======================================
#opendaylight-integration: integration
======================================


Meeting started by jamoluhrsen at 17:02:55 UTC.  The full logs are
available at
http://meetings.opendaylight.org/opendaylight-integration/2017/integration/opendaylight-integration-integration.2017-12-14-17.02.log.html
.



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

* builder  (jamoluhrsen, 17:04:55)
  * gerrit migration is done and lots of helpdesk tickets (we are behind
    on those) dealing with that (e.g. credentials/permission issues)
    (jamoluhrsen, 17:05:40)
  * we have now moved to our new cloud provider, Vexxhost  (jamoluhrsen,
    17:07:18)
  * there are issues being worked through still  (jamoluhrsen, 17:07:27)
  * dfarrell07 noted some missing docker and ubuntu images, which is
    news to the releng team.  (jamoluhrsen, 17:08:06)
  * current issue being debugged is that nodes wont connect, but it's
    currently being debugged  (jamoluhrsen, 17:10:04)
  * the recommendation is that we use the performance nodes, instead of
    standard nodes. they have dedicated cores  (jamoluhrsen, 17:11:04)
  * Vexxhost standard VMs https://vexxhost.com/public-cloud/servers/
    (zxiiro, 17:11:24)
  * LuisGomez wonders what our caps are with Vexxhost, and it looks like
    it might be storage related since the performance nodes require an
    attached volume  (jamoluhrsen, 17:13:50)
  * so essentially we are uncapped and it's up to us and jenkins to
    manage our resources  (jamoluhrsen, 17:15:46)
  * dfarrell07 is noticing that jobs might be taking longer to start
    (jamoluhrsen, 17:18:53)
  * LINK:
    https://jenkins.opendaylight.org/releng/job/packaging-test-rpm-master/
    (dfarrell07, 17:20:08)
  * LINK:
    https://jenkins.opendaylight.org/releng/job/packaging-test-rpm-master/buildTimeTrend
    (jamoluhrsen, 17:21:00)
  * zxiiro is going to try autorelease on a standard node  (jamoluhrsen,
    17:23:33)
  * the default system for ODL is 1 core, 4 GB  (jamoluhrsen, 17:25:14)
  * we are curious to see how cluster jobs operate in this new
    (hopefully more performant) cloud provider. one theory was that some
    flakiness was due to busy/shared/slow networks between systems
    (jamoluhrsen, 17:26:37)
  * wondering if it's possible now for non LF-admins to have ssh access
    to these new vexxhost vms.  (jamoluhrsen, 17:29:24)
  * it should technically be possible for non-LF-admins to access these
    VMs but it might be tricky/complicated  (jamoluhrsen, 17:30:40)

* packaging  (jamoluhrsen, 17:31:27)
  * CD pipeline is in good shape now, with some final patches to get
    merged  (jamoluhrsen, 17:31:59)
  * still facing a weird build issue when done on the community cbs
    build system where it's got a requirement on /usr/bin  (jamoluhrsen,
    17:33:19)
  * opnfv is still having intermittent hangs with ODL (recent carbon
    distro and we are close to sr3)  (jamoluhrsen, 17:34:47)
  * distribution  (jamoluhrsen, 17:35:45)
  * wiki is now updated to show Luis as the PTL for int/dist
    (jamoluhrsen, 17:36:21)
  * the distribution-version feature will be deprecated. email sent and
    no major objections.  (jamoluhrsen, 17:36:47)
  * we probably need to figure out how to "deprecate" this feature in
    the code  (jamoluhrsen, 17:37:39)

* misc  (jamoluhrsen, 17:39:43)
  * LuisGomez brings up a worry that we may lose a lot of the
    work/support/value Vratko was giving to the kernal projects (e.g.,
    contorller, yangtools, netconf, etc)  (jamoluhrsen, 17:40:44)
  * this goes along with a lack of devs in some of these core projects,
    we also need test folks  (jamoluhrsen, 17:42:16)



Meeting ended at 17:52:09 UTC.



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

* jamoluhrsen (29)
* odl_meetbot (3)
* dfarrell07 (1)
* zxiiro (1)
* odp-gerritbot (1)



Generated by `MeetBot`_ 0.1.4