| *** zaitcev has quit IRC | 00:23 | |
| *** zaitcev has joined #openstack-dev | 00:29 | |
| *** AXiS_SharK has joined #openstack-dev | 00:45 | |
| *** jdurgin has quit IRC | 00:59 | |
| *** jwilmes has quit IRC | 01:06 | |
| *** AXiS_SharK has quit IRC | 01:10 | |
| *** AXiS_SharK has joined #openstack-dev | 01:12 | |
| *** AXiS_SharK has quit IRC | 02:15 | |
| *** AXiS_SharK has joined #openstack-dev | 02:16 | |
| *** AXiS_SharK has quit IRC | 02:21 | |
| *** zaitcev has quit IRC | 02:59 | |
| *** zaitcev has joined #openstack-dev | 02:59 | |
| HugoKuo | Still Thinkink about the solution of nova-network :< , any idea? | 03:24 |
|---|---|---|
| *** jbryce has quit IRC | 03:46 | |
| *** Binbinaway has quit IRC | 03:55 | |
| *** Binbinaway has joined #openstack-dev | 03:55 | |
| *** Zangetsue has joined #openstack-dev | 04:28 | |
| *** Zangetsue has left #openstack-dev | 04:34 | |
| *** Zangetsue has joined #openstack-dev | 04:34 | |
| *** AXiS_SharK has joined #openstack-dev | 04:34 | |
| *** AXiS_SharK has quit IRC | 04:47 | |
| *** BK_man has joined #openstack-dev | 04:54 | |
| *** AXiS_SharK has joined #openstack-dev | 05:47 | |
| *** AXiS_SharK has quit IRC | 06:57 | |
| *** Eyk has joined #openstack-dev | 08:11 | |
| *** cloudgroups has joined #openstack-dev | 08:12 | |
| *** cloudgroups has left #openstack-dev | 08:20 | |
| *** Binbinaway has quit IRC | 09:21 | |
| *** Eyk has quit IRC | 12:03 | |
| *** markvoelker has joined #openstack-dev | 12:17 | |
| *** namaqua has joined #openstack-dev | 12:33 | |
| *** lorin1 has joined #openstack-dev | 12:59 | |
| *** namaqua has quit IRC | 13:00 | |
| *** dprince has joined #openstack-dev | 13:17 | |
| *** Zangetsue_ has joined #openstack-dev | 13:27 | |
| *** Zangetsue has quit IRC | 13:28 | |
| *** Zangetsue_ is now known as Zangetsue | 13:28 | |
| *** cloudgroups has joined #openstack-dev | 13:31 | |
| *** jaypipes has joined #openstack-dev | 13:39 | |
| *** cp16net has joined #openstack-dev | 13:57 | |
| *** Zangetsue has quit IRC | 14:03 | |
| *** cp16net has quit IRC | 14:08 | |
| *** AXiS_SharK has joined #openstack-dev | 14:27 | |
| *** AXiS_SharK has quit IRC | 14:31 | |
| *** cloudgroups has left #openstack-dev | 14:32 | |
| *** jkoelker has joined #openstack-dev | 14:36 | |
| *** AXiS_SharK has joined #openstack-dev | 14:43 | |
| sandywalsh | ttx soren sorry, got zip done on dev review duties yesterday. I'll try and make up for it over the next few days. | 14:45 |
| *** dragondm has joined #openstack-dev | 14:49 | |
| vishy | soren: ping | 15:12 |
| vishy | tr3buchet: ping | 15:12 |
| comstud | i need soren too | 15:15 |
| comstud | :) | 15:15 |
| comstud | i have this eventlet patch | 15:15 |
| comstud | for 771512 | 15:15 |
| *** cweidenkeller has joined #openstack-dev | 15:20 | |
| vishy | comstud: I looked at that patch, have you heard at all from the eventlet maintainer? | 15:21 |
| comstud | the convo on the thread for the bug soren filed has stalled | 15:21 |
| vishy | comstud: have you asked redbo to look at it? | 15:22 |
| comstud | he only had a couple comments anyway | 15:22 |
| vishy | he's done some eventlet patching as well | 15:22 |
| comstud | no, i haven't. ok. | 15:23 |
| comstud | i've never talked to redbo | 15:23 |
| *** cweidenkeller has quit IRC | 15:25 | |
| comstud | i'd been stuck in agent land too long, so there's a number of people i don't know | 15:26 |
| comstud | happy to be able to focus on nova now :) | 15:26 |
| vishy | comstud: redbo is on the swift team | 15:27 |
| comstud | ah, mike barton | 15:29 |
| comstud | cools, will ping | 15:29 |
| vishy | comstud: also, notmyname just suggested taking it over to #eventlet to see if anyone there has suggestions | 15:30 |
| comstud | good idea | 15:31 |
| comstud | i think our real issue here is larger than this eventlet problem... | 15:32 |
