15:01:09 <portdirect> #startmeeting openstack-helm
15:01:09 <openstack> Meeting started Tue Jul  2 15:01:09 2019 UTC and is due to finish in 60 minutes.  The chair is portdirect. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:12 <openstack> The meeting name has been set to 'openstack_helm'
15:01:15 <portdirect> #topic rollcall
15:01:19 <mattmceuen> o/
15:01:24 <lamt> o/
15:01:24 <portdirect> lets give it a few mins for folks to arrive :D
15:01:33 <jsuchome> o/
15:02:50 <dwalt> o/
15:03:45 <georgk> hi
15:03:59 <srwilkers> o/
15:04:36 <megheisler> \o
15:04:56 <portdirect> ok - lets kick off
15:05:23 <portdirect> with the upcoming 4th july holiday in usa i expect to be fairly lite from this side of the pond this week
15:05:38 <portdirect> (somthing about spilling tea or somesuch)
15:05:43 <portdirect> #topic Meeting time voting
15:05:45 <gagehugo> o/
15:06:07 <portdirect> so - i think voting for this ends today mattmceuen ?
15:08:25 <portdirect> ok - I think mattmceuen's got called into somthing else
15:08:39 <mattmceuen> sorry I am here
15:08:57 <portdirect> anyway - atm it looks like there is a preference to move to 2 hours before this slot
15:09:06 <mattmceuen> yep - eod today I believe is the end for meeting time voting, since we gave folks another week
15:09:21 <portdirect> though I'll try and get our contributors on the west coast (usa) to chime in
15:09:47 <portdirect> when does the vote end for airship mattmceuen ?
15:10:42 <mattmceuen> EOD today
15:10:51 <mattmceuen> Is how I defined it
15:10:55 <mattmceuen> so it depends when I go home :)
15:11:04 <portdirect> ok - great, once that's in lets sync - and we can make sure they dont end uverlapping :D
15:11:11 <mattmceuen> deal
15:11:13 <portdirect> overlapping even
15:11:40 <portdirect> ok - lets move on - the agenda is currently empty for today
15:12:09 <portdirect> though if you dont mind georgk/cheng1 i'd like to discuss ovs-dpdk?
15:12:26 <georgk> oh, I don't mind
15:12:40 <cheng1> sure portdirect
15:12:41 <georgk> happy to discuss it
15:12:52 <portdirect> #topic ovs-dpdk
15:13:06 <portdirect> so the work you have both been doing here is looking really awesome
15:13:29 <portdirect> i've got a few questions on the chart - though i think once they are sorted, we should be good to go :D
15:13:46 <georgk> it is not just us, others are helping a lot too
15:14:01 <georgk> ok, I saw your comments
15:14:04 <portdirect> my question is how do we intend to test this? as it stands will it be possible to launch vm's via neutron in the nodepool vms?
15:14:26 <georgk> cheng1: <- this is yours
15:14:39 <georgk> :-)
15:14:47 <cheng1> For openstack-helm-infra project, I have already a patch for the CI
15:15:17 <cheng1> https://review.opendev.org/#/c/664220/
15:15:33 <srwilkers> well, that only gets us so far
15:15:40 <cheng1> For openstack-helm project, I will also create a patch.
15:15:50 <srwilkers> okay, that'd be great :)
15:15:52 <portdirect> ok - great - thanks cheng1
15:16:17 <portdirect> i think this will be a great opporunity to make use of the feature-gates flag we introduced to the gates recently
15:16:20 <cheng1> At there, we are going to deploy dpdk-enabled openstack
15:16:28 <portdirect> but lets get there when we have a running vm
15:16:34 <cheng1> with dpdk parameters
15:16:51 <portdirect> georgk: re my worry about restarting the pod, is this valid, or just me being paranoid here?
15:16:53 <cheng1> But I am afraid we can't run VM
15:17:00 <portdirect> cheng1: why not?
15:17:17 <cheng1> Because unstable nested kvm support in the openstack CI
15:17:44 <portdirect> ok - I'll take this up with infra - is that the only blocker?
15:18:01 <georgk> portdirect: that's a valid question and I'll check again how OVS behaves
15:18:26 <portdirect> roger - if its an issue we can just do what we ended up having to do for libvirt
15:18:27 <georgk> the port config remains, but I haven't checked the flows
15:18:27 <cheng1> To my knowledge, yes, this is the only blocker
15:18:54 <cheng1> oh right, another thing to confirm
15:19:18 <georgk> but taht
15:19:21 <georgk> opps
15:19:29 <georgk> but that's not dpdk specific, is it?
15:20:05 <cheng1> The VM provisioned from the nodepool, I would like to know if the interface driver is virtio?
15:20:14 <portdirect> that depends - for std ovs existing flows are unaffected by userspace georgk
15:20:39 <georgk> that's true
15:20:48 <portdirect> cheng1: i think this will vary as well, could you put together a list of needs - and i can see what we can do
15:21:16 <georgk> ok, I need to check. My experience with ODL controlling OVS might interfere here
15:21:25 <cheng1> Sure portdirect , I will send out the email tomorrow
15:21:42 <portdirect> that would be great - thanks so much
15:22:14 <portdirect> anything else here?
15:23:11 <georgk> rihabb2 managed to deploy AIAB with Rocky and bionic
15:23:26 <georgk> just for your information, but I am very happy about this
15:23:57 <portdirect> nice - thats great
15:24:16 <portdirect> we should have stein support merged by eod i hope :D
15:26:52 <portdirect> ok - lets move on
15:27:01 <portdirect> #topic any-other-topics?
15:28:36 <portdirect> ok - if nothing else for today - lets give everyone 30 mins back
15:28:45 <portdirect> thanks everyone :D
15:28:53 <portdirect> #endmeeting