. | 15:03 |
| *** tryggvil_ has quit IRC | 15:03 | |
| *** tryggvil__ has joined #openstack-dev | 15:04 | |
| ttx | I need to find a nova-core volunteer to mark the nova-core ppas deprecated | 15:04 |
| andrewbogott | Oh, so nova-core/trunk isn't just bleeding edge, it's completely unmaintained? | 15:04 |
| ttx | andrewbogott: it's not refreshed anymore | 15:04 |
| andrewbogott | ok! Well, that's encouraging then :) | 15:04 |
| ttx | automatic packages build go to the other (common) one | 15:05 |
| ttx | andrewbogott: sent an email about it but you may have missed it | 15:05 |
| mtaylor | ttx: send me an email to remind me and I'll takecare of it | 15:05 |
| ttx | mtaylor: was about to ask you to do the PROJECT-core PPA tour :) | 15:05 |
| ttx | mtaylor: willdo | 15:05 |
| andrewbogott | ttx: I believe that this is the page that misled me (or, one of them): http://docs.openstack.org/bexar/openstack-compute/admin/content/ch03s02.html | 15:06 |
| andrewbogott | Oh, I guess 'bexar' should've tipped me off :( | 15:06 |
| eglynn__ | jeblair: thanks! | 15:06 |
| ttx | andrewbogott: it used to be all the rage | 15:06 |
| mtaylor | ttx: perhaps we should remove bexar and cactus info? | 15:07 |
| mtaylor | annegentle: ^^^ | 15:07 |
| *** vincentricci has joined #openstack-dev | 15:07 | |
| andrewbogott | Hm... actually, I was also looking at this page, which is outside our sphere of influence... http://www.zeroaccess.org/2011/03/setting-up-openstack-compute/ | 15:10 |
| andrewbogott | Basically my error was searching for 'package' and rather than 'ppa'. It's only when I google for 'ppa' that http://wiki.openstack.org/PPAs shows up. | 15:11 |
| *** rnirmal has quit IRC | 15:11 | |
| annegentle | andrewbogott: yeah I think i'm going to remove /bexar and /cactus from the docs site and the Google CSE this week | 15:11 |
| annegentle | andrewbogott: blogs I can't do much about right :) | 15:11 |
| *** rnirmal has joined #openstack-dev | 15:11 | |
| *** devananda has joined #openstack-dev | 15:12 | |
| andrewbogott | annegentle: Maybe add some keywords to that wiki page like 'debian package' and 'apt' and 'ubuntu' &c. | 15:12 |
| andrewbogott | Not that any of these changes will solve the problem of users not thinking ( <- the actual source of my problem) | 15:12 |
| *** markwash has joined #openstack-dev | 15:13 | |
| bcwaldon | jaypipes: https://review.openstack.org/#change,5509 | 15:17 |
| bcwaldon | chmouel: https://review.openstack.org/#change,5509 | 15:17 |
| jaypipes | jeblair: looks like the GLANCE_TEST_ env variables are still not unset on the py2.7 tox builder: https://jenkins.openstack.org/job/gate-glance-python27/144/console | 15:17 |
| ttx | andrewbogott: feel free to SEO that wikipage with anything necessary :) | 15:18 |
| bcwaldon | jaypipes: I just ran devstack locally with that change and everything went smoothly. We need to get this review in for my other glance change to land | 15:18 |
| jaypipes | bcwaldon: ++ | 15:18 |
| *** dolphm has joined #openstack-dev | 15:20 | |
| jeblair | jaypipes: i just set the variables at eglynn__'s request. | 15:21 |
| jaypipes | eglynn__: just now? | 15:22 |
| jeblair | jaypipes: i completed setting them at 2 minutes past the hour. | 15:22 |
| *** phorgh has joined #openstack-dev | 15:23 | |
| jaypipes | jeblair: k. I fear we need to unset them again... eglynn__, got a sec? | 15:24 |
| jeblair | jaypipes: ok. are you ready for me to do that now? | 15:24 |
| eglynn__ | jaypipes: yep | 15:24 |
| jaypipes | jeblair: yes please. we need to push through some keystone-related critical fixes. | 15:25 |
| jeblair | jaypipes, eglynn__: unset | 15:25 |
| jaypipes | eglynn__: want to chat about latest status with this bug about S3 and 400 vs 200 return? | 15:25 |
| jaypipes | jeblair: ty | 15:25 |
| eglynn__ | jaypipes: reason I wanted those variables set was to preproduce the S3 issue with the new log capture | 15:25 |
| jaypipes | eglynn__: k. did you see the output? | 15:26 |
| eglynn__ | jaypipes: 400 versus 200 status in what context? | 15:26 |
| jaypipes | eglynn__: https://jenkins.openstack.org/job/gate-glance-python27/144/console | 15:27 |
| *** bsza has joined #openstack-dev | 15:27 | |
| jaypipes | sorry, 404 != 200 | 15:27 |
| jaypipes | eglynn__: I don't see any further output in the log capture there, do you? | 15:27 |
| eglynn__ | jaypipes: yep, you're right | 15:28 |
| * eglynn__ confused ... looking into it | 15:28 | |
| jeblair | jaypipes, eglynn__: i set up two new jobs for you: | 15:29 |
| jeblair | https://jenkins.openstack.org/job/dev-gate-glance-python26-swifts3/ | 15:29 |
| jeblair | https://jenkins.openstack.org/job/dev-gate-glance-python27-swifts3/ | 15:29 |
| jaypipes | jeblair: ah, thank you jim | 15:30 |
| eglynn__ | jeblair: thank you! | 15:30 |
| jeblair | they are just like the normal unit test gate jobs, but have the variables set, and they run in silent mode | 15:30 |
| jeblair | so they'll be trigged by all the same changes, but won't affect jenkins vote | 15:30 |
| jeblair | you'll have to visit the jenkins page for the jobs to see them (they won't show up in gerrit at all) | 15:30 |
| eglynn__ | jeblair: excellent | 15:31 |
| jeblair | good luck! | 15:31 |
| eglynn__ | (now to figure out why the log capture isn't capturing ...) | 15:31 |
| *** danwent has joined #openstack-dev | 15:31 | |
