| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-cyborg - 2020-02-27 | |||
| 03:50:49 | shaohe_feng | But I know some Public cloud has support vGPU | |
| 03:51:06 | shaohe_feng | why not we support it? | |
| 03:51:18 | Sundar | brinzhang_: Ho do we distinguish Cyborg from Nova, if both can do all use cases for GPUs? | |
| 03:51:24 | Sundar | *How | |
| 03:52:42 | Sundar | Anyway, I am fine with vGPU support if you all want it. | |
| 03:53:06 | shaohe_feng | and the first propose for smart-nic generic solution in kernel by Redhat also make the device under mdev bus. | |
| 03:53:57 | brinzhang_ | Sundar, I think Cyborg can provide the vGPU architecture to Nova, so Nova can choose the vGPU to use | |
| 03:55:14 | chenke | I think supporting vgpu is not a bad thing for cyborg. So, I agree. | |
| 03:55:50 | Sundar | brinzhang_: I don't know what that means. Nova already has vGPU support without Cyborg. Cyborg can report VFs as attach handles (with SR-IOV) or mdevs as attach handles (with Nvidia GPUs). Both would work AFAICS. It is just that Nova and CYborg overlap in functionality. | |
| 03:56:18 | Sundar | As i said, I am fine with vGPU support in Cyborg. However, I'd rather close the current Nova patch series as is. I have added only 'PCI" as supported type in https://review.opendev.org/#/c/631245/56/nova/virt/libvirt/driver.py@5698 | |
| 03:56:45 | Sundar | One of you could add 'mdev' to that. | |
| 03:57:08 | shaohe_feng | We know k8s, CNI for network, CSI for storage, DPI for device/accelerators, let the right components do the right functions | |
| 03:57:36 | shaohe_feng | It said they want to add mdev in cyborg | |
| 03:58:25 | shaohe_feng | but cyborg's progress is slow, so that they can wait it and add it to nova. | |
| 03:59:21 | Yumeng | Sundar: I think the good side is for the Operators(users). They can manage FPGA and vGPU using just one component cyborg. configuring at cyborg.conf instead of vGPU at nova.conf and FPGA at cyborg.conf. | |
| 03:59:29 | chenke | I think the current nova's patch is high priority. | |
| 04:00:38 | Sundar | Yumeng and all: Sure. No objections to vGPUs or mediated devices in Cyborg. But it is going to take some effort probably to get that change into Nova | |
| 04:00:51 | shaohe_feng | Yes, the cyborg goal is similar to DPI in k8s, manage different accelerators. | |
| 04:01:48 | shaohe_feng | Nova change for cyborg mdev can later. Let's focus on PCI | |
| 04:01:50 | Sundar | Yumeng: My main comment would be to move 'mdev' type to attach handles from CPIDs. Also, it needs some testing. Do you plan to support 3rd party CI for GPUs? | |
| 04:01:59 | xinranwang | Can vGPU be attached like normal PCI device passthrough, if so, I think there is no conflict. | |
| 04:02:20 | Yumeng | xinranwang: it cannot. | |
| 04:03:22 | Yumeng | xinranwang: nova already support this. we can reuse those code. | |
| 04:03:46 | xinranwang | So VGPU has a different way to be attach, comparing with GPU PF? | |
| 04:04:08 | Yumeng | yes, different way. | |
| 04:04:09 | shaohe_feng | but cyborg can support it. The cloud provider can change it in nova by themselves downstream, if they really want to use cyborg. | |
| 04:04:33 | shaohe_feng | yes, MDEV is uuid in mdev bus. | |
| 04:04:43 | shaohe_feng | PCI is bdf in PCI bus. | |
| 04:04:50 | Yumeng | yes, exactly. MDEV is uuid in mdev bus. | |
| 04:05:14 | shaohe_feng | let cyborg support mdev firstly. | |
| 04:05:43 | shaohe_feng | leave the change in nova to cloud provider. | |
| 04:05:57 | shaohe_feng | they have ability to change it. | |
| 04:06:20 | shaohe_feng | at present. | |
| 04:07:51 | shaohe_feng | they will contribute to nova, if they think it is necessary | |
