| *** zhurong has quit IRC | 00:36 | |
| *** chenke has joined #openstack-cyborg | 01:57 | |
| *** s_shogo has joined #openstack-cyborg | 02:47 | |
| *** Sundar has joined #openstack-cyborg | 02:53 | |
| openstackgerrit | YumengBao proposed openstack/cyborg master: Update api-ref: deprecate v1 add v2 device_profile section https://review.opendev.org/690539 | 02:53 |
|---|---|---|
| *** Yumeng has joined #openstack-cyborg | 02:59 | |
| *** shaohe_feng has joined #openstack-cyborg | 03:01 | |
| Sundar | #startmeeting openstack-cyborg | 03:01 |
| openstack | Sundar: Error: Can't start another meeting, one is in progress. Use #endmeeting first. | 03:01 |
| Sundar | #endmeeting | 03:02 |
| *** openstack changes topic to "Pending patches (Meeting topic: openstack-cyborg)" | 03:02 | |
| openstack | Meeting ended Thu Oct 24 03:02:02 2019 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 03:02 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-17-03.01.html | 03:02 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-17-03.01.txt | 03:02 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-17-03.01.log.html | 03:02 |
| Sundar | #startmeeting openstack-cyborg | 03:02 |
| openstack | Meeting started Thu Oct 24 03:02:26 2019 UTC and is due to finish in 60 minutes. The chair is Sundar. Information about MeetBot at http://wiki.debian.org/MeetBot. | 03:02 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 03:02 |
| *** openstack changes topic to " (Meeting topic: openstack-cyborg)" | 03:02 | |
| openstack | The meeting name has been set to 'openstack_cyborg' | 03:02 |
| Sundar | #topic Who's here | 03:02 |
| *** openstack changes topic to "Who's here (Meeting topic: openstack-cyborg)" | 03:02 | |
| Sundar | o/ | 03:02 |
| shaohe_feng | seems you did not finish last meeting | 03:02 |
| s_shogo | Hi all | 03:02 |
| s_shogo | #info s_shogo | 03:02 |
| shaohe_feng | o/ | 03:03 |
| chenke | #info chenke | 03:03 |
| chenke | 0/ | 03:03 |
| chenke | hi all~ | 03:03 |
| shaohe_feng | #info shaohe_feng | 03:03 |
| Yumeng | #info Yumeng | 03:03 |
| Sundar | Hi y'all | 03:03 |
| Sundar | #topic Agenda | 03:03 |
| *** openstack changes topic to "Agenda (Meeting topic: openstack-cyborg)" | 03:03 | |
| Sundar | https://wiki.openstack.org/wiki/Meetings/CyborgTeamMeeting#Agenda | 03:03 |
| shaohe_feng | what does y' means? | 03:03 |
| Sundar | :) "you all" | 03:03 |
| *** xinranwang has joined #openstack-cyborg | 03:04 | |
| Sundar | Anything to add to the agenda? | 03:04 |
| xinranwang | hi all | 03:04 |
| Sundar | Hey xinranwang | 03:04 |
| Sundar | #topic Summit and PTG | 03:04 |
| *** openstack changes topic to "Summit and PTG (Meeting topic: openstack-cyborg)" | 03:04 | |
| Sundar | First, re. project update at the summit, I got clarification after the last meeting. | 03:05 |
| shaohe_feng | will miss the Team dinner :') | 03:05 |
| Sundar | It is actually on. But it is not recorded as video. | 03:05 |
| Sundar | Sorry to have mis-stated before I got the clarification. Yumeng, I will create the slides and we can work on them. | 03:06 |
| Sundar | shaohe_feng: we will get you a good dinner in Beijing ;) | 03:07 |
| shaohe_feng | Thanks | 03:07 |
| shaohe_feng | you will visit Beijing? | 03:07 |
| shaohe_feng | after PTG | 03:07 |
| Sundar | PTG: #link http://ptg.openstack.org/ | 03:08 |
| Yumeng | Sundar: ok. | 03:08 |
| Sundar | All the info for PTG is here. | 03:08 |
| Sundar | It turns out that nearly all projects get only a table, not a separate room. | 03:08 |