| bcwaldon | can a keystoner look at the 'configuringservices.rst' docs today? They seem to be out of date (wrong configuration values, preferring admin_token over full creds, etc) | 15:34 |
| bcwaldon | dolphm, heckj: ^ | 15:34 |
| *** troytoman-away is now known as troytoman | 15:37 | |
| *** rods has joined #openstack-dev | 15:38 | |
| dolphm | bcwaldon: i'll check it out | 15:39 |
| bcwaldon | dolphm: thanks! | 15:39 |
| *** heckj has joined #openstack-dev | 15:42 | |
| *** koolhead11 has quit IRC | 15:42 | |
| *** littleidea has joined #openstack-dev | 15:43 | |
| *** sandywalsh has quit IRC | 15:47 | |
| *** maploin has quit IRC | 15:48 | |
| *** gpernot has quit IRC | 15:50 | |
| *** vizsla has quit IRC | 15:50 | |
| *** littleidea has quit IRC | 15:50 | |
| *** gpernot has joined #openstack-dev | 15:50 | |
| bcwaldon | jaypipes: this might conflict with your tenant-name -> id work: https://review.openstack.org/#change,5517 | 15:53 |
| *** anotherjesse_zz is now known as anotherjesse | 15:53 | |
| *** littleidea has joined #openstack-dev | 15:55 | |
| jaypipes | bcwaldon: actually, I just abandoned the original patchset I made for the tenant name bug... since it modded glance_auth_token that is now removed. | 15:55 |
| bcwaldon | jaypipes: ok, I wasnt sure if you had started on anything locally, that's what I was warning you about :) | 15:56 |
| jaypipes | bcwaldon: you want to take the tenant name bug? I think the thing that needs to be done is just a migration script.. | 15:56 |
| * annegentle laughs at keystoner | 15:56 | |
| bcwaldon | jaypipes: yeah, I can do it | 15:56 |
| dolphm | i want a tshirt | 15:56 |
| dolphm | just sayin | 15:56 |
| bcwaldon | annegentle: I'm surprised we havent made that connection yet | 15:56 |
| jaypipes | bcwaldon: cool, thx man. I'm currently swamped working on https://review.openstack.org/#change,5294 | 15:56 |
| bcwaldon | jaypipes: I'll gladly help you today | 15:57 |
| bcwaldon | jaypipes: for a hamburger tomorrow | 15:57 |
| bcwaldon | or something like that | 15:57 |
| *** thingee has joined #openstack-dev | 15:59 | |
| *** pixelbeat has joined #openstack-dev | 15:59 | |
| *** sandywalsh has joined #openstack-dev | 16:01 | |
| jaypipes | bcwaldon: thx Wimpy :) | 16:03 |
| bcwaldon | that's his name! I couldnt remember what it was | 16:03 |
| bcwaldon | before my time | 16:03 |
| jaypipes | :) | 16:03 |
| heckj | jaypipes: just settling in to do some reviews - are you expecting to do more work on https://review.openstack.org/#change,5294, or is it ready for some eyeballs? | 16:06 |
| *** zaitcev has joined #openstack-dev | 16:08 | |
| jaypipes | heckj: just now pushed the final patchset | 16:10 |
| heckj | sweet - I'll catch it in my run this morning! | 16:10 |
| jaypipes | cheers | 16:12 |
| annegentle | Anyone who wants to review the essex version of the install guide, have at it at https://review.openstack.org/#change,5190 | 16:14 |
| *** danwent has quit IRC | 16:14 | |
| *** bengrue has quit IRC | 16:15 | |
| *** sandywalsh has quit IRC | 16:18 | |
| *** bengrue has joined #openstack-dev | 16:24 | |
| chmouel | dolphm, heckj: https://review.openstack.org/#change,5511 | 16:30 |
| * heckj lookin | 16:30 | |
| *** openstackjenkins has quit IRC | 16:30 | |
| *** openstackjenkins has joined #openstack-dev | 16:31 | |
| *** sandywalsh has joined #openstack-dev | 16:33 | |
| eglynn__ | jaypipes / jeblair: gah! the reason that extra glance log capture didn't appear is that Jenkins builds is driven by tox which calls the nosetests cmd directly instead of glance/run_tests.sh | 16:34 |
| eglynn__ | (the glance logcapture plugin is loaded in run_tests.py) | 16:34 |
| eglynn__ | OTOH running nosetests --glance-logcapture would require the plugin to be installed setuptools stylee instead of being picked up dynamically as it is now | 16:34 |
| *** zul has quit IRC | 16:34 | |
| eglynn__ | jeblair: would it be poss for tox to just call glance/run_tests.sh? | 16:35 |
| *** spiffxp has joined #openstack-dev | 16:38 | |
| *** sandywalsh has quit IRC | 16:38 | |
| *** dubsquared has joined #openstack-dev | 16:39 | |
| jeblair | eglynn__: i don't want to step on mtaylor's toes here, he's leading the tox stuff. let's see if we can get his opinion. | 16:40 |
| jeblair | mtaylor: ^ | 16:40 |
| eglynn__ | jeblair: cool | 16:40 |
| *** sniperd has joined #openstack-dev | 16:40 | |
| *** Gordonz has joined #openstack-dev | 16:42 | |
| heckj | bcwaldon: when you have a moment, would you mind giving https://review.openstack.org/#change,5350 a look | 16:43 |
| *** paulormg has quit IRC | 16:45 | |
| *** jdurgin has joined #openstack-dev | 16:45 | |
| bcwaldon | heckj: yep, going to look over all your docs changes this morning | 16:46 |
| heckj | thxusir! | 16:46 |
| *** PotHix has joined #openstack-dev | 16:47 | |
| *** novas0x2a|laptop has quit IRC | 16:48 | |
| jaypipes | jeblair: just pushed a proposed change to openstack-ci/devstack-gate. Really just looking for some initial feedback... bash scripting ain't my specialty ;) | 16:50 |
| jaypipes | jeblair: changes are attempting to get to the point where we can run tempest against the devstack-vm | 16:50 |
| *** sandywalsh has joined #openstack-dev | 16:51 | |
| *** zul has joined #openstack-dev | 16:51 | |
| *** garyk has quit IRC | 16:51 | |
| *** sdake has quit IRC | 16:52 | |
