| *** furlongm has quit IRC | 00:05 | |
| *** furlongm has joined #openstack-performance | 00:05 | |
| *** rfolco has quit IRC | 00:14 | |
| *** bsilverman has quit IRC | 00:24 | |
| *** arnoldje has quit IRC | 00:46 | |
| *** rfolco has joined #openstack-performance | 01:18 | |
| *** rfolco has quit IRC | 01:39 | |
| *** catintheroof has quit IRC | 02:06 | |
| *** arnoldje has joined #openstack-performance | 02:41 | |
| *** rfolco has joined #openstack-performance | 03:23 | |
| *** rfolco has quit IRC | 03:47 | |
| *** harlowja_at_home has joined #openstack-performance | 04:33 | |
| *** pcaruana has quit IRC | 04:34 | |
| *** harlowja_at_home has quit IRC | 05:01 | |
| *** linuxaddicts has joined #openstack-performance | 05:07 | |
| *** arnoldje has quit IRC | 05:22 | |
| *** dimtruck is now known as zz_dimtruck | 05:52 | |
| *** msimonin2 has quit IRC | 06:12 | |
| *** pcaruana has joined #openstack-performance | 06:55 | |
| *** msimonin has joined #openstack-performance | 07:00 | |
| *** msimonin has quit IRC | 07:02 | |
| *** msimonin has joined #openstack-performance | 07:22 | |
| *** Deng has joined #openstack-performance | 10:14 | |
| *** msimonin has quit IRC | 11:48 | |
| *** msimonin has joined #openstack-performance | 11:48 | |
| *** rfolco has joined #openstack-performance | 11:56 | |
| *** linuxaddicts has quit IRC | 12:22 | |
| *** catintheroof has joined #openstack-performance | 12:46 | |
| *** zz_dimtruck is now known as dimtruck | 12:53 | |
| *** linuxaddicts has joined #openstack-performance | 12:59 | |
| *** linuxgeek_ has joined #openstack-performance | 13:03 | |
| *** v12aml has quit IRC | 13:05 | |
| *** linuxaddicts has quit IRC | 13:07 | |
| *** v12aml has joined #openstack-performance | 13:07 | |
| *** v12aml has quit IRC | 13:10 | |
| *** v12aml has joined #openstack-performance | 13:12 | |
| *** v12aml has quit IRC | 13:12 | |
| *** v12aml has joined #openstack-performance | 13:17 | |
| *** v12aml has quit IRC | 13:18 | |
| *** v12aml has joined #openstack-performance | 13:20 | |
| *** v12aml has quit IRC | 13:21 | |
| *** v12aml has joined #openstack-performance | 13:22 | |
| *** v12aml has quit IRC | 13:25 | |
| *** v12aml has joined #openstack-performance | 13:26 | |
| *** rook has quit IRC | 13:27 | |
| *** v12aml has quit IRC | 13:47 | |
| *** v12aml has joined #openstack-performance | 13:49 | |
| Cassowary | DinaBelova: I have plotted two graphs. | 13:50 |
|---|---|---|
| Cassowary | https://www.dropbox.com/s/5e8xrrctu0rcwx3/CPU%20scaling%20%20-%20continuous%20vs%20increasing%20from%201.png?dl=0 | 13:50 |
| Cassowary | https://www.dropbox.com/s/txpkb8k6mpyexv8/CPU%20scaling%20-%20increase%20from%201.png?dl=0 | 13:50 |
| Cassowary | The first graph shows the VM CPU resize time (y axis) vs number of vCPUs (x axis) of continuous (blue) and resize from a VM with 1 vCPU (green) scenarios. | 13:50 |
| Cassowary | The second graph shows the VM CPU resize time (y axis) vs number of vCPUs (x axis), when resized from a VM with 1 vCPU at each step (The green line in first graph). | 13:50 |
| Cassowary | The error bars show the standard error of the gathered values at each step, as I did resize multiple times to get a mean value. | 13:50 |
| *** dimtruck is now known as zz_dimtruck | 13:54 | |
| *** msimonin1 has joined #openstack-performance | 13:56 | |
| *** msimonin has quit IRC | 13:56 | |
| *** zz_dimtruck is now known as dimtruck | 14:13 | |
| *** arnoldje has joined #openstack-performance | 14:31 | |
| *** Guest51132 is now known as med_ | 15:23 | |
| *** med_ has quit IRC | 15:24 | |
| *** med_ has joined #openstack-performance | 15:24 | |
