| *** furlongm has quit IRC | 00:27 | |
| *** furlongm has joined #openstack-performance | 00:28 | |
| *** harlowja_at_home has joined #openstack-performance | 02:39 | |
| *** harlowja_at_home has quit IRC | 02:54 | |
| *** dimtruck is now known as zz_dimtruck | 03:22 | |
| *** itsuugo has quit IRC | 03:46 | |
| *** itsuugo has joined #openstack-performance | 03:49 | |
| *** luzC has quit IRC | 03:50 | |
| *** luzC has joined #openstack-performance | 03:51 | |
| *** serverascode_ has joined #openstack-performance | 05:44 | |
| *** meteorfox_ has joined #openstack-performance | 05:44 | |
| *** afranc has quit IRC | 05:48 | |
| *** bapalm has quit IRC | 05:48 | |
| *** meteorfox has quit IRC | 05:48 | |
| *** serverascode has quit IRC | 05:48 | |
| *** bapalm_ has joined #openstack-performance | 05:48 | |
| *** afranc has joined #openstack-performance | 05:49 | |
| *** meteorfox_ is now known as meteorfox | 05:49 | |
| *** serverascode_ is now known as serverascode | 05:51 | |
| *** pcaruana has joined #openstack-performance | 06:46 | |
| *** f13o has joined #openstack-performance | 07:00 | |
| *** rmart04 has joined #openstack-performance | 08:13 | |
| *** f13o has quit IRC | 08:18 | |
| *** f13o has joined #openstack-performance | 08:36 | |
| *** v12aml has quit IRC | 08:46 | |
| *** v12aml has joined #openstack-performance | 08:47 | |
| *** msimonin has joined #openstack-performance | 08:54 | |
| *** permalac has quit IRC | 09:02 | |
| *** v12aml has quit IRC | 09:52 | |
| *** v12aml has joined #openstack-performance | 09:53 | |
| *** f13o has quit IRC | 10:08 | |
| *** msimonin1 has joined #openstack-performance | 10:12 | |
| *** msimonin has quit IRC | 10:12 | |
| *** msimonin1 has quit IRC | 11:12 | |
| *** msimonin has joined #openstack-performance | 11:12 | |
| *** f13o has joined #openstack-performance | 11:19 | |
| *** bvanhav has joined #openstack-performance | 11:28 | |
| *** rfolco has joined #openstack-performance | 12:23 | |
| *** zz_dimtruck is now known as dimtruck | 12:34 | |
| *** f13o has quit IRC | 12:55 | |
| *** f13o has joined #openstack-performance | 13:16 | |
| *** dimtruck is now known as zz_dimtruck | 13:21 | |
| *** zz_dimtruck is now known as dimtruck | 13:56 | |
| *** m3m0 has left #openstack-performance | 13:56 | |
| *** doffm_ is now known as doffm | 14:00 | |
| *** msimonin has left #openstack-performance | 14:39 | |
| *** f13o has quit IRC | 15:19 | |
| *** catintheroof has joined #openstack-performance | 15:27 | |
| *** hockeynut has quit IRC | 15:34 | |
| *** dimtruck is now known as zz_dimtruck | 15:41 | |
| *** zz_dimtruck is now known as dimtruck | 15:42 | |
| *** msimonin1 has joined #openstack-performance | 15:45 | |
| *** hockeynut has joined #openstack-performance | 16:00 | |
| *** hockeynut has quit IRC | 16:00 | |
| *** hockeynut has joined #openstack-performance | 16:00 | |
| *** hockeynut has quit IRC | 16:01 | |
| *** hockeynut has joined #openstack-performance | 16:02 | |
| openstackgerrit | Dina Belova proposed openstack/performance-docs: Move all test plans to **ready** state (1.0 version) https://review.openstack.org/354188 | 16:11 |
|---|---|---|
| *** msimonin1 has quit IRC | 16:20 | |
| *** rmart04 has quit IRC | 16:25 | |
| *** bwallis has quit IRC | 16:58 | |
| openstackgerrit | Dina Belova proposed openstack/performance-docs: Add plan<->result references everywhere https://review.openstack.org/354251 | 17:10 |
| openstackgerrit | Dina Belova proposed openstack/performance-docs: Add plan<->result references everywhere https://review.openstack.org/354251 | 17:12 |
| openstackgerrit | Dina Belova proposed openstack/performance-docs: Add correct plan<->result references everywhere https://review.openstack.org/354251 | 17:13 |
| *** catintheroof has quit IRC | 17:30 | |
| bvanhav | DinaBelova: I am currently trying to use OSProfiler with Nova (I know it's not integrated yet, but I locally pulled in the changes to Nova and novaclient) and it seems to partially work. In the profiler output I see the calls to the Nova rest APIs and calls to compute.api, but not to any of the rpc apis, is there something else I need to do to make that work? | 17:32 |
| DinaBelova | bvanhav - can you please share the trace.html somehow to me? frankly speaking I was mostly focused on performance docs last month and did not track the latest rebases on this commit.. | 17:34 |
| bvanhav | Just double checking, the output of the osprofiler command? | 17:36 |
| *** bwallis has joined #openstack-performance | 17:36 | |
| DinaBelova | yep, output in the html | 17:37 |
| bvanhav | DinaBelova: http://paste.openstack.org/show/EbxJqJ9JQ2j9kVkOgdlq/ | 17:40 |
| DinaBelova | bvanhav you copied only part of the file :) html itself is missing, that's JS | 17:48 |
