#fdio-csit: csit weekly 161012
Meeting started by mackonstan at 14:04:49 UTC
(full logs).
Meeting summary
-
- Dave Wallace (dwallacelf,
14:05:24)
- Agenda bashing (mackonstan, 14:05:38)
- Inputs from FD.io projects (mackonstan, 14:05:53)
- TLDK (mackonstan, 14:06:54)
- Peter - re csit patch for TLDK functional -
updates sent over by Ana, addressing review comments (mackonstan,
14:07:41)
- dwallacelf - update from TLDK meeting last week
- TLDK folks are using google test framework for unit tests, detail
to be posted on TLDK project wiki (mackonstan,
14:08:45)
- keith burns talking about csit in the context
of nsh - but no further details is forthcoming (mackonstan,
14:10:26)
- VPP (mackonstan, 14:12:26)
- need to summarize to VPP the csit1701
plan (mackonstan,
14:12:47)
- ACTION: Maciek -
summarize what's done in csit1609, and then summarize proposal for
csit1701 across - i) virl-functional, ii) phy-performance, iii)
vpp-test. (mackonstan,
14:13:36)
- send it to csit-dev, review by ALL csit folks;
and then fwd to vpp-dev. (mackonstan,
14:14:18)
- FD.io CSIT operational status (mackonstan, 14:14:51)
- jobs for 1609 stopped (mackonstan,
14:15:15)
- : https://gerrit.fd.io/r/#/c/3350/
(jgelety,
14:16:56)
- virl machines outage yesterday due to garbage
accumulated over the LF infra outages - see email thread on
csit-dev (mackonstan,
14:19:25)
- ACTION: Peter Mikus
to summarize work required to prevent this happening in the
future (mackonstan,
14:21:37)
- Testbeds upgrade to Ubuntu 16.04 (mackonstan, 14:21:58)
- Peter Mikus status update - VIRL Image 1.4
(Ubuntu Xenial) is now produced and operational. Simulation is
working and nodes are brought up. (mackonstan,
14:22:29)
- irc://barjavel.freenode.net:6697/#infoeter
Mikus - Physical TB3 is reinstalled and working properly. The rest
of TBs are waiting for successful VIRL run so we can switch. (mackonstan,
14:23:05)
- Peter Mikus - Physical TB3 is reinstalled and
working properly. The rest of TBs are waiting for successful VIRL
run so we can switch. (mackonstan,
14:23:51)
- currently blocked by PIP and setuptools not
compatible with TrafficScriptExecutor - wip (mackonstan,
14:27:29)
- https://gerrit.fd.io/r/#/c/3290/
(mackonstan,
14:27:36)
- Next step after fixing above issues with CSIT
code - need to adjust ci-management so VPP jobs will work.
(mackonstan,
14:32:10)
- Performance tests - csit1701 plan (mackonstan, 14:33:22)
- Maciek update - working with dpdk folks on
addressing it, next update after dpdk userspace summit next
week (mackonstan,
14:36:59)
- • Latency - path to further improve
accuracy - work with TRex project (mackonstan,
14:37:14)
- Maciek update - TRex lead confirmed that
improving latency measurement accuracy is possible by using
PTP/ieee1588 supprt in xl710 nics. looking for volunteers to help it
code in TRex. right now no plan. (mackonstan,
14:40:35)
- TLDK update - Anna (mackonstan, 14:41:07)
- TLDK functional test cases added - thanks to
Peter and Miro for good reviews. (mackonstan,
14:41:34)
- DPDK baseline tests - started to work on the
plan to integrate into CSIT - l2fwd, l3fwd, maybe testpmd.
(mackonstan,
14:45:19)
- Functional tests - csit1701 plan (mackonstan, 14:49:07)
- new vpp test infra framework - status -
framework changes tracked on gerrit (mackonstan,
14:49:34)
- https://gerrit.fd.io/r/#/c/3345/
(mackonstan,
14:49:45)
- https://gerrit.fd.io/r/#/c/3237/
(mackonstan,
14:50:40)
- CSIT dev work plan - review, discuss (mackonstan, 14:55:53)
- see jira tickets (mackonstan,
14:55:57)
- CSIT HC tests (mackonstan, 14:56:57)
- CSIT HC tests ready to go production
(mackonstan,
14:57:23)
- https://gerrit.fd.io/r/3018
(mackonstan,
14:57:44)
Meeting ended at 14:58:47 UTC
(full logs).
Action items
- Maciek - summarize what's done in csit1609, and then summarize proposal for csit1701 across - i) virl-functional, ii) phy-performance, iii) vpp-test.
- Peter Mikus to summarize work required to prevent this happening in the future
People present (lines said)
- mackonstan (41)
- collabot (4)
- jgelety (2)
- dwallacelf (1)
Generated by MeetBot 0.1.4.