18:16:39 <dfarrell07> #startmeeting cperf 18:16:39 <collabot> Meeting started Thu Aug 11 18:16:39 2016 UTC. The chair is dfarrell07. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:16:39 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic. 18:16:39 <collabot> The meeting name has been set to 'cperf' 18:20:31 <dfarrell07> #info Jamo, Luis and Nikos are on PTO this week 18:20:36 <dfarrell07> #info All of the NetVirt CSIT dfarrell07 was building on to create scale tests are broken, and infra is broken, so not such great progress this week 18:20:43 <dfarrell07> #info dfarrell07 and acm ran OPNFV TestPerf call this week 18:20:53 <dfarrell07> #info mgkwill and dfarrell07 talk about NetVirt CSIT 18:21:17 <dfarrell07> #info mgkwill has a test patch proposed that does OVS scale, up to 1800 nodes then hit docker issue 18:21:37 <dfarrell07> #info mgkwill reports that folks haven't really reviewed that patch, dfarrell07 has looked at it this week, can review 18:23:06 <dfarrell07> #info mgkwill has been working on Rally, which spins up networks and hosts on those networks and gives info. Seems smart to use that OpenStack built-in, vs brewing our own in Robot 18:25:57 <dfarrell07> #info dfarrell07 may be able to login to OPNFV jumphost and get access to those servers in POD 2, to get dfarrell07 some working dev machines to test all this on (again, upstream CSIT is broken) 18:26:54 <dfarrell07> #action dfarrell07 to try to logging into OPNFV jumphost, see if that can work as dev box 18:27:33 <dfarrell07> #info mgkwill points out that we could also move some servers for dev into POD, leave some for CI and others for dev work, but security implications 18:29:51 <dfarrell07> #info mgkwill and dfarrell07 talk about tests that use openstack vs don't, OVS and networks between them vs full openstack with compute nodes and rally managing 18:29:52 <dfarrell07> https://git.opendaylight.org/gerrit/#/c/37639/ 18:30:35 <dfarrell07> #link https://git.opendaylight.org/gerrit/#/c/37639/ Patch to work on getting running, build on 18:30:49 <dfarrell07> #action dfarrell07 to review that patch^^ 18:32:39 <dfarrell07> #info Al has been working on installing Nikos' most recent Latte program, based on Go, doing latency calcs (closer, but not totally working yet) 18:34:47 <dfarrell07> #info mgkwill sent out info about running O/S in containers with Kola, or parts of Kola. Matt is looking at that, seeing if we can reuse. Making progress, no blockers so far. mgkwill thinks networking will be a problem, uses host networking, not so isolated like docker 18:35:44 <dfarrell07> #endmeeting