*** harshs has joined #openstack-performance | 02:27 | |
*** harshs has quit IRC | 03:28 | |
*** harshs has joined #openstack-performance | 03:56 | |
*** harshs has quit IRC | 04:13 | |
*** shunliz has joined #openstack-performance | 04:36 | |
*** arnoldje has quit IRC | 05:30 | |
*** harlowja_at_home has joined #openstack-performance | 06:04 | |
*** harlowja_at_home has quit IRC | 06:32 | |
*** f13o has joined #openstack-performance | 07:09 | |
*** openstack has joined #openstack-performance | 07:18 | |
-cameron.freenode.net- [freenode-info] channel trolls and no channel staff around to help? please check with freenode support: http://freenode.net/faq.shtml#gettinghelp | 07:18 | |
*** zsli_ has joined #openstack-performance | 08:30 | |
*** shunliz has quit IRC | 08:33 | |
*** xek has joined #openstack-performance | 08:46 | |
*** f13o has quit IRC | 09:08 | |
*** f13o has joined #openstack-performance | 09:09 | |
*** f13o has quit IRC | 09:10 | |
*** f13o has joined #openstack-performance | 09:10 | |
*** f13o has quit IRC | 09:16 | |
*** f13o has joined #openstack-performance | 09:28 | |
*** amaretskiy has joined #openstack-performance | 09:39 | |
*** zsli_ has quit IRC | 09:51 | |
*** shunliz has joined #openstack-performance | 12:07 | |
DinaBelova | klindgren_ - I'm going crazy with my little VM where I'm trying to reproduce your issue via lots of metadata calls :) for some reason my metadata service is not working as expected :D debugging :D | 13:26 |
---|---|---|
*** f13o has quit IRC | 13:31 | |
*** mriedem has joined #openstack-performance | 14:38 | |
*** mriedem has quit IRC | 14:52 | |
*** mriedem has joined #openstack-performance | 14:57 | |
*** shunliz has quit IRC | 15:01 | |
*** zz_dimtruck is now known as dimtruck | 15:17 | |
*** arnoldje has joined #openstack-performance | 15:26 | |
*** mdorman has joined #openstack-performance | 15:38 | |
*** harlowja_at_home has joined #openstack-performance | 15:46 | |
*** harshs has joined #openstack-performance | 16:03 | |
klindgren_ | DinaBelova, cool - I htink you are going to need a relative large number of vm's. As many calls to metadata to the same source, will hit the default cache of 15 seconds. So while it will cause some load - its not going to cause (or should not cause) a ton of load on conductor. | 16:09 |
*** dimtruck is now known as zz_dimtruck | 16:22 | |
*** zz_dimtruck is now known as dimtruck | 16:22 | |
*** harshs has quit IRC | 16:43 | |
*** harlowja_at_home has quit IRC | 16:58 | |
*** arnoldje has quit IRC | 17:21 | |
*** amaretskiy has quit IRC | 17:37 | |
*** arnoldje has joined #openstack-performance | 17:42 | |
DinaBelova | klindgren_ - nope, the issue is that I do not see excepted logs at all... it looks like my nova is behaving not very well - I even cannot ssh to the Vms and check the metadata service inside them - nova guys are helping me right now :( | 17:55 |
DinaBelova | expected* logs | 17:56 |
*** harshs has joined #openstack-performance | 17:58 | |
*** Augie has joined #openstack-performance | 18:37 | |
*** Augie has quit IRC | 18:39 | |
SpamapS | DinaBelova: have you considered doing this with the fake driver? | 18:41 |
*** mriedem has left #openstack-performance | 18:42 | |
*** mriedem has joined #openstack-performance | 18:43 | |
SpamapS | DinaBelova: I'm not sure how to fake out the mac<->port<->instance mapping, but then at least you can have thousands of vms on a laptop. :) | 18:43 |
*** boris-42 has quit IRC | 18:48 | |
harlowja | right, thats the question i have, fake driver should be fine, but networking might not be (and if vm <-> metadata calls are needed then might need a real vm/something) | 18:57 |
DinaBelova | SpamapS - I just wanted to take a look on real metadata load :( | 19:13 |
DinaBelova | and then emulate it with some scripts for instance | 19:13 |
DinaBelova | or whatever | 19:14 |
DinaBelova | but real load is far from me :D | 19:14 |
harlowja | someday u will get real load | 19:14 |
harlowja | when u are worthy | 19:14 |
harlowja | lol | 19:14 |
harlowja | :) | 19:14 |
SpamapS | DinaBelova: indeed, that first step is a big one. :-/ | 19:14 |
DinaBelova | :D | 19:14 |
SpamapS | I'm reading the bit of the metadata code that interacts with lbaas... | 19:16 |
SpamapS | lots of port and subnet listing | 19:17 |
SpamapS | :-/ | 19:17 |
SpamapS | klindgren_: help me out.. this question started when you had to turn off conductor.use_local right? | 19:22 |
klindgren_ | no - we have been 100% remote conductor since havana | 19:23 |
klindgren_ | it started two fold, one during hte performance meeting in tokyo where I brought up that our conductor stuff was eating 3 servers worth of cpu. The re-hashed again when nova merged a change that deprecated local-conductor | 19:25 |
klindgren_ | and a number of large installs that were doing local-conductor was like... um reports of performance issues using remote-conductor - why are we doing this? | 19:25 |
harlowja | :) | 19:28 |
harlowja | some of those large installs may be from me, ha | 19:29 |
harlowja | *cause we haven't turned on remote-conductor yet | 19:29 |
harlowja | and with stuff like this i'm not sure we still want to :-/ | 19:29 |
*** dims has joined #openstack-performance | 19:56 | |
*** alaski_ is now known as alaski | 20:24 | |
*** arnoldje has quit IRC | 20:51 | |
*** andreaf has quit IRC | 21:11 | |
*** andreaf has joined #openstack-performance | 21:19 | |
*** arnoldje has joined #openstack-performance | 21:20 | |
*** dims_ has joined #openstack-performance | 21:43 | |
*** dims has quit IRC | 21:44 | |
*** harshs has quit IRC | 21:58 | |
*** harlowja_ has joined #openstack-performance | 21:59 | |
*** harlowja has quit IRC | 21:59 | |
*** harshs has joined #openstack-performance | 22:05 | |
*** dimtruck is now known as zz_dimtruck | 22:32 | |
*** mriedem is now known as mriedem_away | 22:46 | |
*** dims_ has quit IRC | 22:55 | |
*** dims_ has joined #openstack-performance | 23:25 | |
*** dims_ has quit IRC | 23:26 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!