Wednesday, 2014-03-26

*** kgriffs is now known as kgriffs_afk00:00
*** whenry has quit IRC00:30
*** tjanczuk has left #openstack-marconi00:54
*** rossk has quit IRC01:02
*** reed has quit IRC01:09
*** nosnos has joined #openstack-marconi01:27
*** prashanthr_1 has quit IRC01:55
*** flwang has quit IRC02:00
*** mpanetta has joined #openstack-marconi02:00
*** flwang has joined #openstack-marconi02:41
*** vkmc has quit IRC03:21
*** ayoung has quit IRC03:26
*** nosnos has quit IRC03:28
*** mpanetta has quit IRC04:02
prashanthr_flwang: Hi ,04:34
*** nosnos has joined #openstack-marconi04:36
*** flwang1 has joined #openstack-marconi05:27
*** flwang has quit IRC05:29
*** saju_m has joined #openstack-marconi06:19
*** ykaplan has joined #openstack-marconi06:33
*** flwang1 has quit IRC07:03
*** flwang has joined #openstack-marconi07:06
*** saju_m has quit IRC07:13
*** ykaplan has quit IRC07:28
*** flaper87|afk is now known as flaper8707:57
flwangflaper87: hey hey07:58
flaper87flwang: hey :D07:59
flaper87wow, there was so much movement here yday07:59
flwangflaper87: ready for questions from China?07:59
flaper87flwang: 2 secs08:01
* flaper87 sips some coffee08:01
flaper87flwang: ok, shoot08:02
flwangflaper87: after enable the keystone strategy, should we add config for keystone auth in marconi.conf?08:03
flaper87flwang: yup, you need to add the common keystone configs08:08
flaper87flwang: under the keystone_authtoken section08:08
flaper87just like in glance08:08
flaper87besides that08:09
flaper87you need to also set the auth_strategy08:09
flaper87to keystone08:09
flaper87and that's it08:09
flwangflaper87: I know the auth_strategy, just not sure if I should add the keystone section, seems we didn't mention that in our marconi.conf.sample :(08:10
flwangflaper87: I will post a patch to cover that. you know, we even didn't say what's the options list for auth_strategy08:10
flaper87flwang: mmh, wait, that should have been done by the config generator08:16
flwangwhere is the config generator?08:16
flaper87ahhhh, I know why it wasn't covered08:25
flaper87I think08:25
flaper87flwang: yeah, I figured it out08:29
flaper87fixed08:29
flaper87I'll push a fix in a bit08:29
flwangflaper87: ok08:29
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Add keystone_authtoken to our config generator  https://review.openstack.org/8300408:40
openstackgerritFlavio Percoco proposed a change to openstack/marconi: keystoneclient now supports Py3K  https://review.openstack.org/8300508:40
openstackgerritFlavio Percoco proposed a change to openstack/marconi: Add a help text for auth_strategy  https://review.openstack.org/8300608:40
flaper87flwang: ^^^^08:40
flwanggood08:41
flwanglooks good for me08:42
flwangthanks for fixing it08:42
*** chandankumar_ has joined #openstack-marconi08:42
openstackgerritFei Long Wang proposed a change to openstack/marconi: Cleanup fixtures  https://review.openstack.org/8301609:24
*** flwang has quit IRC09:35
*** chandankumar_ has quit IRC10:22
*** ykaplan has joined #openstack-marconi10:33
*** saju_m has joined #openstack-marconi11:20
*** mwagner_ is now known as mwagner_notHere11:29
openstackgerritSergey Lukjanov proposed a change to openstack/marconi: Switch to oslosphinx  https://review.openstack.org/8304211:29
*** tedross has joined #openstack-marconi11:51
*** mwagner_lap has quit IRC12:00
*** vkmc has joined #openstack-marconi12:03
*** flwang has joined #openstack-marconi12:08
*** oz_akan_ has joined #openstack-marconi12:20
*** oz_akan_ has quit IRC12:25
*** oz_akan_ has joined #openstack-marconi12:25
*** sriram has joined #openstack-marconi12:36
*** mwagner_lap has joined #openstack-marconi13:12
*** sir-o has joined #openstack-marconi13:17
*** ayoung_ has joined #openstack-marconi13:21
*** ayoung_ is now known as ayoung13:22
*** jcru has joined #openstack-marconi13:29
*** amitgandhi has joined #openstack-marconi13:32
*** amitgandhi has quit IRC13:33
*** amitgandhi has joined #openstack-marconi13:33
*** nosnos has quit IRC13:35
*** flwang has quit IRC13:49
*** flwang has joined #openstack-marconi13:50
*** malini has joined #openstack-marconi14:07
*** Kuwagata has joined #openstack-marconi14:15
*** cpallares has joined #openstack-marconi14:20
*** prashanthr_1 has joined #openstack-marconi14:29
prashanthr_1malini: Hi !14:29