| *** sdake has joined #openstack-dev | 16:53 | |
| *** gyee has joined #openstack-dev | 16:55 | |
| *** shevek_ has joined #openstack-dev | 16:56 | |
| mtaylor | ola | 16:56 |
| *** gyee has quit IRC | 16:56 | |
| mtaylor | eglynn__: I would very much like to delete all of hte run_tests.py that exist anywhere | 16:56 |
| mtaylor | eglynn__: so if we could get glance-logcapture published to pypi so that wecan properly build-depend on it, that would make me a happy camper | 16:57 |
| *** adjohn has joined #openstack-dev | 16:57 | |
| mtaylor | eglynn__: also, if we're going to do that, we should make it a gerrit-controlled repo like we did for openstack.nose_plugin, no? | 16:58 |
| *** gyee has joined #openstack-dev | 16:58 | |
| eglynn__ | mtaylor: well its a tiny plugin, seems a bit OTT to set up a repo and publish to pypi etc. ... I'll check first if I can figure out how to load it directly with nosetests | 16:59 |
| mtaylor | eglynn__: ok. I mean, we can also have tox call something else ... but I'd really prefer to mve us away from custom test runners | 17:02 |
| eglynn__ | mtaylor: cool, I'll have a dig and see if I figure out something cleaner than calling run_tests.sh directly | 17:03 |
| mtaylor | eglynn__: cool. thanks! | 17:09 |
| jeblair | jaypipes: cool! left some feedback | 17:09 |
| *** mattstep has quit IRC | 17:10 | |
| *** mattstep has joined #openstack-dev | 17:11 | |
| jaypipes | jeblair: thx man! | 17:11 |
| *** dneary has quit IRC | 17:18 | |
| heckj | markmc: ping | 17:20 |
| *** dachary has quit IRC | 17:21 | |
| *** shevek_ has quit IRC | 17:25 | |
| eglynn__ | heckj: it's a public holiday today in Ireland, so he may not be around ... | 17:26 |
| *** andrewsmedina has quit IRC | 17:26 | |
| heckj | eglynn__ - well darnit. Thanks for the head's up! | 17:27 |
| *** andrewsmedina has joined #openstack-dev | 17:27 | |
| bcwaldon | eglynn__: which holiday is it? | 17:28 |
| bcwaldon | eglynn__: off today for st. patty's day? | 17:28 |
| eglynn__ | bcwaldon: in lieu of St Patricks day (fell on Saturday this year) | 17:28 |
| eglynn__ | yep | 17:28 |
| *** bengrue has quit IRC | 17:29 | |
| Kiall | Yea - Its unlikely anyone in Ireland (bar me) is working today ;) | 17:31 |
| *** anotherjesse is now known as anotherjesse_zz | 17:31 | |
| *** anotherjesse_zz is now known as anotherjesse | 17:32 | |
| Kiall | also.. bcwaldon its St. Paddy's day.. (cant help it - seen "Patty's Day" way too much this year!) | 17:33 |
| bcwaldon | Kiall: I apologize for my American ignorance :) | 17:33 |
| bcwaldon | Kiall: but how do you get Paddy from Patrick? | 17:33 |
| Kiall | That I can't explain! But patty would be even harder ;) St Patrick is a guy | 17:34 |
| bcwaldon | Kiall: well Patty and Patrick both start with Pat, that's why that confusion exists | 17:34 |
| bcwaldon | Kiall: anywho, thanks for correcting me :) | 17:34 |
| Kiall | Lol - As I said, I couldn't help myself! | 17:35 |
| bcwaldon | Kiall: no worries | 17:35 |
| Kiall | On that note.. its time to head home.. cyas | 17:35 |
| bcwaldon | Kiall: until next time... | 17:35 |
| *** mdomsch has quit IRC | 17:37 | |
| *** dachary has joined #openstack-dev | 17:37 | |
| *** vizsla has joined #openstack-dev | 17:37 | |
| *** jakedahn_zz is now known as jakedahn | 17:38 | |
| *** bsza has quit IRC | 17:42 | |
| *** jakedahn is now known as jakedahn_zz | 17:43 | |
| *** vizsla has quit IRC | 17:46 | |
| *** rnirmal_ has joined #openstack-dev | 17:46 | |
| *** rnirmal_ has quit IRC | 17:47 | |
| *** rnirmal has quit IRC | 17:50 | |
| *** novas0x2a|laptop has joined #openstack-dev | 17:50 | |
| *** dtroyer is now known as dtroyer_zzz | 17:50 | |
| *** jakedahn_zz is now known as jakedahn | 17:50 | |
| *** mattray has quit IRC | 18:01 | |
| *** reidrac has quit IRC | 18:01 | |
| *** jakedahn is now known as jakedahn_zz | 18:02 | |
| *** mattray has joined #openstack-dev | 18:03 | |
| *** jakedahn_zz is now known as jakedahn | 18:06 | |
| *** johnpostlethwait has joined #openstack-dev | 18:11 | |
| *** bengrue has joined #openstack-dev | 18:12 | |
| *** n0ano has joined #openstack-dev | 18:14 | |
| rmk | Ugh, I forgot to register for the developer summit and now it's full. | 18:24 |
| eglynn__ | mtaylor: I'm thinking the best way to package glance logcapture plugin would be to add it to openstack-nose as an additional plugin | 18:25 |
| eglynn__ | mtaylor: after recasting that to be more of a collection of plugins as opposed to *the* openstack-nose plugin (singular) | 18:25 |
| eglynn__ | mtaylor: so instead of just with-openstack, we'd have with-nova-test-report, with-glance-logcapture, with-keystone-something-else ... | 18:26 |
| eglynn__ | mtaylor: there's a bit of work there tho' and the initial point of the exercise was to get closure on the mysterious S3-related failures we've been seeing on Jenkins | 18:26 |
| eglynn__ | mtaylor: so in the interests of making headway on the test failures, I'd propose to temporarily switch over to run_tests.sh from nosetests | 18:26 |
| eglynn__ | (with the understanding that this will be reverted once the S3 issue is isolated) | 18:27 |
| *** relateable has joined #openstack-dev | 18:27 | |
| *** lts has quit IRC | 18:28 | |
| eglynn__ | mtaylor: i.e. this simple change to tox.ini: http://fpaste.org/S5is/ | 18:28 |
| *** rnirmal has joined #openstack-dev | 18:28 | |