| 04:08:13 | Sundar | Yumeng: do you plan to support 3rd party CI for GPUs? | |
| 04:08:17 | Yumeng | ok. thanks for the suggestion. shaohe_feng. I think we can talk later about the vGPU. Sundar: pls go back to nova-interaction topic. | |
| 04:08:30 | Yumeng | Sundar: not yet for now. | |
| 04:08:45 | Sundar | ok | |
| 04:09:21 | Sundar | Nova interaction: currently held up on the requirement to support rescheduling. I have a single-node devstack setup, so cannot test rescheduling. | |
| 04:10:55 | Sundar | This does not mean multi-node for CYborg -- the functional tests that I wrote mock the Cyborg APIs, so only need it for the Nova side | |
| 04:11:52 | Sundar | If anybody has any ideas to handle this, please LMK | |
| 04:13:19 | Sundar | That's all I have. Anything else to discuss? | |
| 04:14:11 | Yumeng | ok. not from me. | |
| 04:15:05 | brinzhang | The functional tests by me, I think I should do some search, then to talk, others none of me. | |
| 04:16:09 | Sundar | brinzhang: yes, we need functional tests. You have already started that. What do we need to discuss? | |
| 04:17:22 | brinzhang | Sundar, I think there are need so many work I should consider, wait for enough, I will re take it in ircmeeting | |
| 04:18:10 | brinzhang | Some cleanup patch, can you +A? | |
| 04:18:27 | shaohe_feng | no from me. If you need some help, please ping me on Skype. | |
| 04:18:31 | brinzhang | https://review.opendev.org/#/c/707332/ | |
| 04:18:49 | brinzhang | https://review.opendev.org/#/c/709974/ | |
| 04:19:11 | Sundar | brinzhang: Got it. Sure | |
| 04:19:31 | Sundar | Good. Thanks a lot, everybody. Have a good day! | |
| 04:19:33 | brinzhang | These are not necessary in cyborg or cyborg-specs project, so I want to remove it and make they are clear | |
| 04:19:34 | Sundar | #endmeeting | |
| 04:19:35 | openstack | Meeting ended Thu Feb 27 04:19:34 2020 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 04:19:36 | openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-02-27-03.00.html | |
| 04:19:37 | openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-02-27-03.00.txt | |
| 04:19:38 | openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-02-27-03.00.log.html | |
| 04:19:47 | s_shogo | Only repot: The openstacksdk patch was merged! Thank you,all. https://review.opendev.org/#/c/679914/ | |
| 04:21:26 | Sundar | s_shogo: Congrats! That's good news. It was all your hard work. | |
| 04:21:41 | Sundar | (y) | |
| 04:22:00 | openstackgerrit | Brin Zhang proposed openstack/cyborg-specs master: Switch to sphinx-build and cleanup https://review.opendev.org/707332 | |
| 04:23:26 | brinzhang_ | Sundar: done, thanks | |
| 04:28:07 | openstackgerrit | Brin Zhang proposed openstack/cyborg master: Delete sandbox directory https://review.opendev.org/709974 | |
| 05:59:05 | openstackgerrit | Merged openstack/cyborg-specs master: Switch to sphinx-build and cleanup https://review.opendev.org/707332 | |
| 06:18:04 | openstackgerrit | Brin Zhang proposed openstack/cyborg-specs master: Remove releasenotes directory https://review.opendev.org/707641 | |
| 06:56:09 | openstackgerrit | Merged openstack/cyborg-specs master: Remove releasenotes directory https://review.opendev.org/707641 | |
| 08:43:51 | openstackgerrit | Merged openstack/cyborg master: Delete sandbox directory https://review.opendev.org/709974 | |
| 15:16:02 | openstackgerrit | Dan Smith proposed openstack/cyborg master: Always stop cyborg-api process uwsgi container https://review.opendev.org/710288 | |
| #openstack-cyborg - 2020-02-28 | |||
| 14:20:22 | openstackgerrit | Brin Zhang proposed openstack/cyborg master: Add obj_make_compatible() https://review.opendev.org/703291 | |
