| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-cyborg - 2020-07-23 | |||
| 03:07:32 | songwenping__ | #info songwenping__ | |
| 03:07:49 | Yumeng | #topic policy default refresh | |
| 03:08:27 | Yumeng | I updated this morning https://review.opendev.org/#/c/740542/ | |
| 03:08:59 | Yumeng | now this patch submmited the basic policies and device_profile policies, and basic framwork | |
| 03:10:18 | brinzhang_ | Yumeng: I was reviewed just now, but I am sorry not looked into. in general, LGTM, but that need to add the UT, I left comment inline. | |
| 03:11:05 | Yumeng | both the old and new policies will both work during development process | |
| 03:11:20 | brinzhang_ | Yumeng: after I compeleted thing in my hands, I will check again about the implemention details. | |
| 03:11:51 | brinzhang_ | sound good. | |
| 03:12:42 | brinzhang_ | we should keep old policies and new policies are all run good, thanks | |
| 03:12:46 | Yumeng | brinzhang_: yes we need test. but not as easy as a unit test update in current API tests. that should be another tests for only policies, we should add new policy test framwork, that would be a big work. | |
| 03:13:34 | brinzhang_ | Yumeng: I know, but in the patch if you dont add the UT, I dont think it's good to go. | |
| 03:13:39 | Yumeng | I will submit the policy test in another patch, I just don't want to make the first patch too big. | |
| 03:14:24 | brinzhang_ | you can splite the UT for add new policies, another patch is to add the new policies to the interface | |
| 03:15:04 | Yumeng | for this patch, I have tested the function by current unit test, the test_create case of device_profile, I have pasted the detailed debug info to the commit message: http://paste.openstack.org/show/796229/ | |
| 03:15:09 | xinranwang__ | Policy test framework is a big work. We can just add test in API test first | |
| 03:16:21 | xinranwang__ | For example, add test in device_profile API, test create dp by non admin user, and it fails | |
| 03:16:32 | brinzhang_ | xinranwang__: yeah, we can. But the better way is add as the same time. | |
| 03:16:39 | xinranwang__ | Do you think this make sense | |
| 03:17:17 | brinzhang_ | It make sense to me. But I also hope that it is not a TODO. | |
| 03:18:32 | xinranwang__ | We can add api test in this patch. Policy test framwork can be in a new patch. | |
| 03:18:42 | xinranwang__ | What do you think Yumeng | |
| 03:19:50 | Yumeng | xinranwang__: yes! agree, current API tests will go through the policy check in each test. I will start in the next patch very soon to submit policy test for device_profile | |
| 03:20:49 | brinzhang_ | thanks | |
| 03:21:44 | Yumeng | but for this patch, I think the detailed debug info has shown the policy check process, which can show that this patch basically works, so we can start review. | |
| 03:23:26 | Yumeng | this patch can be merged after the next policy test patch is passed and submitted, but can start review now. what do you think? | |
| 03:24:05 | brinzhang_ | Yumeng, do you test the older policies? I saw you deprecated the old policies, I am not sure it's ok for all apis | |
| 03:24:06 | brinzhang_ | https://review.opendev.org/#/c/740542/10/cyborg/policies/base.py@95 | |
| 03:25:20 | brinzhang_ | Yumeng: this patch I think we (xinranwang__) dont reviewed into, after checked again, we can decide whether let it to go, do you think so? | |
| 03:25:48 | brinzhang_ | is there a milestone? | |
| 03:27:10 | Yumeng | the old policies are "not actually" deprecated, they won't actually deprecated until we set [oslo_policy] enforce_policy=True in cyborg.conf | |
