| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-cyborg - 2019-06-05 | |||
| 03:41:26 | ikuo_o | Thanks. should we choose the two option (fake GPU/generic driver (with tempest) OR fake Cyborg API (functional test)) now? | |
| 03:41:41 | xinranwang | tempest test runs in real env, how to use fake driver with tempest | |
| 03:42:18 | Sundar | xinranwang: the fake driver will return driver OVOs and implement the same API calls as a regular driver | |
| 03:42:39 | Sundar | You mean you cannot put up a VM with a fake driver? | |
| 03:43:28 | xinranwang | yes, we cannot boot VM with accelerators with fake driver | |
| 03:44:08 | Sundar | ikuo_o: We have to implement _some_ tests for the Nova patches to merge. It looks like the 3rd party tempest may take sometime and it is not clear that Nova folks will accept that alone | |
| 03:44:57 | ikuo_o | I got it. | |
| 03:45:20 | Sundar | xinranwang: there may be some way to fake that success too? It has been suggested an an option by Nova folks too. E.g. Line 20 of https://etherpad.openstack.org/p/ptg-train-xproj-nova-cyborg | |
| 03:46:51 | Sundar | Anyways, may be we can take this up in tomorrow's Zoom meeting, because we have only 15 minutes left for today | |
| 03:47:16 | xinranwang | Ok | |
| 03:47:20 | Sundar | #topic Cyborg specs | |
| 03:47:33 | Sundar | Device/driver discovery spec: https://review.opendev.org/#/c/593726 | |
| 03:48:03 | Sundar | We have already implemented it, though the spec is not merged. Do we need the spec? | |
| 03:48:53 | yikun | any different between specs and impletation? If not I think we didn't need it anymore | |
| 03:49:19 | ikuo_o | yikun +1 | |
| 03:49:29 | xinranwang | +1 | |
| 03:49:58 | Sundar | The spec was probably obsolete anyway :) So there may be differences. But that may not be relevant anyway. | |
| 03:50:46 | Sundar | Do the +1s mean 'we do not need it' or 'if no difference, we don't need it' ? | |
| 03:51:28 | ikuo_o | Thanks Sundar. Then the spec seems unnecessary | |
| 03:51:43 | ikuo_o | we do not need it +! | |
| 03:52:00 | Sundar | Good, thanks ikuo_o. I like that answer :) | |
| 03:52:21 | Sundar | I already abandoned it. I'll remove it from Nova spec as well | |
| 03:52:55 | Sundar | specs/index.rst: https://review.opendev.org/#/c/658498/ I think yikun already did that. SO, abandon this? | |
| 03:53:56 | yikun | https://review.opendev.org/#/c/651061/9/doc/source/index.rst@12 | |
| 03:53:58 | yikun | Yes | |
| 03:54:17 | yikun | I fixed it in the other merged patch. | |
| 03:54:25 | Sundar | Thanks, yikun. Done. | |
| 03:54:48 | Sundar | Please review these specs: https://review.opendev.org/#/c/602978/, https://review.opendev.org/#/c/605237/ | |
| 03:55:46 | yikun | OK, add to my review list | |
| 03:55:57 | Sundar | For device profiles, some more changes _may_ be needed in the future for 'nested magic'. See http://eavesdrop.openstack.org/irclogs/%23openstack-placement/%23openstack-placement.2019-06-04.log.html#t2019-06-04T13:17:48 | |
| 03:56:13 | Sundar | But we can always add a change spec | |
| 03:56:40 | Sundar | #topic Cyborg patches | |
| 03:57:00 | Sundar | Q: Why not use asserts? They are complementary to exceptions, IMHO | |
| 03:57:37 | Sundar | Like I said in the code reviews, other projects use them. And they are useful to check internal program logic. E.g. a function must return a list | |
| 03:58:11 | Sundar | If we make exceptions for everything, there will be too many exceptions or, worse, developers will not put the check | |
| 03:58:52 | Sundar | yikun said they produce unclear messages | |
| 03:59:20 | Sundar | But we can say: assert x == y: "Bad state for x, differ s from y" | |
| 03:59:53 | ikuo_o | "They" means exceptions? | |
| 04:00:01 | yikun | Both the ways of the assert or exception is okay to me, :) But I prefer to use the exception. | |
| 04:00:26 | Sundar | ikuo_o: "they" mean asserts | |
| 04:01:14 | Sundar | yikun: OK. For e.g., would you put an exception to check that a function returns a list? | |
| 04:03:07 | Sundar | Anyways, this is food for thought :) | |
| 04:03:37 | yikun | OK, I think assert and exception has their use case. :) so, I said, both okay to me | |
| 04:03:44 | Sundar | If you are all ok, I'll leave the asserts in the code. IMHO, it is a good practice, recommended in software engineering books | |
| 04:03:54 | Sundar | yikun, ok. Thanks ;) | |
| 04:04:02 | ikuo_o | Sure. | |
| 04:04:11 | Sundar | #topic AoB | |
| 04:04:17 | Sundar | There is a Zoom meeting tomorrow | |
| 04:04:36 | Sundar | Mostly to wrap up the spec and patch reviews as much as possible | |
| 04:04:45 | ikuo_o | Ok, I will see you tomorrow. | |
