===================
#fdio-vpp: fdio-vpp
===================


Meeting started by DaveBarach at 14:57:03 UTC.  The full logs are
available at
http://ircbot.wl.linuxfoundation.org/meetings/fdio-vpp/2020/fdio_vpp/fdio-vpp-fdio_vpp.2020-08-11-14.57.log.html
.



Meeting summary
---------------

* CSIT (Maciek reporting)  (DaveBarach, 15:02:28)
  * Continuing to focus on stateful tests, calls-per-second and
    throughput transactions-per-second tests. Initial target is VPP
    NAT44-ED 5-tuple tests for CSIT-2009.  (mackonstan, 15:02:47)
  * New hardware, adding AMD EPYC zen2 performance testbed.
    (mackonstan, 15:03:57)
  * Trending pages updated, extended historical data coverage from 3mths
    to 6mths to fully cover release cycle.  (mackonstan, 15:04:17)
  * Lots of other housekeeping work.  (mackonstan, 15:04:49)

* Host Stack (Florin)  (DaveBarach, 15:05:15)
  * tracking pending timers due to timer pacing algorithm  (DaveBarach,
    15:05:52)
  * network simulator now supports network reordering  (DaveBarach,
    15:07:13)
  * 163.com working on a vcl patch to enable multi-threaded apps w/
    workers  (DaveBarach, 15:07:51)
  * goal: to enable envoy to work with vcl / ld_preload. 30% latency
    reduction reported  (DaveBarach, 15:08:13)

* Documentation (John D)  (DaveBarach, 15:08:39)
  * expect to add infographic from the marketing team over the next week
    or so  (DaveBarach, 15:08:57)
  * 20.05 release notes avaliable on the websit  (DaveBarach, 15:09:08)

* Release Management (Andrew, DaveW)  (DaveBarach, 15:11:12)
  * 20.09 API freeze in about 4 weeks  (DaveBarach, 15:11:22)
  * deprecation tagging included in the API freeze, please make a note
    of it  (DaveBarach, 15:11:57)
  * 280 patches in master heading into 19.08  (DaveBarach, 15:12:32)
  * found out that bursts of more than 20 cherry-picks causes serious
    issues w/ jenkins  (DaveBarach, 15:13:14)
  * Andrew figured out how to dribble cherry-picks into 19.08
    (DaveBarach, 15:13:38)
  * 20.05.1 is complete  (DaveBarach, 15:13:46)
  * Do we actually need to produce CSIT reports for 20.05.1?
    (DaveBarach, 15:14:12)
  * resource limitation argue for not going there.  (DaveBarach,
    15:14:35)
  * AGREED: we will not create CSIT reports for point releases,
    specifically 20.05.1  (DaveBarach, 15:15:36)
  * artifact naming changes waiting for Damjan to return from PTO
    (DaveBarach, 15:16:17)
  * coverity warnings (a number of false positive) 7 remain
    (DaveBarach, 15:17:07)
  * please clean up  (DaveBarach, 15:17:16)
  * sonarcloud seems to catch real bugs in .py code  (DaveBarach,
    15:17:57)
  * ACTION: Andrew to dig into sonarcloud defects  (DaveBarach,
    15:18:19)

* Infrastructure  (DaveBarach, 15:19:24)
  * nomad machines all running 18.04, installed by ansible  (DaveBarach,
    15:19:41)
  * struggling with LF inability to understand / monitor their own infra
    (DaveBarach, 15:20:24)
  * Havent seen Vexxhost timeouts recently  (DaveBarach, 15:21:01)
  * have somewhat better visibility on vm performance through DataDog
    (DaveBarach, 15:21:22)
  * "git checkout" timeouts when 100 Jenkins jobs are trying at the same
    time  (DaveBarach, 15:22:24)
  * Probably no need to upload artifacts on verify job successes
    (DaveBarach, 15:28:05)



Meeting ended at 15:30:35 UTC.



People present (lines said)
---------------------------

* DaveBarach (34)
* mackonstan (4)
* collab-meetbot (3)



Generated by `MeetBot`_ 0.1.4