Wednesday, 2017-04-05

*** crushil has quit IRC01:10
*** mikeH has quit IRC06:28
*** joseppc has joined #openstack-cyborg08:13
*** joseppc has joined #openstack-cyborg08:13
*** jkilpatr has quit IRC10:37
*** jkilpatr has joined #openstack-cyborg11:12
*** mikeH has joined #openstack-cyborg11:39
*** zhipeng has joined #openstack-cyborg12:33
-openstackstatus- NOTICE: The Gerrit service on is being restarted to address hung remote replication tasks, and should return to an operable state momentarily12:50
*** adreznec has joined #openstack-cyborg13:37
*** crushil has joined #openstack-cyborg13:43
*** NokMikeR has joined #openstack-cyborg14:04
*** NokMikeR has quit IRC14:07
*** joseppc has quit IRC14:23
*** mikeH has quit IRC14:39
*** mikeH has joined #openstack-cyborg14:39
*** zhipeng_ has joined #openstack-cyborg14:45
*** zhipeng has quit IRC14:46
*** crushil has quit IRC14:50
*** crushil has joined #openstack-cyborg14:51
zhipeng_let's go15:03
zhipeng_#startmeeting openstack-cyborg15:03
openstackMeeting started Wed Apr  5 15:03:38 2017 UTC and is due to finish in 60 minutes.  The chair is zhipeng_. Information about MeetBot at
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:03
openstackThe meeting name has been set to 'openstack_cyborg'15:03
zhipeng_#info zhipeng crushil15:03
zhipeng_who else do we have at the meeting ?15:04
jkilpatrsorry I'm late15:10
zhipeng_no problem,just saw your patch15:12
zhipeng_#topic BP discussion15:12
zhipeng_I must apologize that due to the cloudnativecon week, I have not been attentive to the api BP15:12
zhipeng_will get to it starting tmr I hope15:13
zhipeng_crushil jkilpatr could you guys walk through the updates ?15:13
jkilpatrcrushil, you first?15:14
crushilUpdated the bp for the generic driver implementation and trying to incorporate all the changes that were requested15:14
crushilDid some reviews on other BPs as well and waiting for the BP owners to come back with a response15:15
zhipeng_crushil cool man, any outstanding issues at the moment ?15:15
crushilThe generic driver implementation bp is at
crushilzhipeng_, Nope, just need more review comments atm15:16
zhipeng_great :)15:16
zhipeng_jkilpatr how about your updates ?15:18
jkilpatrSo I still need to address a review from a Friday. It's not going to be the conductor's responsibility to attach accelerators, so I need to remove it from that part of the loop15:18
jkilpatrthen I think I'll be taking the suggestion to move setup responsiblities out of the conductor.15:18
jkilpatrfor the time being I guess we can maintain a directory of setup playbooks and document how to use them.15:19
jkilpatrI also need to actually write the caching stuff we talked about in the last meeting. Might try and get through that today15:19
jkilpatrthat's all.15:19
zhipeng_i don't remember we had discussion on moving out setup functionalities15:20
zhipeng_is it from the review ?15:20
jkilpatrzhipeng_, it's in the review comments.15:20
zhipeng_my network is dreadfully slow now ... so comments from _gryf ?15:21
jkilpatrRoman doesn't like the idea of having a program run off and install it's own dependencies. Which I can understand.15:21
zhipeng_i thought we had a understanding that agent could do the ansible stuff15:21
zhipeng_okey i will dig into Roman's comment later15:23
jkilpatrThere are arguments from both sides, on one hand ease of setup, on the other the potential conseqences of cyborg running wild with root.15:24
jkilpatrin the end we're planning on making the same playbooks and just having a human press the button instead of the agent.15:24
zhipeng_so if we do the setup as a out-of-band operation15:24
zhipeng_but I do have some scenarios that maybe still the agent should do the setup, despite of the security concerns15:26