| *** linuxgeek_ has quit IRC | 15:32 | |
| DinaBelova | Cassowary ack, interesting.. | 15:39 |
| *** linuxaddicts has joined #openstack-performance | 15:39 | |
| *** msimonin1 has quit IRC | 15:43 | |
| *** rcherrueau has joined #openstack-performance | 15:55 | |
| *** victorhe has joined #openstack-performance | 15:57 | |
| DinaBelova | #startmeeting Performance Team | 16:00 |
| openstack | Meeting started Tue Jul 26 16:00:07 2016 UTC and is due to finish in 60 minutes. The chair is DinaBelova. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:00 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:00 |
| openstack | The meeting name has been set to 'performance_team' | 16:00 |
| DinaBelova | hey folks! | 16:00 |
| DinaBelova | are people around? | 16:01 |
| DinaBelova | ^) | 16:01 |
| DinaBelova | :) | 16:01 |
| *** rohanion has joined #openstack-performance | 16:01 | |
| DinaBelova | rohanion o/ | 16:01 |
| rcherrueau | o/ | 16:01 |
| rcherrueau | 16:01 | |
| *** AugieMena has joined #openstack-performance | 16:01 | |
| rohanion | Hi! | 16:01 |
| DinaBelova | rcherrueau hello :) | 16:01 |
| DinaBelova | AugieMena o/ | 16:01 |
| AugieMena | o/ | 16:01 |
| *** luzC has joined #openstack-performance | 16:02 | |
| rcherrueau | Last week I didn't introduce myself, but I work with msimonin. | 16:02 |
| DinaBelova | ok. let's wait for a couple of moments to see if others are around | 16:02 |
| luzC | o/ | 16:02 |
| DinaBelova | rcherrueau I believe msimonin mentioned you some time, I was sure you are :) | 16:02 |
| DinaBelova | lezbar__ ? | 16:02 |
| rcherrueau | OK great | 16:02 |
| DinaBelova | luzC o/ | 16:02 |
| rcherrueau | msimonin could not be here btw | 16:03 |
| DinaBelova | rcherrueau ack | 16:03 |
| DinaBelova | ok, so let's get started and let's hope lezbar__ will be here as well :) | 16:03 |
| DinaBelova | #topic Summit stuff | 16:04 |
| lezbar__ | hi, sorry, I'm here but will have to leave early | 16:04 |
| DinaBelova | #info first of all, big news - voting for summit presentations has been started! | 16:04 |
| DinaBelova | lezbar__ ack | 16:04 |
| DinaBelova | #link https://www.openstack.org/summit/barcelona-2016/vote-for-speakers | 16:04 |
| DinaBelova | I just wonna to share ways to find performance/scale related talks I'm aware about :) | 16:05 |
| DinaBelova | this summit direct links to the submissions are not working :) so the only way to find these talks will be to search them | 16:06 |
| DinaBelova | either by person or talk name | 16:06 |
| DinaBelova | lemme share talks names one more time :) | 16:06 |
| DinaBelova | #info Let's Oversubscribe the OpenStack environment and try to find the weaknesses and improvements. - talk #1 | 16:06 |
| DinaBelova | #info Chasing 1000 nodes scale - talk #2 | 16:06 |
| DinaBelova | #info OSIC Enables the SDN Sale: Testing MidoNet with Neutron at Scale - talk #3 | 16:07 |
| DinaBelova | #info MidoNet Scalability Testing with Neutron in AWS to 1000 nodes - talk #4 | 16:07 |
| DinaBelova | #info Evaluating OpenStack Performance with the SPEC Cloud (TM) IaaS 2016 Benchmark - talk #5 | 16:07 |
| DinaBelova | lezbar__, I believe #3 and #4 are yours :) | 16:07 |
| lezbar__ | yup :) | 16:07 |
| DinaBelova | are you still planning to push the results to performance-docs prior the summit? :) | 16:08 |
| lezbar__ | of course! | 16:08 |
| DinaBelova | yay :) | 16:08 |
| *** sai has joined #openstack-performance | 16:08 | |
| DinaBelova | nice to read this, /me is really happy :) | 16:08 |
| DinaBelova | I'm pretty sure there are more interesting stuff in the voting process, but I did not have a chance yet to search the vote-for-speakers much | 16:09 |
