03:04:04 <Yumeng> #startmeeting openstack-cyborg
03:04:05 <openstack> Meeting started Thu Jul 16 03:04:04 2020 UTC and is due to finish in 60 minutes.  The chair is Yumeng. Information about MeetBot at http://wiki.debian.org/MeetBot.
03:04:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
03:04:08 <openstack> The meeting name has been set to 'openstack_cyborg'
03:04:13 <Yumeng> #topic Roll call
03:04:17 <Yumeng> #info Yumeng
03:04:21 <s_shogo> #info s_shogo
03:04:34 <brinzhang> #info brinzhang
03:04:42 <Yumeng> #topic Agenda
03:04:43 <Yumeng> https://wiki.openstack.org/wiki/Meetings/CyborgTeamMeeting#Agenda
03:05:45 <Yumeng> s_shogo: can you pls reply the comments or update Programming support https://review.opendev.org/#/c/698190/  ?
03:06:33 <brinzhang> Yumeng: do we have the details docs for device compatibility interface by Chinese? or put together http://lists.openstack.org/pipermail/openstack-discuss/2020-July/015902.html
03:06:50 <Yumeng> policy refresh is ready to review :https://review.opendev.org/#/c/740542/  still have some tempest error, I will fix them.
03:07:22 <brinzhang> I seem missed something with this topic, hope can get catch up that
03:07:24 <Yumeng> brinzhang: yes, I see this mail.
03:07:29 <s_shogo> Yumeng OK, I'm working with the patch,  test issue related to RabbitMQ, and will update that > 698190There are no big issue aside that.
03:08:07 <Yumeng> but we have not any Chinese docs about this.
03:08:23 <xinranwang__> Hi all
03:08:24 <brinzhang> Yumeng: xinranwang or shaohe_feng, may have give more help
03:08:35 <Yumeng> this seems useful for supporting live migration with accelerators.
03:08:43 <swp20> Hi all.
03:08:53 <Yumeng> brinzhang:yes, I will sync with them
03:09:04 <Yumeng> hi swp20,xinranwang__
03:09:30 <brinzhang> I didn't calm down to see this ML, so I want to know more, I think it would be better if there is Chinese material.
03:09:42 <brinzhang> hi xinranwang__, swp20^
03:10:25 <xinranwang__> We need to keep an eye on this maillist, in case there's some work can be done in cyborg.
03:11:30 <Yumeng> xinranwang__: yes! exactly.
03:12:26 <Yumeng> I will dig more and try to reply to ML to check things for cyborg.
03:13:04 <Yumeng> and next review will be on the specs review.
03:13:21 <brinzhang> xinranwang__: ok, later, will seen into
03:15:05 <Yumeng> Inspur FPGA DRIVER: https://review.opendev.org/#/c/730760/  no big issue for me, just  clarify some needed info will make sense to me.
03:15:22 <xinranwang__> As my understanding, it is an initial design discussion, anyway, we can have a look :)
03:15:34 <swp20> Yumeng, i will check your review soon.
03:15:59 <swp20> Yumeng, what's the use of project_id in arq, should we fill it by instance project_id? Pls see the patch:https://review.opendev.org/#/c/738427/
03:16:10 <brinzhang> xinranwang__: yes, looks like an initial design thought for device migration
03:16:33 <brinzhang> xinranwang__: I saw this ML sent by intel, you dont know him?
03:16:58 <Yumeng> project_id in arq is necessary, for example it can be used for policy check
03:17:09 <xinranwang__> I didn't know him before. maybe shaohe's working with them
03:17:56 <brinzhang> yeah, shaohe is missing this meetting
03:19:21 <xinranwang__> btw, just update QAT spec, comments are welcomed. Please check it.
03:19:34 <xinranwang__> https://review.opendev.org/#/dashboard/self
03:19:41 <xinranwang__> sorry, wrong link
03:19:41 <Yumeng> swp20: oh. sounds amazing. how did we pass the unittests without project_id in arq....
03:19:50 <xinranwang__> https://review.opendev.org/#/c/728014/
03:21:01 <Yumeng> xinranwang__: cool. just saw the update. will review again.^
03:21:15 <xinranwang__> lol
03:22:53 <Yumeng> ok. I nothing more important than review to mention today.
03:23:00 <Yumeng> I have
03:23:29 <swp20> Yumeng , should we add api for cleaning devices as there may be some inconsistent data in db?
03:23:30 <brinzhang> Yumeng, swp20: the project_id in some interface is make sense, agree to add
03:24:37 <Yumeng> can you share more info? what's the inconsistent data in db?
03:24:51 <swp20> brinzhang: ok, i will fix the tempest for the project_idpatch.
03:25:19 <brinzhang> swap20: cool, thanks
03:25:48 <Yumeng> swp20: but not in a hurry. you can also ping me in wechat once convenient.
03:25:50 <swp20> Yumeng, such as we discuss at PTG, Cyborg db data is inconsistent with Placement db
03:26:00 <Yumeng> aha.
03:27:19 <swp20> or maybe device table is inconsistent with deployables/attributes adn so on..
03:28:32 <swp20> cool. i will contact with u in wechat.
03:28:34 <Yumeng> from my understanding, if the goal is to solve the inconsistent data in db, it doesn't need to add an api, maybe just some data check, and cleaning by involing current API will make sense
03:28:55 <Yumeng> involking
03:29:49 <Yumeng> anyway, that's a case to case thing. we might need to discuss more details to decide if we need an API
03:30:11 <Yumeng> let's continue in wechat.:)
03:30:12 <swp20> ok, got it. thanks.
03:30:37 <Yumeng> so that's all for today. Thank you guys!
03:30:45 <Yumeng> let's wrap up this meeting and I'll see you all again next week.
03:30:51 <Yumeng> Bye.
03:30:58 <Yumeng> #endmeeting