zhipeng_for example for some dataplane virtualization techniques, you could compose a set up virtual functions into one bigger function15:27
zhipeng_a case wouldbe to compose a virtual FW on a intelligent NIC card15:27
zhipeng_using different dedicated small virtual networking functions which will do q-in-q, switching respectively15:28
zhipeng_jkilpatrt I think the overall question is that , could we have a compromise between the two sides ?15:29
jkilpatrzhipeng_, I think we could find some sort of compromise I'm just not sure where to draw the line15:31
zhipeng_I was thinking the line could be where the root privilege is required15:32
zhipeng_if the root is required on the host side, then maybe it is good, as Roman suggested, to leave to the humans15:32
jkilpatrthat sounds reasonable enough.15:32
zhipeng_but if it is root privilege on the smart device15:32
zhipeng_then we could, at the moment, assume that the conductor could perform the setup with root privilege15:33
zhipeng_it has security concerns as well, no doubt, but i think we could deal with it later on15:33
jkilpatryou're talking about permissions levels on the accelerator itself? What is the "smart device" ?15:34
zhipeng_yes i was. smart device like intelligent NICs I referred to15:35
zhipeng_there are FPGA enabled devices that you could program the "micro" virtual functions15:36
zhipeng_deploy them in different ways to compose different functionalities15:36
zhipeng_it is not widely used yet, but I think it would be a major use case later on15:36
jkilpatrI know FPGA's can be deployed like that but I never thought about user levels on them.15:37
zhipeng_for example in edge computing, where you could dynamically compose your network dataplane15:37
zhipeng_that was just my thinking15:37
jkilpatrit makes sense, just never considered it that way.15:38
zhipeng_if we could have cyoborg agent setup an ansible playbook on a rasberry pie15:38
zhipeng_that would be supper awesome :P15:39
jkilpatrbut why?15:39
zhipeng_(just thinking out loud)15:39
zhipeng_never mind the pie example15:40
zhipeng_so I think I will also comment on the patch, to see what's _gryf's take on this15:40
zhipeng_if in the end , the dynamic composition FPGA use case is still too far fetch, then we could just have the regular playbook procedure15:41
jkilpatrI think that's fine for now. Maybe add some notifications to the operator that things need to be setup.15:43
jkilpatrI think we could detect new accelerators without root15:43
zhipeng_yes that is ture15:43
zhipeng_true ...15:43
zhipeng_fat fingers15:43
zhipeng_#topic AoB15:44
zhipeng_okey any other buisness or topics15:44
crushilNot from me15:46
jkilpatrnot from me either15:48
zhipeng_okey thank you guys for the great work on the bp patches, let's conclude the meeting earlier today :)15:48
zhipeng_see you guys on gerrit :P15:48
openstackMeeting ended Wed Apr  5 15:48:58 2017 UTC.  Information about MeetBot at . (v 0.1.4)15:49
openstackMinutes (text):
*** zhipeng_ has quit IRC15:49
*** jkilpatr_ has joined #openstack-cyborg16:04
*** jkilpatr has quit IRC16:06
*** crushil has quit IRC17:01
*** crushil has joined #openstack-cyborg17:46
*** crushil has quit IRC17:53
*** crushil has joined #openstack-cyborg17:53
*** jkilpatr_ has quit IRC17:54
*** jkilpatr_ has joined #openstack-cyborg18:08
*** crushil has quit IRC18:54
*** crushil has joined #openstack-cyborg18:55
*** mikeH has quit IRC19:28
*** joseppc has joined #openstack-cyborg19:34
*** crushil has quit IRC20:20
*** crushil has joined #openstack-cyborg20:40
*** jkilpatr_ has quit IRC21:42
*** mikeH has joined #openstack-cyborg21:57
*** jkilpatr_ has joined #openstack-cyborg22:20
*** mikeH has quit IRC22:52
*** mikeH has joined #openstack-cyborg23:23

Generated by 2.14.0 by Marius Gedminas - find it at!