| DinaBelova | if I'll find something interesting I'll just publish names of the sessions here | 16:09 |
| DinaBelova | if someone knows something interesting about more interesting sessions - jump in, it's just the right time :) | 16:10 |
| DinaBelova | it looks like nope, ok | 16:11 |
| DinaBelova | today I have only one more agenda item to cover, so let's proceed with osprofiler weekly update | 16:11 |
| DinaBelova | #topic OSProfiler weekly update | 16:11 |
| DinaBelova | rohanion ? | 16:11 |
| rohanion | Hi guys | 16:12 |
| rohanion | I've been working on patching neutron and keystone for the last two days | 16:12 |
| DinaBelova | rohanion with new osprofiler drivers usage? | 16:13 |
| rohanion | Of course | 16:13 |
| rohanion | Alex Ignatyev is on vacation so I'm finishing his tasks | 16:13 |
| DinaBelova | ack | 16:14 |
| rohanion | nothing more special | 16:14 |
| DinaBelova | rohanion ack, thanks for the update) | 16:15 |
| DinaBelova | in fact I have no news so far as well | 16:15 |
| DinaBelova | last week I was busy with relocation stuff, so /me is fully back to work only now :) | 16:15 |
| DinaBelova | hope to finalize my changes regarding osprofiler this week + help rohanion with osprofiler+sahara integratino | 16:16 |
| DinaBelova | integration* | 16:16 |
| rohanion | that would be awesome :) | 16:16 |
| DinaBelova | rohanion really think I can try to do it :) | 16:16 |
| DinaBelova | #topic Open Discussion | 16:16 |
| DinaBelova | do we have something else to cover or we have really quick meeting today? :) | 16:17 |
| DinaBelova | rohanion, rook_, lezbar__, AugieMena, rcherrueau, luzC ? | 16:17 |
| AugieMena | nothing else... voting on interesting summit talks :) | 16:17 |
| rcherrueau | Just for your information, We are working on the deployment script (using kolla) for the 1000 node experiment | 16:18 |
| rcherrueau | We plan to have something that works on our infrastructure for the end of the week | 16:18 |
| DinaBelova | rcherrueau are you planning to share this script? | 16:18 |
| rohanion | Nothing else from me :( | 16:18 |
| rcherrueau | yes | 16:18 |
| DinaBelova | rcherrueau cool, please keep us updated :) | 16:18 |
| rcherrueau | When this will fully work | 16:19 |
| rcherrueau | OK | 16:19 |
| DinaBelova | rcherrueau true :D | 16:19 |
| DinaBelova | ok, it looks like nothing more for today | 16:19 |
| DinaBelova | thanks everyone for coming | 16:19 |
| DinaBelova | bye! | 16:20 |
| lezbar__ | thnx bye! | 16:20 |
| AugieMena | bye | 16:20 |
| rcherrueau | bye! | 16:20 |
| DinaBelova | #endmeeting | 16:20 |
| openstack | Meeting ended Tue Jul 26 16:20:12 2016 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 16:20 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/performance_team/2016/performance_team.2016-07-26-16.00.html | 16:20 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/performance_team/2016/performance_team.2016-07-26-16.00.txt | 16:20 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/performance_team/2016/performance_team.2016-07-26-16.00.log.html | 16:20 |
| *** rcherrueau has quit IRC | 16:20 | |
| *** rohanion has quit IRC | 16:31 | |
| Cassowary | DinaBelova: any idea what this could be? | 16:35 |
| *** msimonin has joined #openstack-performance | 16:56 | |
| *** AugieMena has quit IRC | 17:01 | |
| *** victorhe has left #openstack-performance | 17:05 | |
| DinaBelova | Cassowary nope so far, this looks really suspicious :D | 17:41 |
| doffm | DinaBelova: I'm going to rebase https://review.openstack.org/#/c/254703/ if that is OK. | 17:42 |
| doffm | We need to merge to master internally, so might as well rebase upstream. | 17:42 |
