Wednesday, 2018-08-22

*** annabelleB has quit IRC00:31
*** annabelleB has joined #openstack-cyborg00:36
*** annabelleB has quit IRC00:49
*** edleafe has quit IRC02:46
*** edleafe has joined #openstack-cyborg02:47
openstackgerritSundar Nadathur proposed openstack/cyborg master: Cyborg device discovery specification.
*** nguyenhai93 has joined #openstack-cyborg06:04
nguyenhai93Hi cybord team, please merge the OpenStack Release Bot patches related to Rocky branch06:05
nguyenhai93so that we can process the patches related to python3-first goal. Thanks06:06
*** shaohe_feng has joined #openstack-cyborg08:50
*** shaohe_feng has quit IRC08:54
*** wangzhh has joined #openstack-cyborg14:00
*** Sundar has joined #openstack-cyborg14:03
SundarDo we have a meeting today?14:06
*** zhipeng has joined #openstack-cyborg14:06
*** xinran has joined #openstack-cyborg14:07
wangzhhHi, anybody else?14:10
SundarI am here, waiting for past 10 minutes :)14:10
*** Coco_gao has joined #openstack-cyborg14:10
zhipeng#startmeeting #openstack-cyborg14:10
openstackMeeting started Wed Aug 22 14:10:51 2018 UTC and is due to finish in 60 minutes.  The chair is zhipeng. Information about MeetBot at
Coco_gao#info Coco_gao14:10
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:10
*** openstack changes topic to " (Meeting topic: #openstack-cyborg)"14:10
openstackThe meeting name has been set to '_openstack_cyborg'14:10
Sundar#info Sundar14:11
wangzhh#info wangzhh14:11
zhipeng#info Howard14:11
xinran#info xinran14:11
sum12#info sum1214:11
*** Li_Liu has joined #openstack-cyborg14:11
Li_LiuHi guys14:11
Li_Liusorry for the late, there was a fire drill in my building14:11
wangzhhMaybe something wrong with zuul just now?14:12
wangzhhHi Li.14:12
Coco_gaoHi Li.14:12
zhipeng#topic Rocky finish up14:12
SundarHi, Li, hope all is good14:12
*** openstack changes topic to "Rocky finish up (Meeting topic: #openstack-cyborg)"14:12
zhipengCoco_gao status update on the ovo patch ?14:13
Coco_gaoI updated the code, but still not tested yet.14:14
*** efried is now known as efried_goatin14:14
SundarWhat is our objective in Rocky -- to commit specific patches, or to be able to bring up a VM with some specific use case?14:15
*** jiapei has joined #openstack-cyborg14:16
*** shaohe_feng_robo has joined #openstack-cyborg14:16
Li_LiuSundar, to bring up the VM, we really need to finalized the spec with nova-cyborg interaction14:17
Coco_gaoHi, Sundar, I read your driver-agent api spec, just don't have time to reply, I am sorry.14:17
shaohe_feng_robo#info shaohe_feng_robo14:17
SundarLi_Liu: Thanks, that's what I thought.14:18
Li_Liu#info Li_Liu14:18
SundarCoco_gao: NP, understand14:18
Li_LiuCoco_gao, I will review your patch today14:18
Coco_gaoI still need to finish some functions, I commit this patch on Monday just want everyone to know the OVO first.14:20
Coco_gaoIn case you have some commits for OVO.14:21
Coco_gaocomments, I am sorry.14:21
wangzhhSundar, some data structure  in Coco's patch is different from your new spec.14:21
shaohe_feng_robowill we need to merge this patch this release?14:21
Li_LiuI also have a patch coming up within couple days. it involves calling Agent from Cyborg to do programming14:21
wangzhhCould u help to review it?14:21
Li_Liuzhipeng, can we still merge changes to Rocky yet?14:21
SundarLi_Liu: I think we need to align on the os-acc flow for programming too.14:22
zhipengyes we still can14:22
zhipengprojects are doing rc2 now14:22
Sundarwanhzhh: Sure, I should review Coco's patch. I need to catch up14:22
wangzhhSundar: Thx.14:23
xinranI am trying to make sure if we should finish nova/cyborg interaction in Rocky14:23
Li_LiuSundar, yes, i agree, at least I wanna get the whole flow with rpcapi ready so that os-acc can call it when it wants to.14:23
SundarHonestly, IMHO, we are making piecemeal changes without the big picture :) It is important to get things done of course but, we hope it is in the right direction that will take us forward with Nova and others14:23