| comstud | even if this eventlet problem is real | 15:32 |
| jk0 | vishy: pushed your changes | 15:32 |
| *** cweidenkeller has joined #openstack-dev | 15:33 | |
| comstud | so, i'd like to see eventlet get patched.. but at the same time, i have some changes to make to API | 15:33 |
| comstud | playing around with this more today | 15:33 |
| *** dprince has quit IRC | 15:34 | |
| vishy | comstud: sounds good | 15:34 |
| *** rnirmal has joined #openstack-dev | 15:37 | |
| vishy | jk0: seems fine, but I have a question | 15:39 |
| vishy | jk0: what is making the code block? | 15:40 |
| jk0 | comstud might be able to answer that better than I can | 15:41 |
| jk0 | it cropped up when running our tests (50-100 simultaneous builds) | 15:41 |
| vishy | comstud: ?? | 15:42 |
| jk0 | I can't remember what the exception was but comstud might | 15:42 |
| vishy | jk0: the time.sleep should pause the greenthread and allow other greenthreads to move along | 15:42 |
| jk0 | I think the problem was _find_host() happens before the cast | 15:43 |
| vishy | sure, but there is a sleep in _find_host | 15:44 |
| vishy | very odd | 15:44 |
| jk0 | yeah.. I can't remember what it was | 15:44 |
| vishy | ok maybe when comstud comes back he can enlighten me | 15:44 |
| jk0 | yeah, I think the concern was API keeping the connection open until that was casted, but in some cases it could take up to 10-30 seconds | 15:45 |
| *** AXiS_SharK has quit IRC | 15:45 | |
| *** rnirmal has quit IRC | 15:46 | |
| comstud | i don't have a real answer | 15:46 |
| vishy | i see. But it will still take 10-30 seconds no? | 15:46 |
| vishy | it will just be in a new greenthread | 15:46 |
| comstud | i think we were just stressing the system enough... | 15:46 |
| comstud | that it was taking a while for the DB to get updated with the compute node that was picked | 15:47 |
| comstud | for a particular instance | 15:47 |
| comstud | yes | 15:47 |
| vishy | that explains the timeout, so the thing that is actually fixing it is raising the timeout to 30 | 15:47 |
| vishy | spawning a new greenthread seems unnecessary in that case, no? | 15:48 |
| comstud | i'd prefer not to make an API client user block unnecessarily | 15:48 |
| comstud | if the event doesn't need to wait | 15:48 |
| comstud | but | 15:48 |
| vishy | comstud: but it isn't | 15:48 |
| vishy | comstud: all the time is spent in monkey_patched calls | 15:48 |
| vishy | so it isn't blocking | 15:48 |
| comstud | yeah, i get that. that's not what i meant... | 15:49 |
| comstud | take our control panel for instance... | 15:49 |
| comstud | the call for build is going to sit there and wait | 15:49 |
| comstud | until set_admin_password finds the host | 15:49 |
| vishy | ah i see | 15:49 |
| comstud | if for whatever reason that takes 20 seconds... | 15:49 |
| comstud | yea | 15:50 |
| vishy | ok so the purpose of the greenthread is so that the call can return immediately | 15:50 |
| vishy | makes sense now | 15:50 |
| comstud | yeah | 15:50 |
| vishy | thx | 15:50 |
| comstud | there's no data being returned that's dependent on that finishing first | 15:50 |
| comstud | i don't necessarily like that there's a 2nd cast to set the password, though. | 15:51 |
| comstud | long term, i think compute should handle this when it sees a spawn request | 15:52 |
| *** rnirmal has joined #openstack-dev | 15:52 | |
| vishy | comstud: yes, perhaps that should all be moved down into compute.manager | 15:52 |
| comstud | that would kill this problem overall, and I think just makes more sense | 15:53 |
| comstud | but anyway, the current fix works for now to get us to 50 builds at once | 15:54 |
| comstud | i need 766404 reviewed again | 15:54 |
| comstud | that's another important one related to this | 15:54 |
| comstud | https://code.launchpad.net/~cbehrens/nova/lp766404/+merge/60543 | 15:55 |