| DinaBelova | doffm yep, thank you sir for the help! | 17:42 |
| doffm | I may get the time to address some of the issues in the reviews at some point, but for now just a rebase. | 17:42 |
| DinaBelova | doffm ack, thanks | 17:42 |
| DinaBelova | I had really hard months - so lots of my changes are really not up to date | 17:43 |
| doffm | https://review.openstack.org/#/c/254703/ rebased. | 18:37 |
| *** msimonin has quit IRC | 19:01 | |
| *** msimonin1 has joined #openstack-performance | 19:01 | |
| *** msimonin1 has quit IRC | 19:12 | |
| *** msimonin has joined #openstack-performance | 19:12 | |
| *** msimonin has quit IRC | 19:31 | |
| doffm | DinaBelova: Is there any reason https://review.openstack.org/#/c/310530/ hasn't merged yet? Does it just need another core review and workflow? | 20:18 |
| DinaBelova | doffm yep, I wonna either boris-42 or harlowja to take a look on this | 20:18 |
| harlowja | huh | 20:19 |
| DinaBelova | in fact I'm almost out of patience and I guess if none of them will comment, I'll merge it tomorrow | 20:19 |
| harlowja | who's dina | 20:19 |
| harlowja | lol | 20:19 |
| DinaBelova | ЖВ | 20:19 |
| DinaBelova | :D | 20:19 |
| harlowja | why is DinaBelova chatting at this time | 20:19 |
| harlowja | lol | 20:19 |
| harlowja | go to sleep | 20:19 |
| harlowja | lol | 20:19 |
| DinaBelova | harlowja :D heh, you forgot I've relocated? :) | 20:19 |
| harlowja | oh ya | 20:19 |
| harlowja | that happened? | 20:19 |
| harlowja | done? | 20:19 |
| DinaBelova | yep :D | 20:19 |
| harlowja | uh oh | 20:19 |
| harlowja | the party starts now | 20:19 |
| harlowja | lol | 20:19 |
| DinaBelova | :D so you need to listen to me now :D i'm in your tz :D | 20:20 |
| DinaBelova | review! https://review.openstack.org/#/c/310530/ | 20:20 |
| harlowja | :( | 20:20 |
| harlowja | haha | 20:20 |
| harlowja | yes ma'm | 20:20 |
| DinaBelova | :D | 20:20 |
| harlowja | mongo | 20:20 |
| harlowja | eck | 20:20 |
| harlowja | lol | 20:20 |
| harlowja | try: | 20:20 |
| harlowja | from pymongo import MongoClient | 20:20 |
| doffm | So, we have a guy internally, Mahesh. He has been doing some work on osprofiler, part of which was to look at overhead differences between ceilometer and mongo. | 20:20 |
| harlowja | i hate that stuff, ha | 20:20 |
| harlowja | doing imports in __init__ to detect that a module can't be used feels totally messed up to me :-P | 20:21 |
| DinaBelova | harlowja and https://review.openstack.org/#/c/251343/ is really needed as well btw. so if you don't have very specific comments for now.. | 20:21 |
| harlowja | damn, u in my TZ now | 20:21 |
| harlowja | can't ignore, ha | 20:21 |
| DinaBelova | :D | 20:21 |
| DinaBelova | ahaha, true | 20:21 |
| DinaBelova | doffm ack | 20:21 |
| doffm | I'll just type the results: Scenario, unspecified number of nova boots. | 20:21 |
| harlowja | i mean, ummm, not ignore | 20:22 |
| harlowja | i never ignore | 20:22 |
| harlowja | lol | 20:22 |
| doffm | No profiling avg time for boot: 9.18s | 20:22 |
| doffm | Profiling with ceilometer avg time for boot: 35.7s | 20:22 |
| doffm | Profiling with mongodb avg time for boot: 10.75s | 20:22 |
| doffm | He was using the same rabbitmq server for notifications and messages. Which probably isn't the right thing to do now that we can split them. | 20:23 |
| doffm | And I don't know how big the rabbit server he was using was. | 20:23 |
| harlowja | weird | 20:23 |
| harlowja | why is boot time affected like that | 20:23 |
| harlowja | seems messed up | 20:23 |
| harlowja | profiling isn't supposed to add 25 seconds on, lol | 20:23 |
