| *** dklyle_ has joined #openstack-placement | 01:27 | |
| *** david-lyle has quit IRC | 01:31 | |
| *** bhagyashris has joined #openstack-placement | 01:41 | |
| *** david-lyle has joined #openstack-placement | 02:12 | |
| *** dklyle_ has quit IRC | 02:15 | |
| *** ttsiouts has quit IRC | 03:01 | |
| *** ttsiouts has joined #openstack-placement | 03:02 | |
| *** ttsiouts has quit IRC | 03:07 | |
| openstackgerrit | Sam Morrison proposed openstack/nova-specs master: add spec "flavor-classes" https://review.openstack.org/635408 | 03:52 |
|---|---|---|
| *** tetsuro has joined #openstack-placement | 05:16 | |
| *** e0ne has joined #openstack-placement | 06:10 | |
| *** tetsuro has quit IRC | 06:29 | |
| openstackgerrit | Takashi NATSUME proposed openstack/nova-specs master: Fix warnings in the document generation https://review.openstack.org/631150 | 07:02 |
| *** e0ne has quit IRC | 07:10 | |
| *** efried has quit IRC | 07:11 | |
| *** efried has joined #openstack-placement | 07:12 | |
| *** e0ne has joined #openstack-placement | 07:20 | |
| *** e0ne has quit IRC | 07:31 | |
| openstackgerrit | Takashi NATSUME proposed openstack/nova-specs master: Add more detail and fix code block formatting https://review.openstack.org/628702 | 07:37 |
| *** helenafm has joined #openstack-placement | 08:25 | |
| *** tetsuro has joined #openstack-placement | 09:00 | |
| *** tssurya has joined #openstack-placement | 09:06 | |
| *** ttsiouts has joined #openstack-placement | 09:25 | |
| *** takashin has left #openstack-placement | 09:30 | |
| *** bhagyashris has quit IRC | 09:41 | |
| *** cdent has joined #openstack-placement | 10:00 | |
| *** e0ne has joined #openstack-placement | 10:03 | |
| *** alex_xu has quit IRC | 10:24 | |
| *** tetsuro has quit IRC | 11:52 | |
| *** e0ne has quit IRC | 11:56 | |
| *** e0ne has joined #openstack-placement | 11:57 | |
| *** gibi_off is now known as gibi | 12:34 | |
| efried | o/ gibi, how are you feeling? | 12:45 |
| gibi | efried: thanks for the question. I'm fell pretty OK. I still have some visits planned to the hostpital, e.g. to remove the stitches | 12:50 |
| *** mriedem has joined #openstack-placement | 13:21 | |
| *** takashin has joined #openstack-placement | 13:29 | |
| *** irclogbot_2 has joined #openstack-placement | 14:29 | |
| *** david-lyle has quit IRC | 14:57 | |
| *** takashin has left #openstack-placement | 15:01 | |
| *** purplerbot has quit IRC | 15:11 | |
| *** purplerbot has joined #openstack-placement | 15:13 | |
| *** dansmith has quit IRC | 15:13 | |
| *** dansmith has joined #openstack-placement | 15:14 | |
| *** ttsiouts has quit IRC | 16:07 | |
| *** ttsiouts has joined #openstack-placement | 16:15 | |
| *** rubasov has quit IRC | 16:37 | |
| *** tssurya has quit IRC | 16:50 | |
| sean-k-mooney | so placement is failing to be deployed for me on master with this error in the placement api journal http://paste.openstack.org/show/744692/ | 16:52 |
| sean-k-mooney | is this new? | 16:53 |
| efried | cdent: ^ ?? | 16:53 |
| efried | sounds like maybe a lib missing or out of date | 16:53 |
| * cdent reads | 16:53 | |
| cdent | sean-k-mooney: a) how did you install uwsgi? b) what's after that in the log | 16:54 |
| sean-k-mooney | ya perhaps but this is my srivo setup so it had openstack deployed but it the first time i have powered it up in 2 weeks | 16:54 |
| sean-k-mooney | ill post the full log its short | 16:54 |
| sean-k-mooney | and devstack installed it | 16:54 |
| sean-k-mooney | this is on centos 7 | 16:55 |
| cdent | in any case placement is not responsible for installing or managing the uwsgi install, devstack is, so it might be that | 16:55 |
| cdent | if centos has changed out from under you it might be that uwsgi is not packaging python support by default | 16:55 |
| cdent | _also_ that error message is possible when python is built in and it is still trying to load the plugin, even though it doesn't need it | 16:55 |
| cdent | (which is why I want the rest of the log) | 16:56 |
| sean-k-mooney | http://paste.openstack.org/show/744694/ | 16:56 |
| cdent | sean-k-mooney: if you're seeing that config output, then you've got a working python uwsgi setup, and the service has started | 16:57 |
| cdent | but you should have a bit more log than that | 16:57 |
| sean-k-mooney | yes however you get a 503 from placement if you curl it | 16:58 |