| *** tr3buchet has joined #openstack-dev | 16:03 | |
| vishy | comstud: approved | 16:16 |
| comstud | thanks! | 16:16 |
| dragondm | hmm... Random thought: Shouldn't openstackjenkins be relaying project build statuses here in -dev instead of #openstack ? | 16:19 |
| notmyname | yes | 16:19 |
| *** notmyname has left #openstack-dev | 16:20 | |
| *** notmyname has joined #openstack-dev | 16:20 | |
| *** ChanServ sets mode: +v notmyname | 16:20 | |
| *** bcwaldon has joined #openstack-dev | 16:37 | |
| vishy | comstud: your patch got eaten by the pep monster | 16:41 |
| comstud | strange, i had checked that it passed | 16:42 |
| comstud | i'll take a look | 16:42 |
| comstud | or maybe i didn't after i wrote the test | 16:45 |
| comstud | hehe | 16:45 |
| comstud | ok | 16:45 |
| comstud | *waits for LP to update* | 16:47 |
| comstud | ok | 16:47 |
| *** BK_man has quit IRC | 16:48 | |
| creiht | using launchpad takes me back to the days when I used to have to wait for OS/2 to boot | 16:48 |
| comstud | hehe | 16:48 |
| comstud | vish: updated | 16:49 |
| jk0 | comstud: approved | 16:50 |
| comstud | thnx | 16:51 |
| jk0 | you betcha~ | 16:51 |
| *** jdurgin has joined #openstack-dev | 16:58 | |
| *** BK_man has joined #openstack-dev | 16:58 | |
| *** mszilagyi has joined #openstack-dev | 16:58 | |
| BK_man | vishy: hi! any special configuration for network VlanManager while several hosts? Can't ping instances on compute nodes, on CC - it's ok | 17:03 |
| vishy | BK_man: you have to turn on vlan trunking on your switch | 17:03 |
| vishy | for the vlans you are using | 17:03 |
| vishy | er lets take this back to openstack channel | 17:04 |
| BK_man | vishy: is it documented on wiki or I missed something? | 17:04 |
| *** mszilagyi has quit IRC | 17:04 | |
| *** rnirmal has quit IRC | 17:04 | |
| *** dragondm has quit IRC | 17:07 | |
| *** mattray has joined #openstack-dev | 17:16 | |
| *** zaitcev has quit IRC | 17:29 | |
| *** dprince has joined #openstack-dev | 17:47 | |
| *** dprince has quit IRC | 17:56 | |
| soren | vishy, comstud: Wazzup? | 18:25 |
| soren | comstud: How about we take your eventlet patch and stick an updated eventlet in the PPA and see how it goes? | 18:26 |
| comstud | that'd be cool | 18:27 |
| lorin1 | Possibly stupid question: I'm trying to figure out what code actually invokes ComputeManager.run_instance. I see how a compute node gets scheduled in SimpleScheduler.schedule_run_instnace, but what code actually invokes the run_instance method? | 18:27 |
| comstud | soren: did you look it over completely? i think it makes sense. | 18:27 |
| soren | comstud: Haven't had the surplus brainpower to review it in detail, no. | 18:28 |
| soren | comstud: The approach seems sound, though. | 18:29 |
| comstud | no worries | 18:29 |
| comstud | been exercising it pretty heavily | 18:29 |
| comstud | so i think it's sound | 18:30 |
| soren | I do wish upstream was more interested :-/ | 18:32 |
| comstud | maybe if i do a real branch and do a pull request it'll get more attention | 18:47 |
| comstud | what version is the current ppa based off of? | 18:48 |
| comstud | i guess 0.9.14 | 18:49 |
| comstud | there's a 0.9.15 now | 18:49 |
| soren | I forget. | 18:49 |
| soren | Yeah, 0.9.14 sounds right. | 18:49 |
| soren | comstud: There's a good chance I won't get to it until next week | 18:50 |
| * soren adds to TODO list | 18:50 | |
| comstud | no worries | 18:53 |
| *** dragondm has joined #openstack-dev | 18:56 | |
| *** markvoelker has quit IRC | 19:04 | |
| vishy | soren: i was pinging you about the networking code | 19:38 |
| *** pyhole has joined #openstack-dev | 19:40 | |
| jk0 | lp778678 | 19:40 |
| soren | vishy: Ah. That I have even less mental surplus for. :( | 19:40 |
| vishy | soren: ok | 19:41 |
| jk0 | lp778678 | 19:41 |