malinihello prashanthr_114:33
prashanthr_1Good morning. I had a small question related to patch : https://review.openstack.org/#/c/82343/. Thanks for reviewing it.14:33
malininp..Thank for working on it14:34
prashanthr_1your review comment is actually very valid and I had initially tried to look from that perspective.14:35
prashanthr_1But very unlikely when we add the header copy into the body of the test the clean up code works perfectly.14:36
malinihmmmthe header thing should have nothing to do with the 204s14:37
maliniI suspect tht it fails to cleanUp in some cases14:37
maliniDo you have a way to repro the 204 ?14:37
prashanthr_1Yes I have written it on here :14:38
prashanthr_1https://bugs.launchpad.net/marconi/+bug/128499514:38
maliniawesome! tht will make it easier to troubleshoot14:39
maliniCan you insert a breakpoint before the create queue , to make sure the DB doesnt already have the queue?14:40
malini& Can you also chk if the cleanup is really happening?14:40
prashanthr_1Yes I had done that. Yes the cleanup is actually happening and the surprising thing is that there are no exceptions as well.14:40
*** whenry has joined #openstack-marconi14:41
malinitht is weird..So you are sure that the DB doesnt have the queue with the same name ?14:41
prashanthr_1yes I am sure.14:41
*** chandan_kumar has quit IRC14:42
maliniI am just reading your repro steps14:42
maliniIn Step 4 you have "You will observe that there is a queue named : "asteriskinheader".14:42
maliniDo you see that queue in the DB after the test run ?14:42
prashanthr_1yes.14:43
malinihmm..tht is a problem.14:43
maliniYou should not see that queue there after the test14:43
maliniIt should have been cleaned up14:43
*** chandankumar_ has joined #openstack-marconi14:44
prashanthr_1Yes. The clean up code executes but is not removing this particular queue. All the others are being deleted.14:44
maliniI am wondering if the delete fails when we have the asterisk header14:44
prashanthr_1But actually when you make a copy of the self.header14:45
prashanthr_1the delete works fine14:45
maliniTht might be a bug in the code14:45
*** chandankumar_ has quit IRC14:46
maliniWe should be able to delete with the asterisk in the header14:46
maliniJust like you can insert a queue with */* headers14:46
prashanthr_1Yes we can delete it when we make a copy of the self.header14:46
prashanthr_1and then set the header['Accept'] = '*/*'14:46
prashanthr_1then the cleanup works perfect.14:46
prashanthr_1Althoug I do not know why.14:47
*** chandan_kumar has joined #openstack-marconi14:47
maliniself.addCleanup(self.client.delete, path) uses the default headers14:47
*** jmckind has joined #openstack-marconi14:48
*** chandan_kumar has quit IRC14:48
prashanthr_1Thanks for that info. I did not know that.14:48
*** chandan_kumar has joined #openstack-marconi14:48
prashanthr_1Actually can i know the context of the particular test case ?14:48
*** reed has joined #openstack-marconi14:49
*** sir-o has quit IRC14:50
*** chandan_kumar has quit IRC14:55
*** chandan_kumar has joined #openstack-marconi14:56
maliniprashanthr_1: We should be able to handle headers with */*14:57
malinias in it should work in the same way when we have application/json14:57
prashanthr_1got it.14:58
