Posted	Nick	Remark
#openstack-cyborg - 2019-03-20
03:07:45	Sundar	Just pinged the rest of the team
03:08:26	Yumeng	# info Yumeng
03:08:28	Yumeng	hi all
03:09:00	Sundar	Hi Yumeng
03:09:09	Li_Liu_	Hi Yumeng
03:09:38	Sundar	Just 1 more minute, and then we'll get started
03:10:27	Sundar	#topic Stein Status
03:11:09	Sundar	#link https://review.openstack.org/#/q/status:open+project:openstack/cyborg+branch:master
03:11:36	Sundar	We seem to be in good shape. Start with https://review.openstack.org/#/c/574075/
03:12:07	Li_Liu_	is it merged yet?
03:12:23	Sundar	It is dependent on #link https://review.openstack.org/641529 . So not merged yet
03:12:39	Sundar	I need to ask Coco about the second patch
03:12:58	Li_Liu_	Need to follow up with coco then
03:13:20	Sundar	xinranwang: ANy update on https://review.openstack.org/#/c/624340/ ?
03:14:14	wangzhh	Sorry for late. Meeting conflict just now.
03:14:58	wangzhh	Anyone else?
03:14:58	Sundar	wangzhh: Hi, np
03:15:04	xinranwang	this is taken over by Coco, it's almost the make diff part.
03:15:34	xinranwang	As she said in Wechat group, may need more tests .
03:15:37	Sundar	xinranwang: Too much on Coco's plate? May be we can help her?
03:16:56	Sundar	Only 3 days to go
03:17:26	Li_Liu_	Sundar, I think we can add more test later
03:17:33	Li_Liu_	but get it in first
03:18:12	Li_Liu_	Hi Coco
03:18:30	Sundar	Hi Coco_gao
03:18:30	xinranwang	I am willing to help if needed
03:18:43	Coco_gao	Hi all
03:19:12	Coco_gao	I am working on fixing the bug, sorry for the joining late.
03:19:27	Sundar	Coco_gao: Could we help you by having Xinran focus on https://review.openstack.org/624340, while you focus on https://review.openstack.org/641529 ?
03:20:51	Coco_gao	I almost finish bug fix, will test it today.
03:21:30	Sundar	Coco_gao: for 641529? Great. If that gets in, wangzhh's patch gets merged too!
03:21:50	wangzhh	:) Great
03:21:50	Coco_gao	Both two
03:22:06	Coco_gao	Then add UT next
03:22:27	Li_Liu_	Coco_gao, we can add the UT after the RC cut
03:22:45	Sundar	Coco_gao: Li_Liu_: Agreed ^
03:22:54	Coco_gao	OK.
03:23:33	Sundar	#topic cyborg client
03:23:45	Sundar	What do we need to do for the client?
03:23:58	Sundar	Do we have any owners?
03:24:11	Li_Liu_	you mean the python client?
03:24:15	Sundar	Yes
03:25:02	Li_Liu_	Right now I think we should be good for acc and deployable's v1 APIs
03:25:29	Li_Liu_	1 extra thing I can add is the fpga programmmg command
03:25:36	Sundar	Has anybody tried it out of late?
03:26:02	Sundar	Li_Liu_: Thanks
03:26:03	xinranwang	Hmmm, i think the client is not complete
03:26:40	Li_Liu_	all we can do is listing and update right? xinranwang
03:27:03	xinranwang	just have accelerator list
03:27:28	Sundar	xinranwang: What is missing?
03:28:04	xinranwang	like accelerator show/ delete, and all related to deployables
03:28:27	Li_Liu_	https://git.openstack.org/cgit/openstack/python-cyborgclient/tree/cyborgclient/v1/deployables.py
03:28:54	Li_Liu_	This should do it right?
03:29:21	xinranwang	yes, but also missing shell part
03:29:48	Sundar	I think so. My suggestion is to let things be as they are, and take care of v2 APIs in Train. That will be quite some work
03:30:16	xinranwang	my question is: if we make v1 API deprecated, is it necessary to complete V1 client
03:30:17	Li_Liu_	ok sure
03:30:21	wangzhh	+1
03:30:56	Li_Liu_	not worth the effort and we don't have time for that neither...
03:31:11	Sundar	+1
03:31:47	Sundar	Any differing opinions?
03:31:54	Coco_gao	+1
03:31:58	Yumeng	nope from me +1
03:31:59	xinranwang	agree
03:32:09	Sundar	#agreed Leave the python client as is
03:32:15	Sundar	#topic Documentation
03:32:27	Sundar	Yumeng, thanks for bringing up documentation
03:32:38	Yumeng	Sundar,Li_Liu_: About the api-doc. admin/api.rst is already updated. However if new contributors wants to see how to make a api request, that might not be enough. they might want to see more details like this https://developer.openstack.org/api-ref/compute/?expanded=suspend-server-suspend-action-detail#suspend-server-suspend-action
03:33:12	Yumeng	Dowe want to have some doc-api-progress now? if so, we can do some work that needs to be done both for v1 and v2 , that maybe efficient. the work will include: api-ref creation, deployable api CURD.
03:33:25	Yumeng	if not, then just let it go. ^_^
03:33:30	Coco_gao	#info Coco_gao
03:34:53	Li_Liu_	sounds good to me
03:34:57	Sundar	Yumeng maybe api-ref creation first? We can do deployables if we have time
03:35:42	Sundar	Yumeng: Only 3 days to go, and we don't want you to kill yourself :)
03:35:42	Yumeng	Sundar: ok, agree. then I will do api-ref creation
03:35:47	Sundar	+1
03:35:54	Li_Liu_	+1
03:36:15	yikun	++
03:36:28	Yumeng	Sundar: lol, then maybe buy insurance first
03:36:30	Sundar	#AoB
03:36:33	Coco_gao	Do we have guides on how to write a driver?
03:37:08	Li_Liu_	can wangzhh help on this?
03:37:21	Sundar	Coco_gao: Good point. We should start compiling a list of missing docs.
03:37:50	wangzhh	Li_liu_, Of course.
03:38:18	Li_Liu_	Thanks a lot
03:38:21	Sundar	wangzhh: Thanks!
03:38:25	Coco_gao	thx
03:38:42	Sundar	China Mobile wants to do a demo of ONAP with Cyborg at MWC Shanghai in June :)
03:38:43	wangzhh	Do we have deadline this part?
03:38:58	Li_Liu_	I think that's T release target
03:39:16	Li_Liu_	but since we want to get vendor to work on driver in T
03:39:34	Li_Liu_	having it earlier would be great
03:39:48	wangzhh	It makes sense. I'll try my best.
03:39:55	Li_Liu_	thanks
03:40:10	Sundar	+1. Any issues if the doc gets submitted before Stein gets released?
03:40:29	Sundar	May be we could hold the merge till Stein release
03:41:58	Sundar	wangzhh: Please propose a draft and we can iterate over time.
03:43:12	wangzhh	Yep.
03:43:35	Sundar	wangzhh: Thanks again
03:43:38	Sundar	wangzhh: China Mobile wants to do a demo of ONAP with Cyborg at MWC Shanghai in June. They want to use Cyborg GPU driver for Tesla 100. Do you think the GPU driver can be validated by, say, May beginning?

Earlier   Later