| Sundar | http://ptg.openstack.org/ptg.html Diable is a table, not a room | 03:09 |
| Sundar | *Diablo | 03:09 |
| Sundar | Not sure how well that works, but that's the case for most projects | 03:09 |
| Sundar | Etherpad: The current etherpad (https://etherpad.openstack.org/p/cyborg-ptg-ussuri) will need to be merged into the final one: https://etherpad.openstack.org/p/shanghai-ptg-cyborg | 03:11 |
| chenke | We have three days | 03:11 |
| Sundar | I will take care of it, once we have enough content in the current etherpad | 03:11 |
| Sundar | chenke: True -- but that is a misleading because Wednesday includes the summit too | 03:12 |
| chenke | Yes. | 03:12 |
| Sundar | I have activities on Wednesday, such as office hours for Cyborg | 03:12 |
| Sundar | Also, the project onboarding is part of the PTG this time. So, we will be spending some time talking to newcomers who want to contribute to Cyborg | 03:13 |
| Sundar | I suspect it will really be 2 days of discussion | 03:13 |
| chenke | It seems that time is sufficient. | 03:14 |
| Sundar | Finally: the team photo shoot is part of the PTG: on Thu at 2 pm (after lunch). But that is only 10-15 min | 03:14 |
| Sundar | Does anybody have any questions or comments on the PG or the summit? | 03:15 |
| Sundar | *PTG | 03:15 |
| Sundar | OK, moving on | 03:17 |
| Sundar | #topic Doc patches | 03:17 |
| *** openstack changes topic to "Doc patches (Meeting topic: openstack-cyborg)" | 03:17 | |
| Sundar | The documentation has merged into stable/train -- yay! Thanks, xinranwang, Yumeng, chenke and all for proposing/reviewing it. | 03:18 |
| chenke | Great. | 03:18 |
| Sundar | Now Train has truly left the station for Cyborg. ;) | 03:19 |
| Sundar | I see soem new patches from Yumeng and others for docs. Yumeng, you don't expect this to be backported to Train, right? | 03:20 |
| Yumeng | sundar: no, no need to backported to Train. what do you think? | 03:21 |
| Sundar | We need to check how this differs from other API doc that got merged | 03:21 |
| Yumeng | https://review.opendev.org/#/c/690539/ | 03:21 |
| Sundar | Your patch has more detail. | 03:22 |
| Sundar | OK, anything else on docs? We need similar content for ARQs, etc., I suppose. | 03:23 |
| Yumeng | I took keystone api-ref as a reference, they have deprecated and current APIs | 03:23 |
| Yumeng | here is the keystone api-ref: https://docs.openstack.org/api-ref/identity/ | 03:24 |
| Yumeng | take a look | 03:24 |
| Sundar | Cool. | 03:24 |
| Sundar | #topic Other patches | 03:24 |
| *** openstack changes topic to "Other patches (Meeting topic: openstack-cyborg)" | 03:24 | |
| Sundar | https://review.opendev.org/#/c/685542/ | 03:25 |
| Sundar | My question is, should we expend more effort on the ksa_adapter and older methods, or should we aim to move to openstacksdk? | 03:25 |
| chenke | According to eric's advice. He suggest we use openstacksdk. | 03:27 |
| Sundar | So, what should be the status of this patch? | 03:28 |
| chenke | It is easier to use than before. | 03:28 |
| chenke | If your devstack env test ok. | 03:29 |
| chenke | I suggest to merge this patch and this https://review.opendev.org/#/c/690509/2. | 03:29 |
| Sundar | Ok, I'll try it out. The 2nd one is definitely an improvement | 03:30 |
| xinranwang | Please review this patch too, after async bind merged, GPU is not supportted. https://review.opendev.org/#/c/688239/ | 03:30 |
| chenke | Ye. | 03:30 |
| chenke | Sundar: Thanks. | 03:30 |
| xinranwang | And there is a bug in placement client, please review this one too :) https://review.opendev.org/#/c/688231/ | 03:31 |
