#opnfv-cperf: cperf
Meeting started by jamoluhrsen at 18:16:35 UTC
(full logs).
Meeting summary
-
- tim rozet gave us an updated patch for the
cbench hello world job which will give us an apex deploy on intel
pod2 then calling cperf job. (jamoluhrsen,
18:21:17)
- mgkwill has checked the patch and it looks like
it should work for us. (jamoluhrsen,
18:21:32)
- question is if we can still just manually
trigger cperf after deploy is done. answer is yes (jamoluhrsen,
18:22:56)
- Al Morton (acm,
18:23:34)
- reminder that we can just redploy the overcloud
after full apex deploy is done. (jamoluhrsen,
18:26:19)
- opnfv does not have a sandbox, so we kind of
just have to merge the job(s) and see how they go. (jamoluhrsen,
18:26:56)
- there is someone from huawei that may be
joining cperf as well as veryx resource down the road. (jamoluhrsen,
18:29:10)
- Bhuvan (jamoluhrsen,
18:34:07)
- jamo (jamoluhrsen,
18:34:09)
- Luis (jamoluhrsen,
18:34:11)
- Marcus (jamoluhrsen,
18:34:14)
- it would be good to figure out a way to quickly
swap out controllers inside these already deployed OPNFV
environments. (jamoluhrsen,
18:36:38)
- more discussions of the kinds of tests that
should exist in CPERF. it really should be controller agnostic
tests, and not those that are specific/internal to controllers (e.g.
bulk-o-matic in ODL, or internal java api tests in ONOS)
(jamoluhrsen,
18:43:13)
- no answer yet on how to use licensed product in
opnfv (e.g. veryx tool) (jamoluhrsen,
18:51:50)
- opnfv cperf project may be a good place to run
some of our ODL perf tests that we will use in future (like Boron)
perf report (jamoluhrsen,
18:53:11)
- acm tried to install nstat this week, but
didn't quite get there. (jamoluhrsen,
18:55:03)
Meeting ended at 18:55:50 UTC
(full logs).
Action items
- (none)
People present (lines said)
- jamoluhrsen (17)
- collabot (3)
- acm (1)
Generated by MeetBot 0.1.4.