| pyhole | LP Bug #778678: nova.virt.xenapi.vmops _run_ssl() should write directly to stdin instead of file [Status: Triaged, Assignee: None] https://bugs.launchpad.net/bugs/778678 | 19:41 |
| jk0 | .lbugs | 19:42 |
| pyhole | Launchpad bugs for a team (ex: .lbugs <project> <team>|<user>) | 19:42 |
| jk0 | .lbugs nova soren | 19:42 |
| * jk0 kicks eventlet | 19:43 | |
| *** pyhole has quit IRC | 19:46 | |
| *** pyhole has joined #openstack-dev | 19:46 | |
| *** pyhole has quit IRC | 20:20 | |
| *** pyhole has joined #openstack-dev | 20:20 | |
| jk0 | k, fixed | 20:21 |
| jk0 | .lbugs nova soren | 20:21 |
| pyhole | No bugs found for Soren Hansen | 20:21 |
| jk0 | .lbugs nova jk0 | 20:21 |
| pyhole | LP Bug #767545 in OpenStack Compute (nova): "Password reset not ran on instance create in XenServer" [Assignee: Josh Kearney] https://bugs.launchpad.net/nova/+bug/767545 | 20:21 |
| pyhole | LP Bug #754051 in OpenStack Compute (nova): "instance state polling finds compute instance not in DB" [Assignee: Josh Kearney] https://bugs.launchpad.net/nova/+bug/754051 | 20:21 |
| pyhole | LP Bug #754944 in OpenStack Compute (nova): "self.interfaces_xml is assigned in LibvirtConnection but never used" [Assignee: Josh Kearney] https://bugs.launchpad.net/nova/+bug/754944 | 20:21 |
| pyhole | LP Bug #781435 in OpenStack Compute (nova): "50 builds at once sometimes results in 'Unable to find host for Instance #'" [Assignee: Josh Kearney] https://bugs.launchpad.net/nova/+bug/781435 | 20:21 |
| vishy | sandywalsh: ping | 20:21 |
| jk0 | .lbugs nova nova-core | 20:21 |
| pyhole | LP Bug #740576 in OpenStack Compute (nova): "No XML namespace in OpenStack API" [Assignee: justinsb] https://bugs.launchpad.net/nova/+bug/740576 | 20:21 |
| pyhole | LP Bug #744150 in OpenStack Compute (nova): "unit tests fail if nova-api is running on the system" [Assignee: justinsb] https://bugs.launchpad.net/nova/+bug/744150 | 20:21 |
| pyhole | LP Bug #731448 in OpenStack Compute (nova): "Image Format Instability" [Assignee: justinsb] https://bugs.launchpad.net/nova/+bug/731448 | 20:21 |
| pyhole | LP Bug #744224 in OpenStack Compute (nova): "Zones needs to support ancillary server functions" [Assignee: Sandy Walsh] https://bugs.launchpad.net/nova/+bug/744224 | 20:21 |
| pyhole | LP Bug #744217 in OpenStack Compute (nova): "Zones needs to support servers.get_all" [Assignee: Sandy Walsh] https://bugs.launchpad.net/nova/+bug/744217 | 20:21 |
| pyhole | LP Bug #781878 in OpenStack Compute (nova): "Error during report_driver_status(): host_state_interval" [Assignee: Sandy Walsh] https://bugs.launchpad.net/nova/+bug/781878 | 20:21 |
| pyhole | [...] truncated last 16 users | 20:21 |
| pvo | thanks jk0. handy. | 20:25 |
| jk0 | no problem | 20:25 |
| *** bcwaldon has quit IRC | 20:29 | |
| *** lorin1 has quit IRC | 20:46 | |
| *** AXiS_SharK has joined #openstack-dev | 21:10 | |
| *** mszilagyi has joined #openstack-dev | 21:38 | |
| gholt | mtaylor: Hmm, seems my posts aren't the only ones you get stuck as the author of on the planet openstack feed. Weird. | 21:57 |
| *** AXiS_SharK has quit IRC | 22:20 | |
| *** AXiS_SharK has joined #openstack-dev | 22:23 | |
| *** Eyk has joined #openstack-dev | 22:50 | |
| *** bcwaldon has joined #openstack-dev | 22:52 | |
| *** mattray has quit IRC | 22:53 | |
| *** dragondm has quit IRC | 22:53 | |
| BK_man | .lbugs abrindeyev | 22:54 |
| BK_man | .lbugs nova abrindeyev | 22:54 |
| pyhole | No bugs found for Andrey Brindeyev | 22:54 |
| *** jkoelker has quit IRC | 22:58 | |
| *** mszilagyi has quit IRC | 22:58 | |
| *** mszilagyi has joined #openstack-dev | 23:05 | |
| *** bcwaldon has quit IRC | 23:08 | |
| *** cloudgroups has joined #openstack-dev | 23:44 | |
| *** AXiS_SharK has left #openstack-dev | 23:48 | |
| *** ohnoimdead has joined #openstack-dev | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!