*** sir-o has joined #openstack-marconi15:01
*** balajiiyer has joined #openstack-marconi15:02
*** chandan_kumar has quit IRC15:03
*** malini is now known as malini_afk15:07
*** chandan_kumar has joined #openstack-marconi15:09
*** mpanetta has joined #openstack-marconi15:18
*** sir-o has quit IRC15:22
*** sir-o has joined #openstack-marconi15:22
*** alcabrera has joined #openstack-marconi15:24
*** alcabrera is now known as alcabrera|afk15:33
*** mpanetta_ has joined #openstack-marconi15:36
*** mpanetta_ has quit IRC15:37
*** mpanetta_ has joined #openstack-marconi15:38
*** alcabrera|afk has quit IRC15:39
*** mpanetta has quit IRC15:40
*** saju_m has quit IRC15:43
*** kgriffs_afk is now known as kgriffs15:51
*** rossk has joined #openstack-marconi16:12
flaper87kgriffs: ping16:13
flaper87this channel is so... silent today :(16:14
openstackgerritA change was merged to openstack/marconi: Switch to oslosphinx  https://review.openstack.org/8304216:20
kgriffsflaper87: pong16:21
flaper87kgriffs: hey man :)16:23
flaper87kgriffs: so, -rc1 cut is very close16:23
flaper87anything we should definitely have before the cut ?16:23
flaper87after the -rc1 cut, we can start doing Juno work16:23
flaper87w00000t16:23
*** ykaplan has quit IRC16:25
*** amitgandhi has quit IRC16:26
kgriffsflaper87: I was just waiting on the devstack bug to see if Malini wanted to still try to get it done for Icechouse16:27
kgriffsI've moved the other two bugs out16:27
kgriffsand got the blueprint marked as implemented16:28
kgriffsttx said it's up to us when we cut... idk if he was going to wait and add all the Juno milestones after everyone has rc1 or if he'll do it project-by-project16:28
kgriffsbut, at least we will have the branch made i guess16:29
flaper87ok16:30
flaper87kgriffs: btw, I'm a bit worried about the `marconi-server` always forking :/16:30
flaper87for some reason, I thought that was something that would happen just under devstack16:31
flaper87but instead it always forks and starts as a background process16:31
kgriffsflaper87: yes, I think when someone is just running locally to kick the tires or whatever, you don't want it to fork16:31
flaper87it's kind of annoying since `marconi-server` is what we use for development, demoing, testing16:32
kgriffsso, I think by default it should not fork, but you can pass a CLI option if you want it to daemonize16:32
flaper87does devstack set an env var when it starts ?16:32
flaper87if so, we could use that16:32
kgriffsflaper87: mmm, honestly I'm not a devstack guru16:32
flaper87malini_afk: ^16:32
* flaper87 neither16:32
kgriffsbut yeah, doing it via ENV sounds good16:32
flaper87I'll figure it out16:32
flaper87kgriffs: did you have some time to go through the FAQ yday?16:33
kgriffsI have a vague recollection of seeing a place where you can set vars, but can't say for sure.16:33
kgriffsflaper87: not yet16:33
flaper87kgriffs: no hurries16:33
*** ykaplan has joined #openstack-marconi16:33
kgriffscan you paste the link again? I thought I had it open in a tab, but, well, it is going to be hard to find today. https://dl.dropboxusercontent.com/u/205175/tabs-from-hell.png16:35
flaper87kgriffs: https://etherpad.openstack.org/p/draft-marconi-faq16:36
flaper87kgriffs: lol16:36
kgriffsthanks: now I have one more tab, but at least it is on the end where I can find it. :)16:38
kgriffsflaper87: oz and I started this related pad as well - https://etherpad.openstack.org/p/marconi-mission16:38
*** mpanetta_ has quit IRC16:40
* flaper87 clicks16:40
*** sir-o has quit IRC16:52
*** amitgandhi has joined #openstack-marconi16:54
*** amitgandhi has quit IRC16:55
*** amitgandhi has joined #openstack-marconi16:56
*** chandan_kumar has quit IRC16:57
flaper87kgriffs: there?17:09
flaper87kgriffs: funny, I'm cleaning up some previous presos I did about MArconi and guess what I have in those presos: https://speakerdeck.com/flaper87/marconi-queuing-and-notification-service-for-openstack (slides 11, 12 and 13)17:10
*** mkoderer has quit IRC17:11
*** mkoderer has joined #openstack-marconi17:15
*** ykaplan has quit IRC17:19
kgriffsheh, nice17:21