| 04:05:01 | Sundar | ikuo_o and all, see you all tomorrow :) | |
| 04:05:08 | yikun | ok will join it, and thereare some patches need you guys eyes: | |
| 04:05:15 | yikun | ^ also the huawei ascend driver is also ready to merged | |
| 04:05:15 | yikun | https://review.opendev.org/#/c/660874/ | |
| 04:05:15 | yikun | ^ the generic driver spec have been approved, and the patch is ready for review. | |
| 04:05:15 | yikun | https://review.opendev.org/#/c/658987 | |
| 04:05:29 | Sundar | Great, will review | |
| 04:05:35 | yikun | any question, you can leave your comments, thanks. :) | |
| 04:05:37 | Sundar | Good day, everybody! | |
| 04:05:41 | Sundar | #endmeeting | |
| 04:05:42 | ikuo_o | Bye. | |
| 04:05:44 | openstack | Meeting ended Wed Jun 5 04:05:41 2019 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 04:05:45 | openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-06-05-03.03.html | |
| 04:05:46 | openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-06-05-03.03.txt | |
| 04:05:47 | openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2019/openstack_cyborg.2019-06-05-03.03.log.html | |
| 04:05:52 | Biwei | Bye | |
| 16:02:29 | edleafe | Sundar: I saw this comment: https://leafe.com/timeline/%23openstack-cyborg/2019-06-05T03:36:26 | |
| 16:03:17 | edleafe | Sundar: and wanted to let you know that there may be interest at IBM in also hosting a 3rd party CI for accelerators on Power hardware | |
| 16:03:59 | edleafe | Nothing definite yet, but they generally want to keep up with x86 testing | |
| 16:07:30 | Sundar | edleafe: GTK. Thanks. Please let us know what accelerators you plan to include | |
| 16:07:54 | edleafe | Sundar: will do | |
| 21:00:33 | Li_Liu | Sundar, I am still fighting the jetlag... will try my best for the zoom meeting night | |
| #openstack-cyborg - 2019-06-06 | |||
| 02:08:22 | Sundar | Specs to review by next Wednesday: https://review.opendev.org/658263, https://review.opendev.org/602978, https://review.opendev.org/605237 | |
| 03:08:03 | Sundar | Yikun, could you join me on IRC? | |
| 03:10:43 | yikun | yep, I'm online | |
| 03:11:05 | yikun | So, I think we should do right things in first place. | |
| 03:11:05 | yikun | I found this link when I review cinder code(https://review.opendev.org/#/c/663362), the team is try to clean all assert in cinder project. | |
| 03:11:05 | yikun | https://docs.openstack.org/bandit/latest/plugins/b101_assert_used.html | |
| 03:11:05 | yikun | FYI, here is why assert is not used in openstack. | |
| 03:11:19 | yikun | Sundar: ：), I'm here | |
| 03:12:03 | Sundar | Hi yikun, thanks for joining :) | |
| 03:13:08 | Sundar | We had a discussion today. For the Nova patches to merge, we need both a functional test in Nova that calls fake CYborg APIs *and* a tempest CI with a fake CYborg driver | |
| 03:13:47 | Sundar | XInran and Biwei are looking at the tempest CI in general | |
| 03:14:11 | Sundar | But the fake driver is lot like the generic driver -- I think | |
| 03:14:30 | Sundar | So, we were wondering if you could create a fake driver? | |
| 03:18:46 | Sundar | brb in 10 min | |
| 03:23:22 | yikun | Sure, but I'm not sure this fake driver looks like what? I can take this work, but I need some info from biwei or xinran | |
| 03:24:34 | Sundar | They had questions for me :) | |
| 03:24:44 | Sundar | OK, we need to figure out what this looks like | |
| 03:26:08 | Sundar | Do you know any of the tempest people? | |
| 03:28:27 | xinranwang | Actually, I don't understand why nova need fake driver test, they just wanna make sure that nova can call cyborg correctly and cyborg return the right data. It's all about API layer. | |
| 03:28:30 | yikun | No, I think maybe we can get some help from yumeng? | |
| 03:29:52 | xinranwang | Zhurong from ZTE have experience on tempest plugin. | |
| 03:32:16 | Sundar | xinranwang: Please see comments on line 409 in https://review.opendev.org/#/c/603955/13/specs/train/approved/nova-cyborg-interaction.rst,unified | |
| 03:32:38 | Sundar | yikun: Yumeng is on vacation till end June, IIRC | |
| 03:33:46 | Sundar | xinranwang: GTK. Can we contact that person or zhu.fanglei@zte.com.cn? | |
| 03:37:33 | xinranwang | According to Sean's comments, my understanding is, what nova required is still fake cyborg API, but for gate check, cyborg should have a fake driver to pass the tempest test. And there should be a test in tempest like test_boot_vm_with_accl. | |
| 03:40:06 | Sundar | Hmm, yes, so we need both, as per the comments. | |
| 03:40:14 | Sundar | xinranwang: ^ | |
| 03:41:11 | xinranwang | So no CI env need to be set up. Just code.(nova fixture, fake driver). Right? | |