Coco_gaoSundar, I don't know well on traits, so I don't know my OVO can meet the placement requirements.14:24
SundarLi_Liu: We should first agree on the os-acc flow, right?14:25
SundarCoco_gao: NP, I can help you. Power architecture developers want to support custom traits from the driver, as I mentione din os-acc spec14:25
Sundar*sorry, discovery spec14:25
Coco_gaoBy now, we don't have DB to store traits on Cyborg side.14:26
wangzhhSundar, not only flow but also how to handle different hypervisor, different platform, etc.14:26
Li_LiuSundar, but currently feature I am implementing does not involve os-acc at all, its the REST api to allow fpga programming14:27
SundarCoco_gao: at least for now, we can store custom traits as attributes14:27
SundarLi+Liu: So, there is no expectation that we should use the exact same flow for os-acc, right?14:27
Li_LiuCoco_gao, yes, put them in the attributes list14:28
Sundarwangzhh: Agreed14:28
wangzhhWe need to resolve: data to placement 2. os-acc flow(include some other things) 3. nova intergration14:29
*** annabelleB has joined #openstack-cyborg14:29
Li_LiuSundar, yes, you can say that. but I think we could reuse that piece of code in os-acc anyways14:30
Li_Liuwangzhh, exactly14:30
xinranwangzhh yes14:30
shaohe_feng_roboxinran:  good at all 1, 2, 314:30 data to placement -- Coco_gao   2. os-acc flow(include some other things) -- wangzhh   3. nova intergration -- xinran... are these assignments right?14:31
shaohe_feng_roboLi_Liu: all can be xinran .14:32
Coco_gaoI think so14:32
xinranI am starting design a new API for nova intergration14:33
shaohe_feng_robofor she already do these things, we can avoid duplicated work.14:33
xinranwith lock and batch alloc support14:33
Li_Liuwangzhh, can you help xinran when you have some time?14:33
wangzhhOf course.14:33
sum12are there any implementation specific reviews  of os-acc already ?14:33
shaohe_feng_roboxinran:  great14:34
Li_Liushaohe_feng_robo, it's great that xinran is the expert on all of these :)14:34
zhipengplz do remember that we develop and implement what we agreed for Rocky, at the moment14:34
zhipenganything beyond that is for design discussion starting at the Denver PTG14:35
shaohe_feng_roboLi_Liu: she's new api is for the intergration14:35
wangzhhsum12, Yes,  but It's piecemeal.14:35
Li_Liuso I am assuming there will a batch of patch reviews coming in from you right? xinran14:35
zhipengdo not start development on those design specs that are not merged14:35
wangzhhzhipeng, Agree.14:35
shaohe_feng_robobut pleacement report depend on driver14:36
xinranI just trying to understand if all these should be done in Rocky or not?14:36
shaohe_feng_robonot sure all drivers are ready.14:36
sum12wangzhh: couple you point me to a piece ? I was expecting it to be a separate project14:36
shaohe_feng_roboSundar:  what's the status of the new driver?14:36
xinranyes placement report depend on driver,  did coco already have some works on it ?14:37
shaohe_feng_roboLi_Liu: for interaction,  edleafe  can help on nova side,14:37
sum12found it14:38
xinranLi_Liu:  yes, is it urgent?14:39
wangzhhsum12, OK. :)14:39
Li_Liuxinran, just do it on your own pace, but do you have a eta tho?14:39
openstackgerritNguyen Hai proposed openstack/python-cyborgclient master: Update reno for stable/rocky
shaohe_feng_roboLi_Liu: who report to placement?14:40
Coco_gaowhat kind of data placement need? I just store all of the data.14:40
shaohe_feng_roboagent or conductor?14:40
Sundarshaohe_feng: The driver developer is apparently busy with other stuff. He has reached a stage where rpm and deb packages for OPAE libraries get packaged with the tree. I'll sync up with him and help in testing.14:40
Coco_gaobesides, the driver-specific traits is not support right now.14:41
openstackgerritNguyen Hai proposed openstack/python-cyborgclient master: import zuul job settings from project-config
openstackgerritNguyen Hai proposed openstack/python-cyborgclient master: switch documentation job to new PTI