| Yumeng | Sundar: I think we can merge patch 685542, no more effort on ksa_adapter and openstacksdk for now. | 03:31 |
| Sundar | xinranwang, Yumeng: yes to both | 03:32 |
| Sundar | Any other patches to be highlighted today? I know there are a bunch of others that need review. | 03:33 |
| Sundar | #topic AoB | 03:35 |
| *** openstack changes topic to "AoB (Meeting topic: openstack-cyborg)" | 03:35 | |
| Sundar | If there is nothing else, we can get back 25 minutes. | 03:35 |
| shaohe_feng | Will we support Orchestrator components(or system IAAS sotrware tools) also want to leverage accelerators? | 03:36 |
| shaohe_feng | These components may run on the same host node with VM guest. | 03:36 |
| shaohe_feng | Seems this is out of cyborg control. (should we add a reservation mechanism for them?) | 03:36 |
| Sundar | Could you elaborate or provide links? | 03:36 |
| shaohe_feng | I have attend a meeting about accelerators. | 03:37 |
| shaohe_feng | Get the information, not only VM need accelerator, but maybe some other applications run on the host node also need accelerators | 03:38 |
| Sundar | It is not common in an OpenStack cluster to have VMs and non-VM processes running on the same compute node. | 03:39 |
| shaohe_feng | I meam the non-VM processes are openstack required components | 03:39 |
| shaohe_feng | Not sure you know RDT. | 03:40 |
| Sundar | Yes, I know RDT. Are you referring to RMD? | 03:40 |
| shaohe_feng | No | 03:40 |
| shaohe_feng | RDT | 03:40 |
| shaohe_feng | the use case is: | 03:40 |
| shaohe_feng | Divide cache into 2 parts. | 03:41 |
| Sundar | Resource Director technology (RDT) is a set of Intel processor features for cache/memory management. They can be used in OpenStack in different ways. | 03:41 |
| shaohe_feng | 1 parts are reserved for system IAAS sotrware run on the node | 03:41 |
| shaohe_feng | another parts are reserved for VM. | 03:42 |
| Sundar | That part is outside Cyborg though -- it relates to cache/memory and should be configured by the operator. How does it relate to accelerators? | 03:42 |
| shaohe_feng | We defined, cyborg can only allocate accelerators to nova, right? | 03:43 |
| shaohe_feng | I means like RDT. | 03:43 |
| shaohe_feng | we have 5 accelerators. | 03:43 |
| shaohe_feng | we divided into 2 parts. | 03:44 |
| shaohe_feng | 1 parts have 1 acc, it is out of cyborg control. It is reserved for system IAAS sotrware. Admin allocate them | 03:45 |
| shaohe_feng | another parts have the remain 4 acc, cyborg allocate them to nova. | 03:45 |
| shaohe_feng | similar like RDT use case. | 03:46 |
| shaohe_feng | on we can simple define cyborg. | 03:46 |
| shaohe_feng | if cyborg run on a host node. we not allow to allocate accelerators to system IAAS sotrware any more. | 03:47 |
| shaohe_feng | we can note this on cyborg docs. | 03:47 |
| Sundar | If the admin wants to keep aside some accelerators, he/she can update Placement to mark that resource provider as in-use. | 03:47 |
| Sundar | From Cyborg's POV it created the RP but Placement will never pick that RP because it sees it as in-use | 03:48 |
| xinranwang | As we know, cinder has enable qat driver, I guess shaohe want to know if Cyborg can help cinder manage this kind of accelerator, is that what you want say, shaohe_feng? | 03:49 |
| shaohe_feng | not only cinder | 03:49 |
| shaohe_feng | also for some VPN on the host. | 03:49 |
| chenke | what does POV means? | 03:50 |
| shaohe_feng | and storage | 03:50 |