| doffm | I'm guessing an overloaded rabbit server, doing lots of parallel boots. | 20:23 |
| doffm | Alot of extra rabbit traffic for the notifications. | 20:24 |
| harlowja | ya, weird | 20:24 |
| DinaBelova | doffm probably this might be a reason in case of ceilometer, yep... harlowja - ceilo is really loading cloud controllers, esp. if mongo is located on the same nodes | 20:24 |
| harlowja | makes it hard to get useful profiling if the profiling system itself adds on so much | 20:24 |
| DinaBelova | doffm in fact I would say you need to try set mongo and rabbit mq on separated nodes for ceilo case | 20:24 |
| harlowja | def | 20:24 |
| doffm | I think this could be mitiaged by using a separate rabbitmq for notifications. | 20:24 |
| harlowja | i wouldn't colocate the rabbits | 20:24 |
| harlowja | *couse the rabbits | 20:24 |
| harlowja | co-use-the-same-rabbit-hole | 20:25 |
| harlowja | idk, i'm not a rabbbit | 20:25 |
| DinaBelova | and (again) please check mongo, it has its own troughput affecting everything else of the controllers | 20:25 |
| DinaBelova | if it's located there | 20:25 |
| harlowja | weird seeing DinaBelova type so much | 20:25 |
| harlowja | lol | 20:25 |
| DinaBelova | harlowja :D | 20:25 |
| DinaBelova | deal with it :D | 20:25 |
| harlowja | :-P | 20:25 |
| doffm | DinaBelova: Are you located in the USA? | 20:26 |
| DinaBelova | doffm yep, CA | 20:27 |
| DinaBelova | for a week or so :D | 20:27 |
| harlowja | ha | 20:27 |
| DinaBelova | I mean I've relocated a week ago :) | 20:27 |
| harlowja | immigrants | 20:27 |
| DinaBelova | harlowja ahaha | 20:27 |
| DinaBelova | it's much easier to influence on you here :D | 20:27 |
| harlowja | lol | 20:27 |
| harlowja | well now u can walk over and find me | 20:28 |
| harlowja | which is scary! | 20:28 |
| harlowja | lol | 20:28 |
| DinaBelova | lol | 20:28 |
| dims_ | run and hide harlowja ~ | 20:28 |
| harlowja | i'm trying | 20:28 |
| harlowja | lol | 20:28 |
| doffm | So... The novaclient patch is essentially enabled by importing osprofiler.web. Since rally uses novaclient... | 20:41 |
| doffm | All I have to to is make sure osprofiler.web is in the path. Use a patched novaclient and rally should send out osprofiler enabled requests. | 20:42 |
| doffm | Yeah, that will totally work. *crosses fingers* | 20:42 |
| *** catintheroof has quit IRC | 20:52 | |
| harlowja | DinaBelova sooooo klindgren i think wants to get back into the perf stuff (volunteered, hahaha) anything he can do ;) | 20:52 |
| harlowja | :-P | 20:53 |
| harlowja | right kris | 20:53 |
| harlowja | lol | 20:53 |
| DinaBelova | harlowja klindgren heh, ack, good to know :) | 20:57 |
| klindgren | >_> | 20:57 |
| klindgren | <_< | 20:57 |
| DinaBelova | sorry, need to go afk for ~1 hour | 20:57 |
| klindgren | Do have a lovely stories about keepalives - and terribleness | 20:58 |
| harlowja | that one sounds like a good one for the campfire | 21:00 |
| harlowja | lol | 21:00 |
| *** rfolco has quit IRC | 21:12 | |
| *** linuxaddicts has quit IRC | 21:33 | |
| *** linuxaddicts has joined #openstack-performance | 21:36 | |
| *** msimonin has joined #openstack-performance | 21:40 | |
| *** openstackgerrit has quit IRC | 22:03 | |
| *** openstackgerrit has joined #openstack-performance | 22:04 | |
| *** arnoldje has quit IRC | 22:07 | |
| *** msimonin has quit IRC | 22:27 | |
| *** openstackgerrit has quit IRC | 22:48 | |
| *** openstackgerrit has joined #openstack-performance | 22:48 | |
| *** dimtruck is now known as zz_dimtruck | 22:51 | |
| *** zz_dimtruck is now known as dimtruck | 23:40 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
