03:01:43 #startmeeting openstack-cyborg 03:01:44 Meeting started Wed May 15 03:01:43 2019 UTC and is due to finish in 60 minutes. The chair is Sundar. Information about MeetBot at http://wiki.debian.org/MeetBot. 03:01:45 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 03:01:47 The meeting name has been set to 'openstack_cyborg' 03:01:52 #topic Roll call 03:01:54 #info Li_Liu 03:02:02 #info wangzhh 03:02:03 #info xinranwang 03:02:03 #info Sundar 03:02:17 #info ikuo_o 03:02:24 Hi ikuo_o. 03:02:35 Hi wangzhh. 03:02:43 Please see agenda for today at: https://wiki.openstack.org/wiki/Meetings/CyborgTeamMeeting 03:02:58 Does anybody have anything else to add to it? 03:03:25 Thanks Sundar. 03:03:54 #topic Cyborg specs reviews 03:04:02 #info Yumeng 03:04:07 Hi all 03:04:30 Hi Yumeng 03:04:35 Morning Yumeng. 03:04:37 Top 4 specs in: https://review.opendev.org/#/q/status:open+project:openstack/cyborg-specs 03:05:16 Can we get these 4 reviewed over the next few days? 03:05:26 I and my colleagues are checking "Specification for Cyborg APIs Version 2". 03:05:31 I will try my best 03:05:44 Thanks, ikuo_o, Li_Liu. 03:05:50 Sure. 03:06:15 Great. Thanks wangzhh 03:06:23 #topic Nova spec review 03:06:43 Lots of discussion in https://review.opendev.org/#/c/603955/ 03:06:56 lol 03:07:02 a lot of comments 03:07:27 Does Huawei want a standard accelerator name for Ascend? 03:07:35 Only FPGA and GPU for now 03:07:46 Yes 03:07:58 Ascend is on our agenda 03:08:25 Does it need to be mentioned in the nova spec? 03:08:30 Li_iu: Good. Just to be sure, instead of CUSTOM_ASCEND, you want ASCNED to be added to os-resource-classes, right? 03:08:51 you can check this with Yikun 03:08:55 Yes, efried has said that explicitly. 03:09:13 #info Yikun 03:09:25 Hi Yikun, just in time :) 03:09:32 sorry for late 03:10:23 NP. For Ascend, can we use a custom resource class name like CUSTOM_ASCEND? Or do you want to standardize it? We are standardizing only FPGA and GPU as of now 03:11:59 Is 'Ascend' popular like GPU or FPGA? 03:12:16 I don't know much about ascend... 03:12:30 FPGA GPU is the device type, but ascend maybe is not a standard type just a series, so I think maybe ascend should be cover in some thing like TPU_ASCEND or something 03:12:52 Perhaps it is best to leave it as a custom RC for now. 03:12:54 TPU is very google specific 03:13:08 sorry, NPU 03:13:18 If we propose a standard name, it should be added to the spec, justified and also proposed as a Nova patch. 03:13:31 yikun, I believe you have an AI chip and a security-related accelerator? 03:14:08 NPU can also mean Network processing Unit. NNP = neural network processor 03:14:41 for now, only AI chip, security-related accelerator comes from intel qat. 03:15:47 yikun: OK. You can think about it. I am hoping to converge the Nova spec by May end, while we still have the momentum and attention. 03:16:06 Sundar: ok, sure 03:16:21 #topic Pilot merge plan 03:16:48 I understand some of you find the merged patch to be very large to review. Is that the reason why we are not getting many reviews? 03:18:27 Another way of asking: can you all review the code soon, or suggest ways to get the reviews done? 03:18:29 :) 03:19:02 btw, regarding changing from polling to event driven between nova and Cyborg, I just modify based on the pilot branch right? 03:19:20 or submit a seperated patch? 03:20:05 Li_Liu: I am using the pilot code for various demos from now till mid-June. It will be best not to disrupt the pilot patches. 03:20:06 I'm fine with big patch or smaller one. 03:21:01 Sundar, sure, I will keep it clean in a different patch 03:21:29 Li_Liu and all: I am bit worried about what Nova guys are suggesting: there is no quantitative data on how it will affect Cyborg, right? FPGA programming can take a few seconds, and we are reducing that window of wait if we move the bind to the compute node 03:21:32 for now I will try to review the spec and pilot branch 03:21:56 Li_liu, submit a seperated patch is OK. And pls add sundar's commit as a dependence. 