| xinranwang | yes, cinder is just an example | 03:50 |
| Sundar | chenke: point of view | 03:50 |
| shaohe_feng | DPDK, SPDK also need . | 03:50 |
| shaohe_feng | yes, it is an example | 03:50 |
| shaohe_feng | so does that means, let cyborg run first before other system IAAS sotrware, it report resources to placement first | 03:50 |
| shaohe_feng | and amdin reserve some accelerators, and start the remain system IAAS sotrwares | 03:51 |
| shaohe_feng | and amdin reserve some accelerators to the remain system IAAS sotrwares | 03:52 |
| *** TxGirlGeek has joined #openstack-cyborg | 03:52 | |
| Sundar | Not sure where this discussion is headed. If Cinder is using a device, operator should not configure Cyborg driver for that device. If the admin wants to dedicate some networking/storage devices for DPDK/SPDK and they were claimed by Cyborg, the admin could still mark them in placement, as I said. | 03:52 |
| shaohe_feng | We can add this usage as a guideline in cyborg doc. | 03:53 |
| shaohe_feng | Sundar: can we add this to the cyborg doc? | 03:54 |
| Sundar | shaohe_feng: It may be better to wait for concrete usage by operators and get their feedback, instead of adding hypothetical use cases in the docs -- they could be confusing. | 03:54 |
| shaohe_feng | maybe DPDK as example. | 03:54 |
| s_shogo | I also interested in disabling to allocate some accelerators, as maintenance. | 03:55 |
| Sundar | s_shogo: Only when no VM is using that device, right? | 03:55 |
| s_shogo | Sundar: yes | 03:55 |
| shaohe_feng | Maybe we had better not allow DPDK to use accelerators any more. This can make things more simple. | 03:56 |
| xinranwang | Currently, operator plays this role. | 03:56 |
| shaohe_feng | For example, had better to avoid DPDK to use some network accelerator. | 03:57 |
| shaohe_feng | Then things is simple. | 03:57 |
| Sundar | s_shogo: Would the admin-Placement method work for now? We had plans to add health reporting and enabling/disabling devices. Once we have the /v2/devices API, we could add this functionality. Seems ok? | 03:57 |
| Sundar | shaohe_feng: Cyborg does not support network resources today. | 03:58 |
| shaohe_feng | yes, just a example. | 03:59 |
| shaohe_feng | QAT can be used for some network accelerations | 03:59 |
| shaohe_feng | and it can be used by DPDK. | 03:59 |
| shaohe_feng | not sure cyborg will support QAT | 04:00 |
| s_shogo | Sundar: Yes, the approach is good. I want to discuss the methods continuously with example usecase for maintenance and that's requirements. | 04:00 |
| shaohe_feng | Or cyborg just plan to manager only VM used accelerators. | 04:00 |
| Sundar | shaohe_feng: Do you have specific knowledge that somebody wants to have several QATs in one host, some for VMs, and some for DPDK on the host? | 04:01 |
| Sundar | s_shogo: Sure, let's make this a discussion topic at the PTG. Your I could create an etherpad with the problem statement and some ideas. | 04:02 |
| Sundar | *You or I | 04:03 |
| shaohe_feng | usually，they use kolla to deploy the system IAAS sotrwares include (placement), How do kolla know what's accelerators in placement？ | 04:04 |
| s_shogo | Sundar : Thanks, I'll update the etherpad (part of the topic is already written) | 04:04 |
| Sundar | My suggestion: let's get Nova integ done and make something functional. Then we can start talking about more advanced use cases. | 04:05 |
| Sundar | We are over the time. Is there anything else? | 04:05 |