| *** anotherjesse is now known as anotherjesse_zz | 18:30 | |
| mtaylor | eglynn__: ok. (you want to probably add -N to that) | 18:30 |
| mtaylor | eglso that it doesn't try to make another venv | 18:30 |
| mtaylor | but I'm ok with that for the short term as long as we have an exit strategry | 18:30 |
| eglynn__ | mtaylor: cool, thanks! | 18:31 |
| rmk | Is there any way to still attend the developer summit? | 18:38 |
| *** lts has joined #openstack-dev | 18:40 | |
| eglynn__ | jaypipes: the logcapture saga continues ... https://review.openstack.org/5523 | 18:42 |
| zaitcev | rmk: Write to events@rackspace.com, ask. At some point there was a mix-up and I registered twice. What if they didn't expunge the bad registration? | 18:43 |
| zaitcev | rmk: Not necesserily my fault, but perhaps the total count is conservative. | 18:44 |
| rmk | cool -- I've give that a shot, thanks | 18:44 |
| *** mnewby has joined #openstack-dev | 18:49 | |
| *** anotherjesse_zz is now known as anotherjesse | 18:49 | |
| *** Ryan_Lane has joined #openstack-dev | 18:55 | |
| *** jdg has joined #openstack-dev | 18:59 | |
| *** anotherjesse is now known as anotherjesse_zz | 19:03 | |
| *** anotherjesse_zz is now known as anotherjesse | 19:06 | |
| *** oneiroi has joined #openstack-dev | 19:07 | |
| *** dtroyer_zzz is now known as dtroyer | 19:07 | |
| *** oneiroi has quit IRC | 19:08 | |
| *** oneiroi has joined #openstack-dev | 19:08 | |
| *** dolphm has quit IRC | 19:09 | |
| *** dolphm has joined #openstack-dev | 19:12 | |
| *** dtroyer is now known as dtroyer_zzz | 19:14 | |
| *** adjohn has quit IRC | 19:17 | |
| anotherjesse | heckj: https://review.openstack.org/#change,5525 | 19:18 |
| *** lts has quit IRC | 19:20 | |
| *** zul has quit IRC | 19:21 | |
| *** aclark_ has joined #openstack-dev | 19:25 | |
| *** zul has joined #openstack-dev | 19:26 | |
| *** dtroyer_zzz is now known as dtroyer | 19:27 | |
| *** AlanClark has quit IRC | 19:28 | |
| *** anotherjesse is now known as anotherjesse_zz | 19:28 | |
| *** lts has joined #openstack-dev | 19:32 | |
| *** vincentricci has quit IRC | 19:36 | |
| *** rkukura has quit IRC | 19:38 | |
| jaypipes | eglynn__, jeblair: do we have any idea what is going on with https://jenkins.openstack.org/job/gate-glance-python26/164/console | 19:39 |
| eglynn__ | jaypipes: just looking at it now ... | 19:40 |
| *** dalang has joined #openstack-dev | 19:44 | |
| *** aclark_ has quit IRC | 19:45 | |
| *** asalkeld has joined #openstack-dev | 19:46 | |
| *** aclark_ has joined #openstack-dev | 19:47 | |
| *** anotherjesse_zz is now known as anotherjesse | 19:50 | |
| *** vincentricci has joined #openstack-dev | 19:51 | |
| *** anotherjesse is now known as anotherjesse_zz | 19:52 | |
| andrewbogott | ttx: I'm hoping that there will be one or more sessions at the summit to discuss writing a nova plugin framework and/or establishing ways to isolate extension code from core code. Is http://summit.openstack.org/sessions/view/21 that? | 19:52 |
| *** adjohn has joined #openstack-dev | 19:53 | |
| ttx | andrewbogott: yes | 19:53 |
| andrewbogott | Cool. 55 minutes seems optimistic :) | 19:53 |
| ttx | andrewbogott: we can have follow-up sessions programmed on the spot | 19:54 |
| andrewbogott | Ah, true. | 19:54 |
| ttx | but if there is more room, we can definitely extend :) | 19:54 |
| ttx | The Nova tracks are generally quite full. | 19:54 |
| andrewbogott | Do you know if anyone is planning to show up with an actual plugin-interface design in hand? Or if that would be useful? (Might be putting the cart before the horse.) | 19:55 |
| ttx | andrewbogott: i don't know. I won't have a design :) | 19:56 |
| andrewbogott | ok | 19:56 |
| ttx | but it can't hurt to prepare. | 19:56 |
| Ash | it's called CartHorse | 19:56 |
| *** aclark__ has joined #openstack-dev | 19:57 | |
| andrewbogott | I'll try to at least have a list of demands, if not a design :) | 19:57 |
| *** aclark_ has quit IRC | 19:57 | |
| Ash | <andrewbogott> one.. MILLION dollars | 19:57 |
| *** markmc has joined #openstack-dev | 19:57 | |
| markmc | heckj, dolphm, hey | 19:58 |
| markmc | heckj, dolphm, thanks for pushing ahead | 19:58 |
| heckj | markmc: arent you supposed to be holidaying ? | 19:58 |
| dolphm | markmc: np, hope you approve the giganto patch :) | 19:58 |
| markmc | heckj, yeah :) | 19:58 |
| markmc | dolphm, saw that | 19:58 |
| markmc | can we back off from the giant patch? | 19:58 |
| * markmc thinks the history of individual commits is useful | 19:58 | |
| markmc | it took a bit of work to keep them separate | 19:59 |
| markmc | happy to do whatever rebases to get it in now | 19:59 |
| heckj | markmc: I think dolphm just reapproved, but not sure of the status of that through jenkins | 19:59 |
| markmc | heckj, don't we need a devstack change too? | 19:59 |
| * markmc looks at the devstack failure | 19:59 | |
| heckj | markmc: didn't expect to need one for this | 20:00 |
| markmc | heckj, https://review.openstack.org/#change,5281 | 20:00 |
| dolphm | heckj: markmc: the first attempt only failed pep8 | 20:01 |
| markmc | dolphm, devstack too: https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/2727/ | 20:01 |
| dolphm | markmc: ugh, totally missed that | 20:01 |
| heckj | they need to be able to go in sequentially - submit both at once I'm afraid | 20:02 |