03:22:31 We should try to get some idea of how much the window gets reduced in different scenarios, like launch 1 VM, launch 2 VMs together, etc. 03:23:48 The original suggestion at Stein PTG to do it this way was for multiple VM case. It was from Dan Smith, but he hasn't weighed in on thhe current round of dicussion 03:24:43 Are they blocking the spec merging? 03:24:57 :Sundar, we only have 3 patches now, right? https://review.opendev.org/#/q/status:open+project:openstack/cyborg+branch:feature/cyborg-nova-pilot+topic:cyborg-nova-pilot 03:25:11 if no more objections , we can merge asap 03:25:33 Well, John is saying that we either do it the way he suggests or justify doing it the current way. I don't have data either way. 03:26:31 Li_Liu: will you be putting a Nova patch too? 03:26:55 there will be changes in NOVA for sure 03:27:31 You can submit that as a separate patch too. I can try to do some measurements in a live system with the combined set before we decide fully. 03:27:40 ok 03:28:45 Yumeng: It is really just https://review.opendev.org/626057 . 03:29:46 Yikun and I still have some comments. 03:30:14 wangzhh: Good. Please go ahead with the comments. 03:30:37 I will also add some UT. 03:31:07 Great. 03:31:43 Also, the pilot code makes some changes to the db schema. Just raising that point, if it wasn't clear. 03:32:34 Networked FPGAs may make further changes but, unfortunately, I don't have full clarity on that yet. 03:33:09 BTW, I found this doc on db migration: https://docs.openstack.org/glance/latest/contributor/database_migrations.html 03:33:42 It talks of migration with expansion, contraction, etc. We aren't doing that, right? 03:34:28 Btw, I trying to figure out placement report things, thers is one concern which is, the conductor will only receive the driver_device object with all infos insides. So conductor need to parse it to get the deployable, attribute etc. is it reasonable? 03:35:17 xinranwang: The driver ovo is well-defined, so conductor should be able to get to the inner fields. 03:35:48 Sundar, migration script supports some of these feature. 03:36:14 Ok, good 03:36:17 yes, it should be. No info missing, but I just found it's a little complex for conductor 03:36:46 anyway, i have done like that. I will submit a patch soon, please review it if you got time 03:37:01 Sure 03:37:12 Yep. 03:37:23 sounds good 03:38:37 #action All to review the Cyborg specs, Nova spec and Cyborg pilot code (as much as possible) over the next week or so 03:38:43 #topic Python 3 migration 03:39:15 Bringing this up because I don;t know how much change this will need. SO may be it is best to discuss early. 03:39:32 Sundar: [for the question in migrations script] the migration with expansion, contraction doc is just talk some rule, you can take any other project db migrations as example, we already have some script to complete add change filed in db. 03:40:04 yikun: Good. 03:40:52 ikuo_o: Glad that NTT has taken up Python 3 migration. Do you need any help in doing some investigation to see how much change is needed? (You don;t need to make the changes just yet.) 03:40:53 Yikun, so it will be just the matter of running a script? 03:41:25 Sundar: Yes, now my colleagues are checking the code, and there are some questions. 03:41:26 Li_Liu: I think so 03:41:35 I will send them later. 03:41:56 ikuo_o: Great, thanks 03:42:12 #topic AoB 03:42:24 Anything else for today? 03:43:09 Do we have zoom meeting tmr? 03:43:52 Li_Liu: No. I am actually attending a conference this week (even today). Besides, I don;t have any specific agenda for tomorrow. Do you or others have any? 03:44:08 nope 03:44:17 which conf are you going? 03:44:48 Open Source Technical Symposium 03:45:06 cool :) 03:45:34 This is conference season :) More coming up in June 03:45:55 Thanks for a productive meeting, folks :) 03:45:57 Thanks sundar. I will invite my colleague for next IRC meeting! 03:46:13 Sure, ikuo_o. 03:46:27 Have a good day, everybody! (It is day for Li_Liu too :) ) 03:46:38 Thank you Sundar :P 03:46:40 Have a good day, all! 03:46:46 Bye, all . 03:46:47 #endmeeting