#fdio-csit: CSIT-weekly-160914
Meeting started by mackonstan at 14:02:09 UTC
(full logs).
Meeting summary
- agenda bashing (mackonstan, 14:02:39)
- Inputs from FD.io projects (mackonstan, 14:04:16)
- VPP (mackonstan, 14:04:21)
- dwallacelf (dwallacelf,
14:04:28)
- vpp1609 releasing today, on target (mackonstan,
14:04:35)
- DPDK+VPP focused pre-integration testing - Waterman Cao (mackonstan, 14:07:46)
- no updates from Waterman (mackonstan,
14:08:11)
- TLDK test cases - Anna (Qun Wan). (mackonstan, 14:08:34)
- based on review binaries should be removed,
Anna confirmed she will do it (mackonstan,
14:08:52)
- https://gerrit.fd.io/r/#/c/2492/
< tldk test cases (mmiklus,
14:09:13)
- HoneyComb project (mackonstan, 14:10:41)
- Samuel - baseline test cases are already
developed, but not automated yet (mackonstan,
14:13:08)
- it may be worth to have an overview of HC
testcases in place, and plan, add to agenda for next week
(mackonstan,
14:13:38)
- ACTION: Samuel Elias
- prepare for CSIT HC TCs readout next week on weekly csit
(mackonstan,
14:14:05)
- FD.io CSIT operational status (mackonstan, 14:14:23)
- Peter - performance testbed connectivity
outages: 8-Sep (mackonstan,
14:17:42)
- Matej - functional testbeds - no infra
issues. (mackonstan,
14:18:53)
- Ed offering to help to assist with any
real-time LF infra issue - irc channels: fdio-csit, gmail
account (mackonstan,
14:20:13)
- 8-sep outage tracked under: [FD.io Helpdesk
#29228] JENKINS - again issue with degraded network (mmiklus,
14:20:51)
- CSIT exec time reduction update (mackonstan, 14:21:51)
- Matej provided update - Jan submitted the patch
for review (mackonstan,
14:22:51)
- https://gerrit.fd.io/r/#/c/2855/ (mackonstan,
14:22:58)
- csit1609 release readiness (mackonstan, 14:24:52)
- for functional test report use this job:
csit-vpp-functional-1609-virl (mackonstan,
14:27:44)
- for performance test report use this job:
csit-vpp-perf-1609-all (mackonstan,
14:28:00)
- functional report will list failing test cases
with explanation (mackonstan,
14:34:29)
- performance report will show throughput and
latency numbers, also showing repeatability of results on plots
(over multiple job runs) (mackonstan,
14:36:09)
- VM-virtio, VPP-vhost-user numbers - update Peter. (mackonstan, 14:40:37)
- expecting improved performance with patched
vhost-user code (mackonstan,
14:40:53)
- Latency measurements - update Miroslav (mackonstan, 14:41:26)
- Miro proposed a patch to measure latency at
10%NDR and 50%NDR in addition to 100%NDR (mackonstan,
14:42:08)
- https://jenkins.fd.io/view/csit/job/csit-vpp-perf-master-long/1052/console
< verify job (mmiklus,
14:42:25)
- Review failing tests CSIT Jira 1609-fail (mackonstan, 14:43:54)
- https://jira.fd.io/issues/?filter=10667
(mackonstan,
14:44:17)
- Samuel: csit-402 documented, csit-409 - will
do (mackonstan,
14:44:33)
- Matej: all my CSIT-xxxx are documented
(mackonstan,
14:45:00)
- Test report generation (mackonstan, 14:45:28)
- Peter - updated parser and report, and the
graphs (mackonstan,
14:46:03)
- https://wiki.fd.io/view/CSIT/VPP-16.09_Test_Report_Draft
(mackonstan,
14:46:07)
- some performance anomalies between tb1-or-tb2
and tb3 (mackonstan,
14:52:21)
- ACTION: Peter, Maciek
to investigate (mackonstan,
14:52:30)
- verify kernel versions in tb1, tb2, tb3
(mackonstan,
14:57:39)
- CSIT dev work plan - wiki plan, jira tracking (mackonstan, 14:58:54)
- vpp1612 planning meeting on 20-Sep, Suggestion:
follow one week later with csit1612. (mackonstan,
14:59:33)
Meeting ended at 15:07:39 UTC
(full logs).
Action items
- Samuel Elias - prepare for CSIT HC TCs readout next week on weekly csit
- Peter, Maciek to investigate
People present (lines said)
- mackonstan (45)
- collabot (5)
- mmiklus (3)
- dwallacelf (1)
Generated by MeetBot 0.1.4.