Earlier  
Posted Nick Remark
#openstack-cyborg - 2018-04-04
14:06:38 Li_Liu Welcome Melissa!
14:06:50 Melissa_S Likewise :)
14:07:33 Sundar Hi Melissa, welcome! Good to have you here.
14:07:58 Melissa_S Thanks Sundar :)
14:08:51 kosamara #info kosamara
14:08:58 kosamara hi!
14:09:04 zhipeng Hey !
14:09:11 Yumeng__ #info Yumeng__
14:09:14 Yumeng__ Hi
14:09:23 zhipeng Hi Yumeng :)
14:09:42 Yumeng__ Hi zhipeng
14:10:16 chucksong Hi everyone!
14:10:37 zhipeng Hi chucksong
14:10:44 zhipeng don't believe we met before
14:10:52 zhipeng can you introduce yourself a little bit ?
14:11:12 chucksong yes, I am Chuck Song of Xilinx
14:11:27 chucksong Currently taking charge of the FaaS team
14:11:41 chucksong I suppose we are going to talk about the Cyborg project.
14:11:45 zhipeng FPGA as a service ?
14:11:48 chucksong yes
14:11:50 Sundar Hi Chuck, welcome!
14:11:56 Li_Liu Hi Chuck
14:12:12 Li_Liu We have a lot people joining :)
14:12:15 chucksong hi Sundar and LiLiu
14:12:41 zhipeng let's get into business :)
14:12:50 zhipeng #topic sub-team report
14:13:22 zhipeng I think we could start with release subteam
14:13:49 zhipeng #link https://review.openstack.org/558223
14:14:16 zhipeng #info cyborg has been added for milestone based release
14:14:44 zhipeng that was part of Rocky goal
14:15:05 Sundar Is Apr 19 the spec freeze day for Cyborg too?
14:16:08 zhipeng yes Sudnar the week of Apr 16 to Apr 20
14:16:23 zhipeng will be our very first deadline for spec freeze :)
14:16:44 Melissa_S what do you need delivered on the Xilinx side before freeze?
14:17:45 Melissa_S Pardon me - I am new to the project :)
14:18:00 zhipeng Melissa_S altho currently all the drivers are in-tree which means Xilinx driver also needs a spec
14:18:00 Li_Liu I think the major thing we need is Xilinx Device driver. But that doesn't have to be before spec freeze I think
14:18:17 zhipeng but driver spec freeze should be expected for Rocky MS-2
14:18:28 zhipeng which will be week of Jun 04-Jun08
14:18:30 Li_Liu Does the vendor specific driver need a spec Howord?
14:18:44 zhipeng Li_Liu at the moment yes
14:18:47 Li_Liu ok
14:19:10 zhipeng but yes as Li_Liu said, plz don't wait for the deadline :)
14:19:12 Sundar Not to jump the gun :) but could I ask folks to review https://etherpad.openstack.org/p/Cyborg-Nova-Multifunction so that I can update https://review.openstack.org/#/c/554717/ ?
14:19:38 Sundar This is based on the latest proposal from Nova community
14:19:51 zhipeng Sundar will go into that very soon :)
14:19:56 Li_Liu sure, I will take a look
14:20:03 zhipeng Let's finish the subteam report first
14:20:32 zhipeng Melissa_S chucksong plz refer to https://releases.openstack.org/rocky/schedule.html#r-release
14:20:41 zhipeng for the general milestone sched
14:21:10 Sundar Thanks, Li_Liu and Howard
14:21:45 Melissa_S Yes - thank you.
14:22:09 zhipeng Cyborg's specific sched could be found http://eavesdrop.openstack.org/meetings/openstack_cyborg/2018/openstack_cyborg.2018-03-14-14.07.html
14:22:28 zhipeng okey moving on
14:22:50 zhipeng Yumeng__ could you update the progress on your side ?
14:23:54 Yumeng__ I sent out a update about the high time synchronisation card on the mail list
14:24:05 Yumeng__ Pls check
14:24:29 Yumeng__ And launchpad migration to storyboard will be done soon.
14:25:00 zhipeng did you just send the clock driver proposal ?
14:25:05 zhipeng or maybe I missed it ?
14:25:50 Yumeng__ 1 second
14:26:36 zhipeng BTW in case everyone not follow the storyboard migration, plz refer to
14:26:39 zhipeng #link https://review.openstack.org/#/c/558327/
14:26:52 Yumeng__ https://docs.openstack.org/infra/manual/zuulv3.html
14:27:14 Yumeng__ Pls go ahead, I will send out the right link
14:27:28 zhipeng #link https://review.openstack.org/558821
14:28:02 zhipeng #info storyboard migration for cyborg
14:28:47 zhipeng thx Yumeng__ :) doc subteam has been very responsive and active
14:29:17 zhipeng Melissa_S chucksong Dutch signed on as co-lead for the driver sub-team
14:29:35 zhipeng does one of you guys maybe interested in take the role ?
14:29:52 zhipeng help update and coordinate the driver development for Rocky ?
14:30:31 Melissa_S Chuck and I can work on this
14:30:57 zhipeng great :)
14:31:11 Melissa_S I think you could put Chuck's name down - he is on the apps side
14:31:24 zhipeng duly noted :)
14:31:30 Melissa_S we will find the correct internal team for this project
14:31:35 zhipeng #topic critical rocky spec review
14:32:02 zhipeng first up, Sundar's Nova-Cyborg interaction spec
14:32:22 chucksong zhipeng, yes I will take the role
14:32:25 zhipeng Sundar could you help get everyone up to speed ?
14:33:26 Sundar Sure, Howard
14:33:27 zhipeng #info Sundar cyborg-nova spec discussion
14:33:30 zhipeng #link https://etherpad.openstack.org/p/Cyborg-Nova-Multifunction
14:33:46 zhipeng #link https://review.openstack.org/#/c/554717/
14:34:01 Sundar We had arrived at a way of representing FPGA components in Nova, and worklfows for all use cases based on hat representation, back in the PTG
14:34:26 Sundar The spec that Howard provided above codifies that
14:35:01 Sundar However, there was a race condition for one use case, which has been debated in the community. An analogy was found with a vGPU issue, and a joint solution was proposed
14:35:27 Sundar http://lists.openstack.org/pipermail/openstack-dev/2018-March/128888.html
14:35:41 zhipeng #link http://lists.openstack.org/pipermail/openstack-dev/2018-March/128888.html
14:36:03 Sundar Based on that, I worked out the details, in a way that should cover GPUs etc. but probably focuses more on FPGAs.
14:36:16 Sundar #link https://etherpad.openstack.org/p/Cyborg-Nova-Multifunction
14:36:38 Sundar I think this is a better solution than the PTG flow :) both for Nova and Cyborg
14:37:34 Sundar For Cyborg, this means that we don't need to keep accelerator usage details in Cyborg DB -- just use placement info :)
14:38:16 Sundar Eventually, if/when Nova supports preferred traits, maybe we don't need the weigher -- but we can evaluat that when it becomes real
14:38:54 Sundar For the new folks in Cyborg, this may sound cryptic :) I am open to walking folks through the proposal in a call if needed
14:39:27 Sundar Howard, do you want me to explain the crux of the proposal here?
14:39:28 Li_Liu Sundar, regardless, we still need to track individual assignment for each device in Cyborg, right?
14:40:04 zhipeng Sundar it would be great to have a brief walk through and the central take aways :)
14:40:07 Sundar Li_Liu: Yes, we will be handling acceleratr -PF/VF matching
14:40:46 Sundar That is not involve din the scheduling, because Placement counts accelerator resources

Earlier   Later