| 14:20:53 | openstackgerrit | Brin Zhang proposed openstack/cyborg master: WIP: Add description column to device_profiles db https://review.opendev.org/703285 | |
| 14:21:04 | openstackgerrit | Brin Zhang proposed openstack/cyborg master: WIP: Add 'description' field to the dp's object https://review.opendev.org/703293 | |
| #openstack-cyborg - 2020-02-29 | |||
| 09:34:12 | openstackgerrit | chenker proposed openstack/cyborg master: Remove useless interfaces in cond https://review.opendev.org/710553 | |
| 09:36:00 | openstackgerrit | chenker proposed openstack/cyborg master: Remove useless interfaces in cond https://review.opendev.org/710553 | |
| #openstack-cyborg - 2020-03-02 | |||
| 01:33:40 | openstackgerrit | chenker proposed openstack/cyborg master: Remove useless interfaces in cond https://review.opendev.org/710553 | |
| 03:10:46 | openstackgerrit | chenker proposed openstack/cyborg master: Remove useless interfaces in cond https://review.opendev.org/710553 | |
| 07:53:22 | openstackgerrit | YumengBao proposed openstack/cyborg master: Fix attach_handle_type of gpu device with improper value https://review.opendev.org/709433 | |
| 10:57:34 | openstackgerrit | Merged openstack/cyborg master: add support for multi node deployments to fake driver https://review.opendev.org/708674 | |
| 13:59:02 | openstackgerrit | Xinran WANG proposed openstack/cyborg master: [WIP]Microversion Support https://review.opendev.org/696860 | |
| 15:02:19 | openstackgerrit | sean mooney proposed openstack/cyborg master: move setting to devstack/settings https://review.opendev.org/708770 | |
| 15:34:30 | openstackgerrit | sean mooney proposed openstack/cyborg master: WIP add cyborg multinode job https://review.opendev.org/709641 | |
| 16:35:36 | openstackgerrit | Merged openstack/cyborg master: Add obj_make_compatible() https://review.opendev.org/703291 | |
| #openstack-cyborg - 2020-03-03 | |||
| 01:50:15 | Sundar | Hi chenke | |
| 01:50:30 | chenke | Hi | |
| 01:50:42 | chenke | https://review.opendev.org/#/c/673735/40/nova/tests/unit/compute/test_compute_mgr.py | |
| 01:51:03 | chenke | Do your means Dan's advice will report error? | |
| 01:51:08 | Sundar | Yes | |
| 01:51:28 | chenke | You can write like this:https://github.com/openstack/cyborg/blob/acbc64f3bedc4f4946c5d12ea4fd7dc3d941667a/cyborg/tests/unit/db/test_db_device_profile.py#L85-L87 | |
| 01:51:50 | Sundar | That's what I did | |
| 01:52:19 | Sundar | It fails saying it throws AcceleratorOp exception. | |
| 01:53:11 | Sundar | What is wrong with the current test? I am not trying to check that the code raises an exception. But the test case mocking will force an exception. So, we shouldn't need to assert for that. | |
| 01:55:32 | chenke | Emmm. Before that, I thought you were unfamiliar with the use of assertRaises, but it wasn't actually the problem. | |
| 01:56:16 | chenke | It seems I still need to read the code carefully. | |
| 01:58:27 | Sundar | ok, np, thanks | |
| 02:02:32 | chenke | About line 6289,6290 with self.compute._build_resources(*args): | |
| 02:02:47 | chenke | I think you should delete this. | |
| 02:06:43 | chenke | just remove the line raise exception.NovaException. | |
| 02:07:20 | openstackgerrit | Merged openstack/cyborg master: Remove useless interfaces in cond https://review.opendev.org/710553 | |
| 02:22:22 | Sundar | chenke: self.compute._build_resources is a context manager.When an exception is thrown inside the with statement, that causes a specific code path inside _build_resources_, which is where AcceleratorOp excepotion gets raised. | |
| 02:23:25 | chenke | I know. I means write like this: self.compute._build_resources(*args) | |