18:14:23 #startmeeting cperf 18:14:23 Meeting started Thu Jul 20 18:14:23 2017 UTC. The chair is dfarrell07. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:14:23 Useful Commands: #action #agreed #help #info #idea #link #topic. 18:14:23 The meeting name has been set to 'cperf' 18:21:17 o/ 18:25:17 #info dfarrell07 talks about wanting to hand off PTL roll to mattw4, who is doing great perf work and would be a great next person in this seat 18:25:31 #info jamoluhrsen talks about CPerf POD still needing IP fixes 18:25:42 #info LuisGomez talks about upcoming test work, scale and clustering 18:25:51 #info LuisGomez is looking at OpenFlow, as always 18:36:22 dfarrell07, drinking beer!!! 18:36:26 * jamoluhrsen is jealous 18:36:40 jamoluhrsen: cava and aperol, fancy ;) 18:37:04 #info discussion about Sai's perf results he talked about on Int/Test call today, what to do next 18:37:20 #info we want to do something like we did for the perf whitepapers in the past 18:37:38 #info Sai has detailed writeup already, which we can publish as the "how to do this yourself" tech doc 18:37:56 #info would need to create a high-level, human-readable version for general consumption 18:38:06 #info would do this for the next tests, Carbon SR1 18:38:21 #info dfarrell07 would rope in the marketing ppl once we have these results 18:38:37 #info marking ppl can help direct us, get it published and PR'd, do Linux.com interviews and such 18:38:48 #info discussion about tooling with mattw4 and sai_ 18:39:21 #info talking about testing sai_ is doing, if it can be implemented in a cperf job 18:39:42 #info jamoluhrsen is of course master of creating jobs, so should be able to take anything reasonably automated and make a job out of it 18:39:58 #info starting state is ODL with OpenStack, which we already have in CPerf via Apex 18:40:05 #info next steps are fairly easy to automate 18:40:40 #info first step is to figure out steps to run sai_'s test against any generic openstack cloud 18:41:22 #info mattw4 points out that his container based scale deployment could maybe be the base for this testing 10:47:12 #endmeeting