| heckj | (er, can't submit both at once) | 20:02 |
| markmc | yeah | 20:04 |
| markmc | we had this with glance too | 20:04 |
| dolphm | devstack needs to support both for a moment? | 20:04 |
| heckj | basically, yeah | 20:05 |
| markmc | https://bugs.launchpad.net/bugs/917593 | 20:05 |
| uvirtbot | Launchpad bug 917593 in openstack-ci "Co-ordinating glance/nova/keystone changes with devstack" [Critical,Won't fix] | 20:05 |
| heckj | intentionally set up so we walk the code forward as we go, making sure we dont introduce breaking changes | 20:06 |
| markmc | eglynn__, there by any chance? | 20:07 |
| eglynn__ | markmc: yep | 20:08 |
| markmc | eglynn__, see above? need devstack to support both | 20:08 |
| markmc | eglynn__, what did you do for glance again? | 20:08 |
| bcwaldon | markmc: how do I make a cli option required? | 20:08 |
| bcwaldon | markmc: with connon config stuff | 20:08 |
| bcwaldon | markmc: common* | 20:08 |
| *** dolphm has quit IRC | 20:09 | |
| markmc | bcwaldon, at the moment, just check if it's set, raise an exception if not | 20:09 |
| markmc | bcwaldon, we could add required=True flag or something, though | 20:09 |
| eglynn__ | markmc: the chicken and egg issue when I reworked the paste config? | 20:09 |
| markmc | eglynn__, yeah | 20:09 |
| bcwaldon | markmc: I would consider that a necessary feature | 20:09 |
| markmc | bcwaldon, does optparse have it? | 20:09 |
| bcwaldon | markmc: argparse does | 20:10 |
| markmc | bcwaldon, feel free to file against openstack-common | 20:10 |
| eglynn__ | markmc: in that case we just copied the paste config in the *.conf as well as putting in the separate *-paste.ini | 20:10 |
| markmc | bcwaldon, right, I know argparse has it | 20:10 |
| bcwaldon | markmc: argparse is replacing optparse, anywho | 20:10 |
| markmc | bcwaldon, yep | 20:10 |
| bcwaldon | markmc: reason we didnt go with that? | 20:10 |
| markmc | bcwaldon, yes, can't remember offhand though | 20:10 |
| eglynn__ | markmc: sorrta a special case I guess, just duplicating the paste config was sufficient to work around | 20:10 |
| markmc | bcwaldon, looked into it, pretty confident we can switch to it later | 20:10 |
| markmc | bcwaldon, think it was just I didn't want to switch both nova and glance to something new and glance was using optparse | 20:11 |
| eglynn__ | markmc: whereas your keystone cfg changes are more intrusive, right? | 20:11 |
| bcwaldon | markmc: ok | 20:11 |
| *** anotherjesse_zz is now known as anotherjesse | 20:11 | |
| markmc | eglynn__, I made identical changes to glance before your stuff, not sure how that was pushed through | 20:11 |
| eglynn__ | markmc: was devstack gating merges at that time? | 20:11 |
| markmc | eglynn__, yeah | 20:12 |
| markmc | eglynn__, I'll find it, one sec | 20:12 |
| *** troytoman is now known as troytoman-away | 20:12 | |
| *** danwent has joined #openstack-dev | 20:13 | |
| heckj | markmc: since that big one failed, I'm fine with going back to the single commits and doing them individually. | 20:13 |
| markmc | heckj, ok, pushed another rebase | 20:13 |
| markmc | heckj, should be able to get them all in except the last two without the devstack change | 20:14 |
| * markmc pokes at devstack | 20:14 | |
| markmc | ah, we weren't gating on devstack when we did this for glance | 20:15 |
| *** novas0x2a|laptop has quit IRC | 20:16 | |
| *** novas0x2a|laptop has joined #openstack-dev | 20:16 | |
| * heckj nods | 20:17 | |
| markmc | glance devstack change was https://review.openstack.org/2343 | 20:17 |
| eglynn__ | markmc: a-ha | 20:17 |
| eglynn__ | markmc: resurrect your idea on pairing the devstack patch with the keystone/glance/nova change that would break the old devstack? | 20:19 |
| markmc | eglynn__, think we want to move quicker than that | 20:19 |
| markmc | something like | 20:19 |
| markmc | in glance checkout | 20:19 |
| markmc | if ! git rev-parse 35372906d >/dev/null 2>&!; do sed -i .... etc/keystone.conf; done | 20:20 |
| markmc | and then after we've merged the keystone changes | 20:20 |
| markmc | merge your proper config file update? | 20:20 |
| eglynn__ | markmc: you meant 'in keystone checkout' or? | 20:21 |
| eglynn__ | (as opposed to glance) | 20:22 |
| markmc | yep, sorry | 20:22 |
| *** anotherjesse is now known as anotherjesse_zz | 20:22 | |
| *** anotherjesse_zz is now known as anotherjesse | 20:23 | |
| eglynn__ | markmc can you explain: if ! git rev-parse 35372906d ... | 20:24 |
| markmc | eglynn__, should check whether the commit 35372906d exists in the repo | 20:24 |
| eglynn__ | markmc: gotcha, nice | 20:24 |
| markmc | eglynn__, really we want a clever way of saying "is 35372906d an ancestor of HEAD" | 20:24 |
| markmc | eglynn__, but I can't think of something nice offhand | 20:24 |
| *** anotherjesse is now known as anotherjesse_zz | 20:24 | |
| eglynn__ | markmc: so devstack uses a pre-canned version of the keystone.conf, so that rev-parse could go into stack.sh, and drive selection between a new or old config file template | 20:26 |
| eglynn__ | (i.e. devstack clones keystone, then copies over a keystone.conf it prepared earlier ...) | 20:27 |
| markmc | right | 20:28 |
| eglynn__ | should work | 20:28 |
| eglynn__ | should run the idea by some devstack core folks, dtroyer say ... | 20:28 |
| markmc | actually, can't do a sed | 20:29 |
