Wednesday, 2017-04-12

*** crushil has quit IRC00:12
*** jkilpatr_ has quit IRC00:21
*** jkilpatr_ has joined #openstack-cyborg00:32
*** crushil has joined #openstack-cyborg01:07
*** crushil has quit IRC02:20
*** crushil has joined #openstack-cyborg02:29
*** crushil has quit IRC03:44
*** crushil has joined #openstack-cyborg04:17
*** crushil has quit IRC04:48
*** crushil has joined #openstack-cyborg05:01
*** crushil has quit IRC05:16
*** jkilpatr_ has quit IRC05:35
*** jkilpatr_ has joined #openstack-cyborg05:47
*** adreznec has quit IRC06:55
*** adreznec has joined #openstack-cyborg06:56
*** joseppc has joined #openstack-cyborg07:30
*** joseppc has joined #openstack-cyborg07:30
*** jkilpatr_ has quit IRC07:34
*** jkilpatr_ has joined #openstack-cyborg07:35
*** jkilpatr_ has quit IRC07:43
*** jkilpatr_ has joined #openstack-cyborg07:43
*** jkilpatr_ has quit IRC09:06
*** jkilpatr_ has joined #openstack-cyborg09:07
*** zhipengh[m] has quit IRC09:11
*** ttk2[m] has quit IRC09:11
*** jkilpatr_ has quit IRC09:11
*** jkilpatr_ has joined #openstack-cyborg09:12
*** ttk2[m] has joined #openstack-cyborg09:13
*** zhipengh[m] has joined #openstack-cyborg09:34
*** jkilpatr_ has quit IRC10:49
*** jkilpatr_ has joined #openstack-cyborg11:20
*** mikeH_ has joined #openstack-cyborg12:04
*** mikeH_ has quit IRC12:59
*** crushil has joined #openstack-cyborg13:03
*** mikeH has joined #openstack-cyborg13:03
*** EugeneC has joined #openstack-cyborg13:36
*** NokMikeR has joined #openstack-cyborg13:37
*** EugeneC has left #openstack-cyborg13:38
*** EugeneC has joined #openstack-cyborg13:41
*** EugeneC has quit IRC13:50
*** ewchan has joined #openstack-cyborg13:51
*** zhipeng has joined #openstack-cyborg13:59
ewchanHi. This is Eugene Chan from Huawei. I would like to be an observer of the meeting today.14:01
zhipengcool :)14:02
NokMikeRHI Eugene, welcome.14:02
zhipengour meeting will start one hour later14:02
zhipengbut we could always chat :P14:02
ewchanoh. thanks. it must be the dst changes. as it listed as 10am est in the cyborg site.14:04
zhipengi forgot to change lol14:05
zhipengthx for reminding14:05
ewchanno problem :)14:05
NokMikeRwas the meeting normally held at 15:00 UTC?14:06
zhipengNokMikeR yes I think at least for the Pike cycle14:06
*** zhipeng has quit IRC14:25
*** zhipeng has joined #openstack-cyborg14:35
*** zhipeng has quit IRC14:54
*** zhipeng has joined #openstack-cyborg14:57
*** jkilpatr_ has quit IRC14:59
*** jkilpatr has joined #openstack-cyborg15:00
*** ewchan has quit IRC15:05
*** ewchan has joined #openstack-cyborg15:05
zhipengh[m]Hey guys15:06
zhipengh[m]My home network is very unstable15:07
zhipengh[m]#startmeeting openstack-cyborg15:07
openstackMeeting started Wed Apr 12 15:07:26 2017 UTC and is due to finish in 60 minutes.  The chair is zhipengh[m]. Information about MeetBot at
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:07
openstackThe meeting name has been set to 'openstack_cyborg'15:07
zhipengh[m]#topic BP outstanding issues15:07
*** zhipeng has quit IRC15:08
_gryfI think we lost zhipengh[m]15:16
crushil_gryf, I think so too15:16
jkilpatrso I updated my review Id appreciate if someone could look at it15:16
_gryfjkilpatr, will do.15:16
crushiljkilpatr, I will too15:16
jkilpatrhas anyone else done another patchset I should look at?15:17
crushiljkilpatr, Can you please remove the WIP from the commit message and give it a proper commit message?15:17
jkilpatroh sure that's easy.15:17
_gryfmy bp still needs some attention. esp in context of its correctness15:17
jkilpatr_gryf, link?15:18
crushiljkilpatr, _gryf Can you look at please?15:18
_gryfall cyborg work ^^15:18
_gryfand that's mine ^^15:18
_gryfcrushil, yup, I'll have it on my queue15:19
crushil_gryf, I already gave it a first look15:19
_gryfcrushil, yes, I appreciate it :)15:19
crushil_gryf, Can you address those comments?15:19
_gryfI will.15:20
crushil_gryf, Thanks15:20
_gryfI just have limited time for doing this15:20
jkilpatrso for the time being we are going to direct passthrough attachment of hardware?15:20
_gryfsince my division have different goals now, so that I have to do it at my free time15:20
jkilpatrI don't think there are any other options that we can implement for a first release.15:20
_gryfjkilpatr, that should be easy15:21
_gryfmaybe that was already discussed, and I didn't grasp it - does cyborg supposed to spin VMs?15:22
jkilpatrfrom what I understand no15:24
jkilpatrfrom what I understand we can't touch the VM's at all without seriously screwing with Nova we have to let nova do everything15:24
jkilpatrmeaning schedule, spin up, and maybe call out to us to attach if it allows.15:24
_gryfthat what I thought, hence my questions on your BP (cyborg conductor/agent)15:25
_gryfI wasn't on PTG, that's why I'm wondering if there was some architectural decisions made regarding interaction between cyborg and rest of openstack services15:26
jkilpatr_gryf, now that I know of, we should probably actually ask the nova team about this15:26
jkilpatrget clear outlines for what we can and can't do, because I've got nothing but hearsay15:27
_gryfI guess, that we can at least build a prototype with minimal changes in Nova, just to sell the idea15:28