* kgriffs is in a meeting. :p17:22
*** kgriffs is now known as kgriffs_afk17:22
*** whenry has quit IRC17:36
*** malini_afk is now known as malini17:40
*** prashanthr_1 has quit IRC17:40
*** whenry has joined #openstack-marconi17:46
malinio/17:46
*** saju_m has joined #openstack-marconi17:57
cpallareshi malini :)18:00
malinihello cpallares!18:00
cpallaresmalini: how are you doing?18:01
malinigood :) how are you?18:01
cpallaresmalini: I'm doing good, though a bit hungry. :P18:03
* malini gives cpallares a peanut butter pop-tart18:04
* cpallares noms on that peanut butter pop-tart18:04
*** rossk has quit IRC18:23
*** rossk has joined #openstack-marconi18:23
*** oz_akan_ has quit IRC18:31
*** oz_akan_ has joined #openstack-marconi18:36
*** flwang has quit IRC18:41
*** mpanetta has joined #openstack-marconi18:55
*** alcabrera has joined #openstack-marconi18:57
*** flwang has joined #openstack-marconi18:57
*** mpanetta has quit IRC19:00
*** mpanetta has joined #openstack-marconi19:01
*** jcru has quit IRC19:03
*** kgriffs_afk is now known as kgriffs19:04
kgriffsflaper87: btw, are you going to make the g+ stream tomorrow watchable on youtube?19:05
kgriffsmalini: ping19:07
malinikgriffs: pong19:07
kgriffshi!19:07
malinihello!19:07
kgriffshope you are having a good day. :D19:07
kgriffsso, I a couple things19:08
kgriffsfirst, final decision on the devstack mongo bug?19:08
kgriffsregarding rc1 vs. Juno19:08
maliniI think it'll be safe to move it to Juno19:08
kgriffsok.19:08
maliniI will follow up with the ceilometer team to find what they are doing19:09
maliniThe last I heard, they were not going to upgarde mongo at the gate19:09
kgriffshelp me remember to do a bug triaging session when the J-1 opens up; we have a bunch of bugs and bps that need to be reassigned milestones19:09
kgriffsmalini: WAH?19:09
kgriffshow can they not upgrade it at the gate?19:09
kgriffsthat seems pretty unworkable19:10
malinithey want to keep the same version tht comes with the distro :|19:10
maliniTht was the last thing I heard19:10
kgriffsblah. And since we aren't integrated, we have less of a say on what the distro version is19:10
kgriffsmalini: can't we just install and run our own?19:10
malinihmm..I don't know if we can do that19:11
maliniBut I'll chk what we can do19:11
kgriffsok, well, if we can't get a mongo version that supports TTL indexes then we can't run that test in gate, period.19:11
kgriffsbut maybe we can do an out-of-band testing19:12
maliniThey have something known as 3rd part testing19:12
maliniparty*19:12
maliniI have no clue what that is19:12
kgriffswe could stand up jenkins or something and have it run independent, that could do not just mongo driver test, but perf and stuff19:12
maliniBut I will explore that option as well19:12
kgriffssounds like a plan19:12
maliniCeilometer does not run against mongo at the gate19:12
kgriffsmalini: do they run Redis in the gate?19:12
*** mpanetta has quit IRC19:13
kgriffsare we going to be constrained to just sqlite? that seems silly.19:13
maliniI dont know - but will chk tht as well19:13
*** mpanetta has joined #openstack-marconi19:13
kgriffskk19:13
maliniIf redis comes default in a distro, they probably will run it at the gate19:13
kgriffsby "distro" do you mean Linux distro or OpenStack distro?19:14
maliniLinux distro19:14
kgriffsahhh19:14
malinitht's why we are stuct at Mongo 2.0.419:14
malini:|19:14
maliniI saw your comment in the mysql devstack bug19:15
kgriffsok, well, you can make it clear we aren't sure about Redis, but we are looking at several backends and need to be able to run at least one in the gate19:15
kgriffsmalini: thanks for looking into this!19:17
kgriffsmalini: you might also ask for an ETA on when the distro(s) will be upgrading to a new mongo version19:18
malinikgriffs: any time :)19:18
maliniI also saw the discussion with falper87 on forking the process19:18