| markmc | it would need to be something like cp files/keystone.conf.new files/keystone.conf | 20:29 |
| eglynn__ | yep | 20:29 |
| markmc | dtroyer, there? | 20:30 |
| dtroyer | markmc: yo, catching up | 20:30 |
| markmc | dtroyer, basically just need to switch to a new keystone conf file | 20:30 |
| markmc | dtroyer, but the new one doesn't work with current keystone | 20:30 |
| eglynn__ | the keystone.conf.new could just be pulled from: https://review.openstack.org/#change,5281 | 20:31 |
| dtroyer | markmc: ok. how do we know which one to use? | 20:31 |
| markmc | dtroyer, if ! git rev-parse 35372906d >/dev/null 2>&!; then ... | 20:31 |
| *** dolphm has joined #openstack-dev | 20:31 | |
| *** zigo has quit IRC | 20:31 | |
| *** bengrue has quit IRC | 20:32 | |
| markmc | dtroyer, well, if git rev-parse 35372906d >/dev/null 2>&!; then use_new_conf; fi | 20:32 |
| *** dprince has quit IRC | 20:34 | |
| *** Vek has quit IRC | 20:35 | |
| markmc | dtroyer, simpler, if grep keystone.common.wsgi etc/keystone.conf >/dev/null; then use_new_conf; fi | 20:35 |
| markmc | dtroyer, i.e. check whether the keystone checkout has the new config | 20:36 |
| jeblair | markmc: there's another pattern for handling this kind of change that i think is pretty elegant | 20:36 |
| jeblair | markmc: check out https://github.com/openstack-dev/devstack/blob/master/stack.sh#L881 | 20:36 |
| jeblair | where the nova api-paste.ini file from _nova_ is used as the template | 20:37 |
| jeblair | rather than one that lives with devstack. | 20:37 |
| markmc | absolutely, would much rather that | 20:37 |
| dtroyer | ok, I think that will work. we'll want to carry the new file inside devstack rather than pull it from gerrit just to keep the external dependencies down | 20:37 |
| dtroyer | jeblair, markmc: I actually have a bug on doing exactly that | 20:38 |
| dtroyer | I think it's within reach for keystone | 20:38 |
| jeblair | dtroyer: cool. i think that will reduce the number of times we run into this in the long run. | 20:38 |
| jeblair | only _really fundamental_ changes would need switching logic in devstack | 20:38 |
| jeblair | you guys seem on it. i'll leave you to it; just wanted to make sure that wasn't overlooked. | 20:39 |
| dtroyer | I'd love to do it for all projects, but following the changes to glance recently shows me that some of them are just not going to be possible because of the variations in pipeline configs, etc. | 20:39 |
| markmc | current diff between devstack and keystone confs | 20:40 |
| markmc | http://paste.openstack.org/show/11137/ | 20:40 |
| markmc | admin_token, compute_port, sql.connection, sql.idle_timeout, catalog.template_file | 20:42 |
| markmc | that looks to be the only important changes | 20:42 |
| justinsb_ | You could use this script to set config values cleanly: https://github.com/justinsb/openstack-simple-config/blob/master/utils/openstack-config-set | 20:42 |
| justinsb_ | I borrowed the idea from Fedora's packaging | 20:43 |
| markmc | justinsb_, yeah, we should really have that upstream somewhere | 20:44 |
| justinsb_ | markmc: I'm hoping to create a project / area within Openstack that will make packaging easier | 20:44 |
| markmc | justinsb_, ok, cool | 20:44 |
| justinsb_ | markmc: http://summit.openstack.org/sessions/view/26 | 20:45 |
| markmc | justinsb_, heh, user config in paste config | 20:46 |
| markmc | justinsb_, definite bugbear of mine too :) | 20:46 |
| markmc | heckj, there? | 20:48 |
| markmc | hmm, lunch time on west coast | 20:49 |
| * markmc assumes heckj has gone for a surf or something | 20:49 | |
| markmc | that's what you guys all do, right? :) | 20:49 |
| *** anotherjesse_zz is now known as anotherjesse | 20:50 | |
| *** anotherjesse is now known as anotherjesse_zz | 20:51 | |
| justinsb_ | markmc: Yes, but we normally take our phones with us. | 20:52 |
| heckj | markmc: sorry, here - in a meeting | 20:52 |
| markmc | justinsb_, we're not so considerate when we go for our lunchtime 4 pints of guinness :) | 20:53 |
| markmc | heckj, np | 20:53 |
| heckj | 4 pints of guiness? Sounds damn good right now | 20:53 |
| russellb | +1 | 20:53 |
| markmc | heckj, was just going to say, you can approve https://review.openstack.org/4557 and everything below it | 20:54 |
| markmc | heckj, then we can figure out something about devstack | 20:54 |
| heckj | on it | 20:54 |
| *** dneary has joined #openstack-dev | 20:54 | |
| ttx | vishy: yo | 20:54 |
| markmc | heckj, cool | 20:55 |
| heckj | markmc: take a look at https://review.openstack.org/#q,status:open+keystone,n,z - is that the right set | 20:56 |
| heckj | markmc: do I need to hit the others in that big set? I just followed "needed by" and it was a short list... | 20:57 |
| markmc | heckj, this shouldn't be in it: https://review.openstack.org/4558 | 20:57 |
| *** dolphm has quit IRC | 20:57 | |
| *** rkukura has joined #openstack-dev | 20:57 | |
| markmc | heckj, nor should this: https://review.openstack.org/4559 | 20:57 |
| markmc | heckj, everything in https://review.openstack.org/#q,status:open+project:openstack/keystone+owner:markmc,n,z except those two | 20:58 |
| heckj | pulled 4558, 4559 | 20:58 |
| markmc | ok, I think I have a solution | 21:02 |
| markmc | I'll add the old factory methods back to https://review.openstack.org/4558 | 21:02 |
