14:33:12 <fzhadaev> #startmeeting Fuel Casual
14:33:12 <collabot> Meeting started Tue Aug  2 14:33:12 2016 UTC.  The chair is fzhadaev. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:33:12 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:33:12 <collabot> The meeting name has been set to 'fuel_casual'
14:33:24 <fzhadaev> #info Fedor Zhadaev
14:33:39 <Greg_E_> #info Greg Elkinbard
14:33:48 <mskalski> #info Michal Skalski
14:34:09 <fzhadaev> Hi Greg
14:34:19 <Greg_E_> Hi Fedor
14:35:54 <z0d> #info Peter Barabas
14:39:35 <fzhadaev> I have the topic/question: Are there scenario owners who didn't fill their rows in https://wiki.opnfv.org/display/SWREL/Colorado+Scenario+Status ?
14:41:30 <fuel> #info David Chou
14:46:56 <Greg_E_> Fedor, do we have a list of critical defects that need to be targeted before stable branch creation?
14:49:32 <fzhadaev> Looking at our status table https://wiki.opnfv.org/display/fuel/Colorado+Release+Status - we have no critical issues ))
14:50:29 <fzhadaev> But actually we're unable to deploy OVS with DPDK for now. Me and mskalski are trying to debug this issue.
14:51:01 <fzhadaev> *OVS-scenario with DPDK
14:58:46 <Greg_E_> looks like we have a number of scenarios with no status or inactive that we are not testing
15:00:08 <fzhadaev> we have almost all noha scenarios in inactive state
15:00:37 <fzhadaev> I mean for all of them we have the same scenario with HA, which is active
15:00:53 <Greg_E_> os-nosdn-kvm-ha - no status
15:01:39 <Greg_E_> os-nosdn-ovs-ha, os-nosdn-vlan-ha — inactive
15:02:13 <fzhadaev> OVS is in debug now
15:02:26 <Greg_E_> os-odl_l2-sfc-ha - no status
15:02:34 <fuel> Thanks a lot for Michal Skalski's help to enable os-nosdn-kvm-ha, up to now, it run 4 times, 2 times pass, and 2 times fial, we need Michal's continuous help us to debug.
15:03:12 <mskalski> fzhadaev: dpdk with mirantis packages + e1000 driver works, then maybe problems can be specific for BCM57840 nic
15:03:33 <mskalski> fzhadaev: ^ virt env
15:04:16 <Greg_E_> We never tested broadcom
15:04:20 <Greg_E_> only intel
15:04:31 <Greg_E_> and qlogic tested qlogic
15:04:31 <mskalski> fuel: I use your ticket to request intel pod access
15:05:33 <fuel> Got it, thanks.
15:05:49 <Greg_E_> for the failed scenarios we should add a bug id or at least a comment on the release page
15:07:16 <mskalski> Greg_E_: understood, in fuel it is defined as nic which support dpdk https://github.com/openstack/fuel-web/blob/stable/mitaka/nailgun/nailgun/fixtures/openstack.yaml#L2188-L2217 but found at least one mail thread about problems with this card and dpdk on ubuntu
15:09:37 <Greg_E_> I think this is a copy from DPDK tree
15:09:43 <Greg_E_> testing was limited
15:10:00 <Greg_E_> Intel Niantic and Forterville by us
15:10:17 <Greg_E_> and Qlogic is supposed to be testing theirs
15:10:45 <Greg_E_> Michal Chernik did the testing you can check with him
15:10:50 <DanSmithEricsson> #info Dan Smith (coming from other meeting)
15:14:49 <mskalski> on lf-pod2 we have "Cisco Systems VIC Ethernet NIC" which also in theory should support dpdk, but we are waiting for appropriate vlan configuration there
15:15:25 <mskalski> so we have two types of NICs which we are not sure if will work
15:16:02 <Greg_E_> yes
15:16:16 <Greg_E_> our DPDK testing was very limited for 9.0
15:17:17 <Greg_E_> but Teselkin should still help you to debug issues on both NICs
15:17:28 <Greg_E_> even if there is no official release support
15:18:05 <Greg_E_> Do we support IPv6 in Colorado?
15:19:00 <mskalski> and this is about mirantis packages, ovs version from ovs plugin do not respect settings from /etc/default/openvswitch-switch and don't run vswitchd with --dpdk switch, but it is probably easy to fix
15:19:29 <mskalski> in ovs-ctl script
15:21:22 <mskalski> Greg_E_: Dmitry Teselkin ?
15:26:32 <Greg_E_> yes
15:26:42 <Greg_E_> Head of our Linux team
15:26:44 <mskalski> ok thanks
15:26:56 <Greg_E_> he will need hardware access
15:27:12 <Greg_E_> we did not build any test beds with Cisco or BMC
16:04:34 <fzhadaev> #endmeeting