14:03:26 <shaohe_feng> #startmeeting openstack-cyborg-driver
14:03:27 <openstack> Meeting started Mon Jul  2 14:03:26 2018 UTC and is due to finish in 60 minutes.  The chair is shaohe_feng. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:03:28 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:03:31 <openstack> The meeting name has been set to 'openstack_cyborg_driver'
14:03:49 <shaohe_feng> #topic Roll Call
14:04:02 <shaohe_feng> #info shaohe
14:05:52 <Helloway> #info Helloway
14:07:23 <shaohe_feng> #topic let's continue the last discussion on LC3.
14:08:39 <shaohe_feng> zhenghao has send a email about more os-acc details
14:09:10 <shaohe_feng> we have discussed that nova should make some changes for its DB
14:10:29 <shaohe_feng> Maybe RequestSpec and Instance should add a new field for accelerator info.
14:11:32 <shaohe_feng> also cyborg should standardize its API define.
14:11:49 <shaohe_feng> edleafe, are you online?
14:12:52 <shaohe_feng> can you elaborate when/how nova will call cyborg to get the allocated accelerators information
14:13:21 <edleafe> shaohe_feng: yes, but in another meeting, too
14:13:55 <shaohe_feng> and when/how nova will call cyborg to the bind the accelerator to an instance.
14:14:19 <shaohe_feng> edleafe, OK, we can discuss it later. Thanks
14:14:54 <shaohe_feng> let move the next topic
14:14:57 <shaohe_feng> #topic current status of drivers
14:16:55 <shaohe_feng> #link https://etherpad.openstack.org/p/cyborg-driver-tasks
14:17:40 <shaohe_feng> AOB
14:18:39 <shaohe_feng> ^ zhaohongbo, shashaguo, Helloway
14:19:49 <shaohe_feng> OK. seems most of us are on vocation. let's the meeting earlier. Thanks.
14:20:12 <shaohe_feng> let's end the meeting earlier.
14:20:14 <edleafe> shaohe_feng: sorry for the delay. Right now there is nothing set in Nova for calling cyborg. The whole process is still being discussed.
14:20:42 <shaohe_feng> edleafe, have you discuss with other developers?
14:20:53 <zhaohongbo> I am testing opae, and will join later
14:21:59 <shaohe_feng> edleafe,  I have seen scheduler report client has support claim_resource
14:22:36 <shaohe_feng> what should cyborg do to support it?
14:22:58 <edleafe> shaohe_feng: a lot of the discussions have been postponed while they figure out how to migrate existing nested allocation from their current flat design to a true nested design
14:23:25 <edleafe> That's holding up other development work
14:25:17 <shaohe_feng> edleafe, got it, thanks for the info. so is it also  holding up accelerator resource report?
14:27:18 <edleafe> Not holding it up exactly, but more like distracting attention away from it
14:28:46 <shaohe_feng> OK, we can continue the accelerator resource report. thanks.
14:28:57 <shaohe_feng> If no more, we can end the meeting.
14:29:20 <hongbozhao> ok
14:30:08 <shaohe_feng> #endmeeting