| markmc | so it will support the old and new config | 21:02 |
| heckj | markmc: Ok - take a look now - should be set. | 21:03 |
| markmc | and remove them again in https://review.openstack.org/4559 | 21:03 |
| markmc | we approve 4558 | 21:03 |
| markmc | get the new config into devstack | 21:03 |
| *** zzed has quit IRC | 21:03 | |
| markmc | then approve 4559 | 21:03 |
| *** Gordonz has quit IRC | 21:03 | |
| heckj | markmc: sounds like a lovely waltz | 21:03 |
| *** aclark__ has quit IRC | 21:03 | |
| markmc | heckj, looks good | 21:03 |
| *** zzed has joined #openstack-dev | 21:04 | |
| notmyname | ttx: swift 1.4.8 is at commit b8e1513ad7de8b88c95986e20b9a03d5b8207fd6 | 21:06 |
| ttx | notmyname: ok, will cut milestone-proposed in my tomorrow morning | 21:07 |
| notmyname | ttx: thanks. I'll try to get a go-nogo by wed pm for a release on thurs | 21:07 |
| ttx | notmyname: sounds good. | 21:08 |
| eglynn__ | markmc: yep, thats way neater | 21:09 |
| markmc | heckj, ok, pushed new versions of 4558 and 4559 | 21:09 |
| markmc | heckj, I guess it's best to just wait until all the others are merged before doing the dance | 21:09 |
| heckj | markms: yeah, sounds look a good plan | 21:09 |
| markmc | heckj, https://review.openstack.org/5281 is the devstack commit | 21:09 |
| *** zzed has quit IRC | 21:18 | |
| *** lts has quit IRC | 21:19 | |
| *** dolphm has joined #openstack-dev | 21:19 | |
| *** jdg has quit IRC | 21:24 | |
| *** mancdaz1203 has quit IRC | 21:28 | |
| *** mancdaz1203 has joined #openstack-dev | 21:30 | |
| *** johnpostlethwait has quit IRC | 21:30 | |
| *** johnpostlethwait has joined #openstack-dev | 21:31 | |
| eglynn__ | jaypipes: found the issue that was causing https://review.openstack.org/5523 to bomb out on 2.6, looks like setUp() wasn't called reliably for skipped tests | 21:36 |
| eglynn__ | fixed in patch set 2 | 21:36 |
| *** rnirmal has quit IRC | 21:44 | |
| *** vizsla has joined #openstack-dev | 21:46 | |
| *** utlemming has quit IRC | 21:48 | |
| *** adjohn has quit IRC | 21:50 | |
| *** vizsla has quit IRC | 21:50 | |
| *** utlemming has joined #openstack-dev | 21:52 | |
| vishy | comstud: can you check this for any issues: https://review.openstack.org/#change,5518 ? | 21:52 |
| *** rgoodwin is now known as rgoodwin_away | 21:54 | |
| *** rgoodwin_away is now known as rgoodwin | 21:54 | |
| *** apevec has quit IRC | 21:56 | |
| vishy | pvo, comstud: I think we are shipping rc-1 tonight unless one of you guys has a reason that we need to delay... | 21:58 |
| *** dolphm has quit IRC | 22:00 | |
| *** adjohn has joined #openstack-dev | 22:00 | |
| *** rgoodwin is now known as rgoodwin_away | 22:02 | |
| *** rgoodwin_away is now known as rgoodwin | 22:04 | |
| justinsb_ | vishy: You're talking about nova, right? I looked through my bugs. I'm wondering if bug #939386 could be a sign of bigger escaping problems in XML. I'm checking it out. | 22:08 |
| uvirtbot | Launchpad bug 939386 in nova "Console-Output does not escape XML correctly" [Medium,Confirmed] https://launchpad.net/bugs/939386 | 22:08 |
| *** Mandell has joined #openstack-dev | 22:09 | |
| *** crobinso has quit IRC | 22:09 | |
| *** kbringard has quit IRC | 22:10 | |
| vishy | justinsb_: are you going to submit a patch for that? | 22:11 |
| justinsb_ | vishy: I'm evaluating it now. If it's just console-output, I don't think it's too serious. If it's all XML output, I'll report back. | 22:12 |
| vishy | justinsb_: thanks | 22:13 |
| *** dolphm has joined #openstack-dev | 22:17 | |
| *** ywu has joined #openstack-dev | 22:19 | |
| rmk | Running the Glance RC package with precise -- on install it attempts to do a db_sync and throws an exception: | 22:29 |
| rmk | TypeError: Invalid argument(s) for Table: ['extend_existing', 'mysql_engine'] | 22:29 |
| rmk | Granted, the default config is sqlite | 22:29 |
| rmk | So that looks like a possible bug? | 22:29 |
| *** Mandell has quit IRC | 22:32 | |
| comstud | vishy: We had a patch coming for fixing an issue searching by IPs in nova compute API... not sure if that's made it up for review yet | 22:33 |
| vishy | comstud: haven't seen it | 22:33 |
| *** dolphm has quit IRC | 22:33 | |
| comstud | vishy: checking | 22:33 |
| vishy | rmk: old version of sqlalchemy | 22:33 |
| vishy | rmk: needs useexisting | 22:34 |
| vishy | rmk: I guess the patch to glance to switch it should be reverted | 22:34 |
| rmk | vishy: Thanks -- this was on 11.10 | 22:34 |
| vishy | bcwaldon, jaypipes: ^^ | 22:34 |
| vishy | rmk: ah that makes sense, yes oneiric that would happen | 22:34 |
| rmk | I'd rebuilt all the packages from Precise for it. I can upgrade sqlalchemy but yeah it'll break oneiric people. | 22:34 |
| *** markmc has quit IRC | 22:34 | |
| rmk | Though hopefully most are looking to go to 12.04 now that it's an LTS. | 22:35 |
| bcwaldon | file a bug for us :) | 22:35 |
| *** dolphm has joined #openstack-dev | 22:35 | |
| rmk | Sure, though is it a bug or do you want to require a newer sqlalchemy? | 22:35 |
| bcwaldon | can you file a bug with what you see as broken, then we'll be able to discuss the appropriate fix when everyone understands | 22:36 |
| rmk | cool will do | 22:36 |
| bcwaldon | I havent been following your convo with vishy, so I don't really know the problem yet | 22:36 |
| *** Vek has joined #openstack-dev | 22:36 | |