| DinaBelova | with the part of the data | 17:49 |
| DinaBelova | you can just upload html output as a file it to the google drive or whatever | 17:49 |
| bvanhav | Yeah... the entire file is over 3400 lines long... and pastebin is truncating it at 1800 | 17:51 |
| DinaBelova | bvanhav just upload html file itself to somewhere I can download it from | 17:51 |
| bvanhav | Yep, working on it | 17:52 |
| bvanhav | DinaBelova: try https://drive.google.com/file/d/0B4YQonUprQV4cDdoMERFN3VURW8/view?usp=sharing | 17:53 |
| DinaBelova | yep, working | 17:53 |
| * DinaBelova looking | 17:53 | |
| DinaBelova | bvanhav was it VM show request? | 17:54 |
| bvanhav | It was actually a rally run that tried to setup and delete a handful of servers | 17:55 |
| DinaBelova | so what kind of request to nova it was? nova boot or nova delete? | 17:56 |
| DinaBelova | different commands will get different trace ids | 17:56 |
| DinaBelova | this particular request looks like grabbing nova vm details | 17:57 |
| DinaBelova | before killing it | 17:57 |
| DinaBelova | do you have an opportunity to identify what nova api call it was? | 17:58 |
| bvanhav | I think it's a details get | 17:58 |
| DinaBelova | yep, it looks like | 17:58 |
| DinaBelova | bvanhav I'm asking because nova uses rpc to communicate with nova conductor and nova-computes on the remote hosts, for instance. for vm info request it may communicate directly to the local hypervisor wrapper (so you see osapi_compute service in the trace, not the nova-compute) | 18:03 |
| DinaBelova | so it may be normal not to observe rpc type of calls | 18:06 |
| bvanhav | Ok, I'm going to be trying a straight boot (skipping rally) and see what I see | 18:07 |
| DinaBelova | bvanhav ack | 18:07 |
| *** hockeynut has quit IRC | 18:09 | |
| *** hockeynut has joined #openstack-performance | 18:11 | |
| bvanhav | DinaBelova: on a straight boot rpc came through just fine, says it's something I setup wrong with rally... thank you for the help | 18:13 |
| DinaBelova | bvanhav ack, np :) | 18:14 |
| doffm | DinaBelova: So it looks to me like bvanhav has things working as expected. | 18:22 |
| doffm | Single requests made through novaclient seem to create good trace results. | 18:22 |
| doffm | I guess we were kind of hoping that we could get a single trace for an entire rally scenario. | 18:23 |
| doffm | Which would require us to pass a parent trace ID to novaclient. | 18:24 |
| DinaBelova | doffm it looks like yes.. the thing is that from nova perspective (on server side) every api call to it is separated, and that's the reason osprofiler treat is as separated call (as in fact it's just lib used inside nova). and rally (as a client) is just doing bunch of requests to variour services | 18:30 |
| DinaBelova | so to allow scenario-size profiling some modifications needs to be done (e.g. opportunity to pass some synthetic super trace id, that's generated in rally itself) | 18:31 |
| *** catintheroof has joined #openstack-performance | 18:34 | |
| DinaBelova | doffm i think this might be workaround via initializing profiler in rally itself, so rally will be using it to generate the root trace id and pass it everywhere (to nova in our case) | 18:35 |
| DinaBelova | this seems to be workable solution to me | 18:35 |
| doffm | DinaBelova: bvanhav has already initialized the profier in the rally 'task' class and generated a root trace id. Its just detached from the rest of the traced. | 18:38 |
| doffm | *traces | 18:38 |
| doffm | So yeah, we will need to pass it to all of the novaclient calls somehow. | 18:38 |
| doffm | Fun. :) | 18:38 |
| DinaBelova | :D heh, yeah, sound challenging... | 18:39 |
| *** msimonin has joined #openstack-performance | 18:44 | |
| *** catintheroof has quit IRC | 19:28 | |
| *** f13o has joined #openstack-performance | 19:37 | |
| *** bvanhav has quit IRC | 19:42 | |
| *** msimonin has quit IRC | 19:48 | |
| *** catintheroof has joined #openstack-performance | 20:00 | |
| *** msimonin has joined #openstack-performance | 20:37 | |
| *** msimonin has quit IRC | 20:42 | |
| *** f13o has quit IRC | 20:59 | |
| *** bwallis has quit IRC | 20:59 | |
| *** bwallis has joined #openstack-performance | 21:16 | |
| *** msimonin has joined #openstack-performance | 21:38 | |
| *** msimonin has left #openstack-performance | 21:38 | |
| *** hockeynut has quit IRC | 21:38 | |
| *** dimtruck is now known as zz_dimtruck | 22:07 | |
| *** msimonin1 has joined #openstack-performance | 22:27 | |
| *** catintheroof has quit IRC | 22:35 | |
| *** bwallis has quit IRC | 23:20 | |
| *** bwallis has joined #openstack-performance | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!