| shaohe_feng | Yes, the meeting tells use DPDK can use accelerators. But they does not say DPDK can not live with VM on same host. | 04:06 |
| shaohe_feng | Ok. | 04:07 |
| shaohe_feng | seems, we did not support much advanced use cases at present. | 04:07 |
| shaohe_feng | step by step | 04:07 |
| Sundar | shaohe_feng: Let's take this offline. | 04:08 |
| Sundar | Good. Thanks a lot, everybody. Have a good day! | 04:08 |
| Sundar | #endmeeting | 04:08 |
| *** openstack changes topic to "Pending patches (Meeting topic: openstack-cyborg)" | 04:08 | |
| openstack | Meeting ended Thu Oct 24 04:08:20 2019 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 04:08 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-24-03.02.html | 04:08 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-24-03.02.txt | 04:08 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-10-24-03.02.log.html | 04:08 |
| s_shogo | If you have time, a specific topic related to intel PAC > mainly intel folksI'm trying to build devstack env with intel-PAC board, but the latest devstack config seems to be lack of main body of OPAE driver, is out of scope in devstack? | 04:10 |
| s_shogo | Not in a hurry topic. | 04:10 |
| Sundar | s_shogo: Yes, the OPAE kernel driver is not installed by devstack. Mainly because that requires the right kernel headers etc. and can go wrong in many ways | 04:10 |
| Sundar | Do you have the right documentation to install OPAE drivers? | 04:11 |
| s_shogo | Sundar: OK, thanks. I got it. | 04:11 |
| s_shogo | Sundar: I always refer the official manual and works well. | 04:11 |
| Sundar | Ok, great | 04:12 |
| s_shogo | Thanks all, good evening and lunch, bye | 04:14 |
| *** s_shogo has quit IRC | 04:14 | |
| chenke | bye~ | 04:14 |
| *** TxGirlGeek has quit IRC | 04:23 | |
| *** shaohe_feng has quit IRC | 04:26 | |
| *** Sundar has quit IRC | 04:26 | |
| *** efried has quit IRC | 05:04 | |
| *** efried has joined #openstack-cyborg | 05:11 | |
| *** links has joined #openstack-cyborg | 05:18 | |
| *** xinranwang has quit IRC | 07:33 | |
| openstackgerrit | YumengBao proposed openstack/cyborg master: Update api-ref: deprecate v1 add v2 device_profile section https://review.opendev.org/690539 | 07:57 |
| *** chenke has quit IRC | 09:41 | |
| *** Yumeng has quit IRC | 11:47 | |
| openstackgerrit | chenker proposed openstack/cyborg master: Enable openstack-cover-jobs in zuul check for cyborg https://review.opendev.org/690950 | 12:40 |
| openstackgerrit | chenker proposed openstack/cyborg master: Bugfix: _needs_programming lack one situation process https://review.opendev.org/690983 | 14:28 |
| openstackgerrit | chenker proposed openstack/cyborg master: Bugfix: _needs_programming lack one situation process https://review.opendev.org/690983 | 14:35 |
| *** links has quit IRC | 14:43 | |
| *** TxGirlGeek has joined #openstack-cyborg | 15:12 | |
| *** TxGirlGeek has quit IRC | 15:21 | |
| *** TxGirlGeek has joined #openstack-cyborg | 16:22 | |
| openstackgerrit | Merged openstack/cyborg master: Update docstring for apply_patch https://review.opendev.org/690517 | 18:09 |
| *** TxGirlGeek has quit IRC | 21:26 | |
| *** TxGirlGeek has joined #openstack-cyborg | 21:29 | |
| *** TxGirlGeek has quit IRC | 21:58 | |
| *** TxGirlGeek has joined #openstack-cyborg | 22:01 | |
| *** TxGirlGeek has quit IRC | 22:27 | |
| *** TxGirlGeek has joined #openstack-cyborg | 22:31 | |
| *** TxGirlGeek has quit IRC | 22:34 | |
| *** TxGirlGeek has joined #openstack-cyborg | 22:38 | |
| *** TxGirlGeek has quit IRC | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!