| *** Mandell has joined #openstack-dev | 22:36 | |
| bcwaldon | rmk: thanks! | 22:37 |
| comstud | vishy: the search-by-IP bug is only a bug when using quantum/melange.. i guess it'll be fixed with a larger re-work | 22:39 |
| vishy | bcwaldon: it is the issue with sqlalchemy versions useexisting vs extend_existing | 22:39 |
| comstud | vishy: i doubt anyone really uses the search-by-ip function anyway, so I think it's okay to live with that bug being in essex | 22:39 |
| vishy | bcwaldon: glance was updated to the new non-deprecated version which breaks with old versions of sqlalchemy | 22:39 |
| vishy | comstud: ok | 22:39 |
| bcwaldon | vishy: I don't have any previous context related to that... | 22:39 |
| bcwaldon | vishy: ok, sqlalchemy bug | 22:39 |
| comstud | so, i think as far as we're concerned, we're good to go | 22:39 |
| bcwaldon | vishy: great... | 22:39 |
| *** dachary has quit IRC | 22:42 | |
| vishy | jaypipes: in regards to your ratelimit bug | 22:42 |
| vishy | jaypipes: I put a bug on nova to have a way to disable rate limiting when I broke that flag in devstack | 22:43 |
| vishy | jaypipes: but I was told that it isn't vital for tempest because there is a workaround, so I untargeted it | 22:43 |
| jaypipes | vishy: hmm :) | 22:43 |
| *** Mandell has quit IRC | 22:44 | |
| *** mattray has quit IRC | 22:44 | |
| jaypipes | vishy: well, the "workaround" is to let the tempest client sleep for the number of seconds needed, which makes tempest run in about an hour. ;) | 22:44 |
| vishy | jaypipes: we could do a hacky simple version and just add a new pipeline | 22:44 |
| jaypipes | vishy: running out to dinner with the wifey... we'll chat later about it? :) | 22:44 |
| vishy | sure | 22:45 |
| *** Mandell has joined #openstack-dev | 22:53 | |
| justinsb_ | vishy: It's a xml / minidom bug I think: http://paste.openstack.org/show/11168/ | 22:54 |
| justinsb_ | vishy: Some characters escaped, some not escaped | 22:54 |
| justinsb_ | vishy: Trying to figure out if it's minidom or our usage that's to blame | 22:54 |
| *** dtroyer is now known as dtroyer_zzz | 22:56 | |
| vishy | justinsb_: we shouldn't be using xml.dom.minidom | 22:59 |
| vishy | justinsb_: it should be lxml only | 22:59 |
| vishy | there is a patch out to convert most of it | 22:59 |
| *** dolphm has quit IRC | 23:00 | |
| *** Mandell has quit IRC | 23:01 | |
| *** Mandell has joined #openstack-dev | 23:01 | |
| justinsb_ | vishy: Do you have a link on that patch? wsgi.py still using minidom... | 23:02 |
| *** dolphm has joined #openstack-dev | 23:03 | |
| rmk | bcwaldon: https://bugs.launchpad.net/glance/+bug/959787 | 23:03 |
| uvirtbot | Launchpad bug 959787 in glance "Glance database creation fails using RC1 on Ubuntu 11.10" [Undecided,New] | 23:03 |
| *** Mandell has quit IRC | 23:06 | |
| *** andrewsmedina has quit IRC | 23:06 | |
| vishy | justinsb_: https://review.openstack.org/#change,5422 | 23:07 |
| vishy | justinsb_: i think it missed the minidom usages though | 23:08 |
| justinsb_ | vishy: Yes - I don't see nova/api/openstack/wsgi.py in that list | 23:08 |
| justinsb_ | vishy: I think this is the bug http://bugs.python.org/issue5752 | 23:09 |
| justinsb_ | vishy: I don't think it's particularly serious; I think the XML produced is just invalid; I don't think it's a vulnerability or anything like that. | 23:10 |
| justinsb_ | vishy: But it does mean we're potentially producing invalid XML anywhere we're outputing a string | 23:11 |
| justinsb_ | vishy: In this case the problematic character is 0x1B, the escape character | 23:11 |
| justinsb_ | vishy: I don't think tabs and newlines create invalid documents, just documents that don't exactly reflect the input | 23:11 |
| justinsb_ | vishy: I think how you want to fix this is going to be your call | 23:12 |
| justinsb_ | vishy: If you want to fix it at all! | 23:12 |
| vishy | console output should strip ctrl characters | 23:12 |
| justinsb_ | vishy: OK ... I can implement that pretty easily. | 23:12 |
| justinsb_ | vishy: We'll just have to remember that this isn't really fixed, just hidden! | 23:13 |
| vishy | jaypipes: when you get back: does this work: https://review.openstack.org/5543 ? | 23:13 |
| *** thingee has quit IRC | 23:16 | |
| *** yoshiotu has joined #openstack-dev | 23:25 | |
| *** dachary has joined #openstack-dev | 23:26 | |
| *** yoshiotu has left #openstack-dev | 23:29 | |
| *** Kiall has quit IRC | 23:32 | |
| *** Gordonz has joined #openstack-dev | 23:32 | |
| *** Kiall has joined #openstack-dev | 23:33 | |
| justinsb_ | vishy: Stripping tricky characters out of console output https://review.openstack.org/#change,5545 | 23:34 |
| justinsb_ | vishy: I don't really mind if that makes it in or not. I can just use the JSON output. | 23:34 |
| *** phorgh has quit IRC | 23:34 | |
| *** andrewsmedina has joined #openstack-dev | 23:37 | |
| *** test400 has joined #openstack-dev | 23:41 | |
| *** Gordonz has quit IRC | 23:41 | |
| *** vincentricci has quit IRC | 23:47 | |
| *** adjohn has quit IRC | 23:47 | |
| *** adjohn has joined #openstack-dev | 23:50 | |
| *** vricci has joined #openstack-dev | 23:50 | |
| pvo | bcwaldon: vishy: no huge blockers here. | 23:54 |
| pvo | pull the chute! | 23:54 |
| *** novas0x2a|laptop has quit IRC | 23:56 | |
| *** adjohn has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!