==============================
#fdio-csit: csit weekly 161026
==============================


Meeting started by mackonstan at 14:00:46 UTC.  The full logs are
available at
http://ircbot.wl.linuxfoundation.org/meetings/fdio-csit/2016/fdio-csit.2016-10-26-14.00.log.html
.



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

* : Inputs from FD.io projects - VPP  (mackonstan, 14:03:25)
  * Damjan providing update - new features: snat, security groups
    (mackonstan, 14:04:09)
  * Maciek, Dave proposal - use the new make test infra, once merged
    with initial baseline tests, to execute per patch verify jobs for
    functional tests.  (mackonstan, 14:12:51)
  * use that as a main functional verification test framework going
    framework, incl. as an acceptance criteria for any new VPP node code
    (mackonstan, 14:13:35)
  * Dave Wallace and Jan Gelety to provide oversight of make test
    framework and tests, and updates on vpp and csit calls  (mackonstan,
    14:21:30)
  * Dave Wallace and Jan Gelety also agreed to review all patches into
    vpp make test, and +1 once kosher  (mackonstan, 14:22:01)

* DPDK+VPP focused pre-integration testing.  (mackonstan, 14:24:14)
  * some work in the background - some progress.  (mackonstan, 14:24:51)

* TLDK - no update.  (mackonstan, 14:24:59)

* FD.io CSIT operational status - VIRL testing jobs - csit-vpp,
  vpp-csit.  (mackonstan, 14:25:33)
  * issues with vpp restart - brute force solution -
    https://gerrit.fd.io/r/#/c/3572/  (mackonstan, 14:26:30)
  * 3572 is short-term and going into csit oper branch now  (mackonstan,
    14:27:36)
  * proper fix is to increase http://dut_setup.sh/ execution timeout to
    100s +  (mackonstan, 14:30:18)
  * this is due to as of ubuntu16.04 we're using systemd to drive vpp
    restart, and systemd has 90sec timeout  (mackonstan, 14:31:39)
  * brute force solution cherry-picked to oper-161024 branch:
    https://gerrit.fd.io/r/#/c/3598/  (jgelety, 14:31:59)

* Perf testing jobs - csit-vpp, vpp-csit.  (mackonstan, 14:41:11)
  * Jan reporting occasionally qemu report error after parsing output of
    /proc/memory for hugepages variables  (mackonstan, 14:42:16)

* Performance tests  (mackonstan, 14:48:19)
  * dpdk rls baseline - work in progress  (mackonstan, 14:48:40)
  * trending baseline tests -
    https://jenkins.fd.io/view/csit/job/csit-vpp-perf-master-long/plot/RFC2544%3ANDR/
    (mackonstan, 14:50:34)
  * multi-core - non-repeatable results - needs investigation - Peter,
    Maciek to lead, all others to chip in  (mackonstan, 14:52:12)

* Functional tests - vpp make test framework  (mackonstan, 14:54:51)
  * initial test infra and sample test cases patches - to be merged
    after rebase and code-review  (mackonstan, 15:00:17)
  * LINK: https://gerrit.fd.io/r/#/c/3345/  (mackonstan, 15:00:22)
  * Functional tests - csit virl  (mackonstan, 15:00:30)

* Functional tests - csit virl  (mackonstan, 15:00:41)
  * all tests that need driver functionality stay in csit virl
    functional e.g. vhost, SPAN, dpdk, af_packet, tuntap.  (mackonstan,
    15:01:38)
  * need to update jira  (mackonstan, 15:03:15)

* CSIT HC tests  (mackonstan, 15:04:02)
  * Samuel relaying request from OPNFV/FDS team to test HC+VPP running
    on Centos  (mackonstan, 15:05:18)
  * this requires onboarding of Centos 7.2 into VIRL, so that HC CSIT
    tests described at links below can be executed in the environment
    desired for FDS  (mackonstan, 15:06:36)
  * LINK: https://wiki.fd.io/view/CSIT/Honeycomb/Summary  (mackonstan,
    15:06:51)
  * ACTION: Samuel to email description to csit-dev of what's needed
    from VIRL Centos VM to conduct HC tests  (mackonstan, 15:07:35)
  * let's cover it from now on as part of HC project update at the
    beginning of csit calls  (mackonstan, 15:08:00)



Meeting ended at 15:08:20 UTC.



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

* mackonstan (36)
* collabot (4)
* jgelety (1)



Generated by `MeetBot`_ 0.1.4