_gryfalso communicate with nova team about that using irc/mailing list might be helpful15:29
_gryfmaybe reserving a slot for that on upcoming summit would be good (if possible, since it is not ptg)15:29
jkilpatrI may or may not be there I haven't gotten final word yet. I didn't get a talk in :(15:31
*** joseppc has quit IRC15:31
* _gryf will not attend summit, unless he'll find a position in other company who will let him to do so :)15:32
jkilpatrthere's been some chatter about the placement API, we need look into that.15:34
_gryfyes there is. currently it is qualitative aspect discussed (implementation already has started)15:37
_gryfand also nested resource providers is the next topic15:37
_gryfespecially the last feature might be really interesting for us in context of exposing VF from accelerators (I assume that GPU also using virtual functions for virtualization)15:38
jkilpatrso nova will have an api to support virtual functions?15:39
jkilpatralso zhipengh[m] is less than 1 light minute away, looks like Saturn will know about his messages before we do.15:40
jkilpatrfun fact Saturn is just over 1 light hour away.15:40
_gryfjkilpatr, not sure about pci subsystem, but AFAIK plan was to completely map old resource handling with placement API som my cautious guess is yes15:42
_gryf*about pci subsystem plans*15:42
jkilpatrmight work with some things, but I get the feeling a lot of work would be required to make drivers compatible with that sort of thing, like manufacturer side work.15:44
_gryfjkilpatr, in which context?15:46
_gryflike, in exposing right part of the virtualized (or not) subsystem to the vm via libvirt/xen/whatever?15:46
_gryfsince in the end we have to have a way to make VM to utilize accelerator15:47
_gryfpassing pci device is the simplest way, but its implementation in Nova is quite horrible15:47
jkilpatr_gryf, ok maybe I'm confused are we talking about an api for direct pci attachment or an api to mux multiple vm's to a single physical resource15:47
_gryfi guess both15:48
_gryfsome accelerators might be consumed only by one vm, since the amount of accelerated functions is… one15:48
_gryfin that case one of the possible way for attachment to the VM is through PCI15:49
_gryfor sort of SR-IOV-like, which basically are the same as PCI passthrough15:50
_gryfotoh, we can have an accelerator which provides several VF (or some other way, like exposing devices on /dev filesystem)15:51
_gryfso there would be many-to-many relation between accelerators and VMs15:51
_gryfwhat I was talking about was a way to directly attach some accelerated function (using pci pass, vf pass, device pass) to the hypervisor15:52
_gryfon top of that we can start building "drivers" for cyborg15:53
jkilpatrfor some accelerators we need to start with 1-1 vm accelerator and wait for others to develop muxing15:54
jkilpatrother things you can just go ahead and do it, for example most network acceleration doesn't care15:54
_gryfevery cyborg driver have to have knowledge how it may be attached to the vm15:54
_gryfI guess network acceleration is the other type of acceleration, since we do not pass any accelerated functions to the VMs directly, but put such VM within accelerated networking functions, which neutron is handling15:56
jkilpatrI think it's still under the scope of cyborg to make sure those functions are working15:56
jkilpatreven if it's not a neutron thing15:56
jkilpatrand won't that be a wonderful pattern breaker when it comes to driver design.15:56
crushilI think we should end the meeting and we can keep talking about all this after the meeting?15:57
_gryfI guess only zhipengh[m] can stop the meeting15:57
crushilWe can try15:57
jkilpatrhope zhipengh[m] is ok15:58
jkilpatranyawys matrix should cache this for him and he can comment when he's back online15:58
_gryfok. thank you guys. Let me update my bp regarding crushil comments, and I'll jump into another round of review.16:01
*** NokMikeR has quit IRC16:03
*** ewchan has quit IRC16:07
zhipengh[m].... Network sucks today16:09
jkilpatrwelcome back!16:09
zhipengh[m]I've been cut off from time to time16:10
zhipengh[m]Let me go through the minutes later lol, thank you guys for keeping up the good work16:11
zhipengh[m]Let me end the meeting first just in case ...16:11
openstackMeeting ended Wed Apr 12 16:11:23 2017 UTC.  Information about MeetBot at . (v 0.1.4)16:11
openstackMinutes (text):
*** mikeH has quit IRC16:44
*** mikeH has joined #openstack-cyborg16:45
*** crushil has quit IRC17:55
*** crushil has joined #openstack-cyborg18:02
*** mikeH has quit IRC18:05
*** jkilpatr_ has joined #openstack-cyborg18:49
*** jkilpatr has quit IRC18:50
*** jkilpatr_ has quit IRC19:41
*** jkilpatr_ has joined #openstack-cyborg19:54
*** crushil has quit IRC21:00
*** crushil has joined #openstack-cyborg21:24
-openstackstatus- NOTICE: Restarting Gerrit for our weekly memory leak cleanup.21:25
*** crushil has quit IRC21:37
*** jkilpatr_ has quit IRC22:15
*** jkilpatr_ has joined #openstack-cyborg22:49
*** crushil has joined #openstack-cyborg23:04
*** jkilpatr_ is now known as jkilpatr23:10

Generated by 2.14.0 by Marius Gedminas - find it at!