shaohe_feng_roboSundar: good. thanks .14:41
openstackgerritNguyen Hai proposed openstack/python-cyborgclient master: add python 3.6 unit test job
xinranLi_Liu:  i am afraid that it's quite hard to implement all in this release ...14:41
sum12xinran: agreed14:41
Coco_gaoyeah, I think the work to heavy for xinran14:41
Li_Liushaohe_feng_robo, agent report to cyborg, then cyborg uses the placement client to report to placement, right?14:42
Coco_gaosince the time is limited before Rocky14:42
xinranCoco_gao:  have you already do some work on placement report which related to your current work?14:42
shaohe_feng_roboLi_Liu: OK.14:42
*** annabelleB has quit IRC14:42
xinranLi_Liu:  yes, cyborg use placement client to report resource14:43
Li_Liuxinran, I know, as I said, just do it on your own pace. but could you give estimations on them?14:43
Coco_gaoI didn't report to placement by now. I14:43
*** shaohe_feng_robo is now known as shaohe_feng_14:44
Coco_gaoI just collected the driver data, store in db.14:44
shaohe_feng_every card is a sub-provider14:45
wangzhhCoco, After diff. :)14:45
xinranLi_Liu:  I will probably submit first version of new api allowing nova/cyborg interaction next week14:45
Li_LiuCoco_gao, that should be good for now. next we will figure out during the reporting step, how to invoke placement client apis14:46
shaohe_feng_Li_Liu:  the placement report does not need many  code14:46
Li_Liuxinran, sounds good14:46
shaohe_feng_just need to know the  deloyable well.14:46
Li_Liushaohe_feng_, yup14:47
xinranto be clear, did we agreed to improve the current API, I mean, add a new API like DeployableActions to do alloc and dealloc?14:47
wangzhhxinran, I suggest to imporve them in S release.14:47
shaohe_feng_xinran: you can introduce your new API briefly14:47
*** annabelleB has joined #openstack-cyborg14:48
Coco_gaoLi_Liu, where can I store the acc_topology data?14:48
xinranwangzhh:  yes that what i desired14:49
xinranshaohe_feng_:  sure14:49
SundarFolks, I want to clarify that all commits now should be considered provisional. We may need to change them substantially to meet the use cases and Nova requirements14:50
Li_LiuCoco_gao, the acc_topology is not to be saved tho. It helps you to structure the deployable relationships14:51
sum12wangzhh: do you plan to pull in any patches for the Rocky release for os-acc ?14:51
Li_LiuCoco_gao, in case it's not clear to you, we can discuss it further once you feel better14:52
shaohe_feng_Sundar: will you  work on the image upload ? auto parser the bitstream files, and generate the glance metadata?14:52
wangzhhsum12: No.14:52
xinranI am trying to add a new API representing  all actions of deployables e.g. allocation/deallocation.  for example:  def post(self, body) with body like {"FPGA":"vendor_id":"","type":""...,"GPU","vendor_id":""...}. And in this method, deployable_list will be call with filters.14:53
wangzhhshaohe_feng_, Before upload image, we need download first, in agent. Do you think so?14:54
SundarShaohe: Sure.14:54
xinranTo support batch alloc/dealloc, there will be a loop to call deployable_list(...,filters) that's my current solution. Does anyone of you have some better suggestion?14:56
wangzhhxinran, for batch alloc/dealloc, we should have new APIs, and I think it's better to have a spec.14:57
wangzhhWe should disscuss how to handle failure.14:58
Li_Liuwangzhh, you mean the looping case?14:58
wangzhhLi_Liu, No, for the new design.14:59
Coco_gaoLI_Liu, I just mess up,can we list the remaining work and the timeline for that?14:59
xinranthere is no spec currently15:00
wangzhhOh, Sorry. Misunderstand u. xinran.15:00
Li_LiuCoco_gao, you mean the remaining work on the agent reporting?15:01
wangzhhCurrent implemetion15:01
Coco_gaoon cyborg rocky.15:01
wangzhhdoesn't need spec.15:01
SundarI need to leave now for the next call. Thank you all. Bfn!15:01