maliniWe can export env variables in devstack19:18
kgriffsmaybe upgrading isn't the right term. More like, is there an ETA on when the older distro version will no longer be supported for openstack deployments19:18
kgriffsmalini: great, I was hoping we could19:19
maliniwe should probably ask RAX to stop supporting 12.0419:19
kgriffsflaper87: is there a bug filed for the forking issue?19:19
malinitht might be easier ;)19:19
kgriffsheh19:19
kgriffsmalini: updated https://bugs.launchpad.net/devstack/+bug/129213619:22
maliniThird part testing might be our only option if we ever want to run Tsung tests http://ci.openstack.org/third_party.html#testing-your-ci-setup19:27
maliniwrong (sort of) link19:28
malinihttp://ci.openstack.org/third_party.html19:28
kgriffsoh, cool19:30
kgriffsthat looks like just the thing we need19:30
kgriffsit can even vote once it is stable. nice!19:30
malini& We can also test against newer mongo..woohoo19:31
kgriffsmalini: would you mind giving the team an overview of this in the next mtg?19:31
maliniI just found that link :-P19:31
kgriffsI just want to raise awareness.19:32
kgriffswhen we have a more concrete proposal, that can go to the ML for discussion19:32
maliniI will figure out how we can use that & am sure mpanetta, oz_akan_ would LOVE to setup jenkins for us ;)19:32
oz_akan_:)19:33
kgriffsmalini: I can help set up too19:33
kgriffswe should maybe look at some non-jenkins options too19:33
kgriffsdrone.io or something, just to do our part in helping the community experiment with alternates19:34
maliniwe sure give the community lot of alternates ;)19:34
kgriffsheh19:36
kgriffsI mean the broader community, not just OpenStack19:36
kgriffs:p19:36
kgriffsmalini: actually, there is a proposal to write a replacement for jenkins19:36
kgriffsnot sure how I feel about that19:37
kgriffsmalini: https://github.com/drone/drone19:39
kgriffsthey work out of the SF geekdom space iirc19:39
kgriffsmalini: http://docs.drone.io/databases.html#mongodb19:44
kgriffsI assume that "just works" because they have mongo in their docker image19:44
kgriffsI wonder what version of mongo?19:44
kgriffsand whether you can add custom images? I'm sure you can if you self-host drone.io19:45
flaper87kgriffs: mmh, I think there is one19:55
flaper87not sure19:55
*** jergerber has joined #openstack-marconi19:58
*** tedross has quit IRC20:00
*** jcru has joined #openstack-marconi20:04
*** saju_m has quit IRC20:11
*** tedross has joined #openstack-marconi20:15
*** malini is now known as malini_afk20:20
*** oz_akan_ has quit IRC20:21
*** saju_m has joined #openstack-marconi20:29
*** alcabrera has quit IRC20:31
*** balajiiyer has quit IRC20:32
*** oz_akan_ has joined #openstack-marconi20:35
*** saju_m has quit IRC20:37
*** sriram has quit IRC20:39
*** mpanetta has quit IRC21:13
*** oz_akan_ has quit IRC21:16
*** Kuwagata has quit IRC21:31
*** zyuan has quit IRC21:44
*** zyuan has joined #openstack-marconi21:45
*** zyuan_ has joined #openstack-marconi21:49
*** zyuan_ has quit IRC21:49
*** megan_w has quit IRC21:53
*** kgriffs_afk has joined #openstack-marconi21:59
*** amitgandhi has quit IRC22:14
*** mwagner_lap has quit IRC22:17
*** zyuan has quit IRC22:20
*** zyuan has joined #openstack-marconi22:21
*** zyuan has quit IRC22:21
*** zyuan has joined #openstack-marconi22:22
*** zyuan has quit IRC22:22
*** zyuan has joined #openstack-marconi22:22
*** tedross has quit IRC22:23
*** jergerber has quit IRC22:28
*** whenry has quit IRC22:33
*** jmckind has quit IRC22:37
*** tedross has joined #openstack-marconi22:50
*** mpanetta has joined #openstack-marconi23:01
*** tedross has quit IRC23:17
*** flaper87 is now known as flaper87|afk23:18
*** mkoderer has quit IRC23:21
*** mwagner_lap has joined #openstack-marconi23:38
*** mpanetta has quit IRC23:52
*** vkmc has quit IRC23:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!