| sean-k-mooney | i hit the paste limit i can output it to a file and throw it on a web server one sec | 16:59 |
| cdent | if you're seeing 503 then it is likely the apache configuration is not right and mod_uwsgi_proxy is over-configured | 17:00 |
| cdent | devstack doesn't tolerate reconfiguration of apache well | 17:00 |
| cdent | I'd suggest looking in the virtual host configurations (not sure of the exact path, cuz it's different on centos for deb-based machines, yeah) | 17:01 |
| cdent | what seems to happen is that too many lines are configured and the right one is ignored | 17:02 |
| sean-k-mooney | well it running under uwsgi so it should not have apache right? | 17:02 |
| cdent | in devstack it goes through apache | 17:03 |
| cdent | apache2 -> mod_proxy_uwsgi -> uwsgi (run by systemd) | 17:03 |
| cdent | I believe that was sdague's architecture | 17:03 |
| sean-k-mooney | cdent: that is the full output of journalctl https://logs.seanmooney.info/manual/placement.log | 17:04 |
| cdent | yeah, the uwsgi process is fine, the problem will be earlier in the stack probably apache2 | 17:05 |
| sean-k-mooney | but ok ill see if i can fid the missing so | 17:05 |
| cdent | you don't need the so you already have python support | 17:05 |
| sean-k-mooney | oh ok | 17:05 |
| cdent | the plugin thing is a red herring and can be ignored. it's just one of uwsgi's pecadillos | 17:05 |
| sean-k-mooney | do you know if there were any devstack changes recently related to this | 17:06 |
| sean-k-mooney | i noticed that things seam to not be running on port anymore | 17:06 |
| sean-k-mooney | most service are subfolders on port 80 | 17:06 |
| cdent | placement hasn't run on a port in a really long time | 17:06 |
| cdent | not sure about other things | 17:06 |
| cdent | i haven't looked at devstack recently | 17:07 |
| sean-k-mooney | at this point neutron was the only thing that was not on port 80 | 17:07 |
| *** e0ne has quit IRC | 17:08 | |
| cdent | sean-k-mooney: if you do discover that it is in the apache config and can see the bug in the devstack code that would likely be a great fix | 17:13 |
| *** helenafm has quit IRC | 17:20 | |
| *** ttsiouts has quit IRC | 17:28 | |
| *** ttsiouts has joined #openstack-placement | 17:29 | |
| *** dklyle has joined #openstack-placement | 17:29 | |
| *** ttsiouts has quit IRC | 17:31 | |
| *** ttsiouts has joined #openstack-placement | 17:32 | |
| *** ttsiouts has quit IRC | 17:32 | |
| *** cdent has quit IRC | 17:52 | |
| *** belmoreira has quit IRC | 18:07 | |
| *** e0ne has joined #openstack-placement | 18:15 | |
| *** e0ne has quit IRC | 18:22 | |
| *** e0ne has joined #openstack-placement | 18:42 | |
| *** ttsiouts has joined #openstack-placement | 19:05 | |
| openstackgerrit | Adam Spiers proposed openstack/os-traits master: Add HW_CPU_AMD_SEV trait https://review.openstack.org/635608 | 19:28 |
| efried | gibi: Can you please remind me how you wound up reconciling which bw allocation corresponds with which request group? | 19:34 |
| *** ttsiouts has quit IRC | 19:51 | |
| mriedem | efried: this? https://review.openstack.org/#/c/616239/ | 19:55 |
| mriedem | that's my next in that series to review, i just haven't mustered the energy | 19:55 |
| *** ttsiouts has joined #openstack-placement | 19:57 | |
| *** ttsiouts has quit IRC | 20:01 | |
| *** e0ne has quit IRC | 21:19 | |
| openstackgerrit | Sam Morrison proposed openstack/nova-specs master: add spec "flavor-classes" https://review.openstack.org/635408 | 21:41 |
| efried | mriedem: Thank you, yes. Okay, so it hasn't merged yet. And also, this sucks, for the client to have to reengineer work that was already done on the placement side. I hope people will see this and start coming around to the idea of providing this mapping in the allocation candidate response payload instead. | 22:40 |
| efried | jaypipes: ^ | 22:41 |
| efried | edleafe: ^ | 22:41 |
| *** mriedem is now known as mriedem_afk | 22:43 | |
| efried | that may have come across as unfairly harsh; the idea has support, just needs to be worked through. We should do that. Device management (cyborg) will need it as well. | 22:44 |
| edleafe | efried: just heading out for the day now. Will look over that tomorrow | 22:45 |
| efried | edleafe: it's really this spec: https://review.openstack.org/#/c/597601/ | 22:46 |
| efried | I'll take the liberty of adding you :P | 22:46 |
| edleafe | ack | 22:46 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!