*** Sundar has quit IRC15:02
Li_Liu Coco_gao, for rocky, in my opinion, I hope we can merge the ovo stuff that you are currently working on. anything on top of that is a bonus :)15:04
Li_Liuplus, xinran's new api and my programming api15:05
Li_Liuwangzhh, what's left on your plate for Rocky?15:05
xinranthere are 1. driver discovery and data structure. 2. placement report. 3. nova integration (new api)  and what else ?15:05
*** efried_goatin is now known as efried15:06
wangzhhAgent diff and save, related to Coco's patch.15:06
Li_LiuI thought the placement client report is finished didn't it?15:06
Li_Liuwangzhh, sounds good15:07
Coco_gaobut I am afraid, Sundar's Spec have difference with my code.15:07
wangzhhLi_liu, you mean we have finished reporting to placement?15:08
xinranSundar's spec mention that diff should be done in conductor. Did we have an agreement on it?15:08
Coco_gaothe api15:08
wangzhhNot yet. xinran said , she is working on it.15:08
wangzhhxinran, not in current release.15:09
Li_Liuthese code from shaohe is the placement client right?15:09
Coco_gaonope, zhipeng think it should be done on agent in this release15:09
xinranok thanks15:09
shaohe_feng_Li_Liu: yes, client15:10
wangzhhYes, but we should intergrate them in agent.15:10
Coco_gaoI think so15:10
Li_Liuyup, it's just the matter when to use the client15:10
shaohe_feng_wangzhh: we do not need to download15:10
Coco_gaoI think it should be used after diff.15:11
wangzhhshaohe_feng_, so what if we want to program?15:11
xinranCoco_gao:  yes, exactly15:12
wangzhhShouldn't we download bitstream first?15:12
Coco_gaoBye, I need to go bed.15:13
shaohe_feng_Coco_gao: have a good dream.15:13
Li_Liubye, Coco_gao, take a good rest15:14
Li_Liushaohe_feng_, wangzhh, are you guys talking about the same step here?15:14
wangzhhMaybe we can hold a zoom meeting later?15:16
shaohe_feng_we should not be in same page15:16
shaohe_feng_we mya be not in same page15:16
Li_Liuya, it's too late now, let's wrap up. I think we are clearer on the work assignment for R. let's get it done15:18
Li_Liuxinran, let us know if you need help, especially wangzhh and me can off load some of your work, right wangzhh :)15:19
wangzhhOf course. :)15:19
xinranok, thanks guys :)15:20
Li_Liuone last thing. remember to post your topics for PTG no matter if you are coming to Denver or not15:20
Li_Liuand let us know if you are coming to PTG15:21
wangzhhNot sure. :)15:22
Li_Liuok, have a good night/day wherever you are :)15:23
*** openstack changes topic to "A zuul config error slipped through and caused a pile of job failures with retry_limit - a fix is being applied and should be back up in a few minutes"15:23
openstackMeeting ended Wed Aug 22 15:23:57 2018 UTC.  Information about MeetBot at . (v 0.1.4)15:24
openstackMinutes (text):
*** Li_Liu has quit IRC15:25
*** zhipeng has quit IRC15:26
*** annabelleB has quit IRC15:28
*** openstackgerrit has quit IRC15:31
*** efried has quit IRC15:36
*** efried has joined #openstack-cyborg15:36
*** annabelleB has joined #openstack-cyborg15:44
*** efried is now known as efried_rollin16:14
*** annabelleB has quit IRC16:23
*** jiapei has quit IRC16:35
*** annabelleB has joined #openstack-cyborg16:42
*** openstackgerrit has joined #openstack-cyborg16:48
openstackgerritcoco-Gao proposed openstack/cyborg master: Add driver-side oslo_versioned objects.
*** Coco_gao has quit IRC17:19
*** nguyenhai93 has quit IRC17:25
*** sum12 has quit IRC17:25
*** dims has quit IRC17:25
*** wangzhh has quit IRC17:28
*** dims_ has joined #openstack-cyborg17:36
*** xinran has quit IRC17:52
*** annabelleB has quit IRC18:09
*** ChanServ has quit IRC18:16
*** ChanServ has joined #openstack-cyborg18:22
*** sets mode: +o ChanServ18:22
*** efried_rollin is now known as efried19:47
openstackgerritSundar Nadathur proposed openstack/cyborg master: Cyborg/Nova/Glance interaction in compute node, including os-acc.
*** ChanServ has quit IRC22:49
*** ChanServ has joined #openstack-cyborg23:03
*** sets mode: +o ChanServ23:03

Generated by 2.15.3 by Marius Gedminas - find it at!