| 03:27:53 | Yumeng | the old policies work for now, they just marked as deprecated, not actually deprecated. | |
| 03:28:54 | Yumeng | emmm. I think we should add tests for new policies, for the old policies, we will deprecate them after all the new policies are implemented | |
| 03:29:36 | brinzhang_ | agree, so I think this change should have UT to cover the old policies and the new policies, and checked the enforce_policy config (of course you can do in another patch, but it will be not easy to review). | |
| 03:30:15 | brinzhang_ | no, after you completed the new policies, we also need to keep the old policies run until W or X release | |
| 03:30:46 | brinzhang_ | To ensure backward compatibility, we should not remove it so quickly. | |
| 03:31:05 | brinzhang_ | What do you think? Yumeng, xinranwang__, songwenping__ | |
| 03:31:20 | songwenping__ | Yumeng: you can fix the ut and pep8 first, ths. | |
| 03:32:05 | Yumeng | brinzhang_: yes, that's what I planed. I have NOTE them in the code. deprecate_since W | |
| 03:32:25 | xinranwang__ | Sorry, I am back | |
| 03:32:39 | xinranwang__ | brinzhang_: yes, agree | |
| 03:33:08 | Yumeng | xinranwang__: oPS, will see the pep8 error. don't know why. works fine in my env. | |
| 03:35:32 | brinzhang_ | ok. I have nothing to say of this patch now | |
| 03:35:45 | brinzhang_ | I have two patches need to go, it was submited for a looong time. https://review.opendev.org/#/c/734406/ and https://review.opendev.org/#/c/734403/ | |
| 03:35:59 | Yumeng | brinzhang_: anyway. this patch defines the overall framwork, I wanna you guys help to review it. | |
| 03:36:07 | brinzhang_ | Yumeng, xinranwang__, and all: Hope you can review all, thanks | |
| 03:36:42 | brinzhang_ | Yumeng: of course ^^, it's my pleasure | |
| 03:36:48 | songwenping__ | bingzhang_: will look into this patch. | |
| 03:36:52 | Yumeng | brinzhang_: about the milestone, that was set to milestone-2, but we will see if we need postpone | |
| 03:36:55 | xinranwang__ | https://www.irccloud.com/pastebin/z94K8fNV/ | |
| 03:37:32 | Yumeng | brinzhang_: yse, will review | |
| 03:37:40 | brinzhang_ | thanks | |
| 03:38:25 | songwenping__ | brinzhang_: sorry for error spell. | |
| 03:38:45 | brinzhang_ | xinranwang__: IRC cloud need to be jump the firewall, otherwise we cannot connect to it | |
| 03:39:07 | Yumeng | xinranwang__: aha. thanks xinran. I will fix it. | |
| 03:39:47 | xinranwang__ | Oh, I thought is works for PRC, never mind | |
| 03:40:30 | Yumeng | Inspur FPGA driver spec: https://review.opendev.org/#/c/730760/ | |
| 03:40:32 | Yumeng | songwenping__: just reviewed this morning. pls see my reply. | |
| 03:40:56 | Yumeng | we can merge this before milestone-2 | |
| 03:40:57 | songwenping__ | Yumeng: yeah, i see it. | |
| 03:41:12 | Yumeng | also Intel QAT driver ready for review:https://review.opendev.org/#/c/728014/ | |
| 03:41:20 | Yumeng | I will review again today. | |
| 03:41:32 | xinranwang__ | thanks a lot | |
| 03:41:33 | songwenping__ | thanks for review, i will check again. | |
| 03:42:05 | Yumeng | And shogo's is also milestone-2. Programming support ready for review: https://review.opendev.org/#/c/698190/ | |
| 03:42:17 | Yumeng | a big work. | |
| 03:42:26 | brinzhang_ | songwenping__, xinranwang__:will review later | |
| 03:42:39 | xinranwang__ | thanks brinzhang_ | |
| 03:43:24 | Yumeng | xinranwang__: can you also ask haibing to help review this, I'm not expert in programming. | |
| 03:43:30 | songwenping__ | thanks brinzhang__ | |
| 03:43:45 | xinranwang__ | Yes, sure | |
| 03:45:08 | Yumeng | thanks! | |
| 03:45:22 | Yumeng | I have nothing else to mention for today. | |
| 03:45:50 | Yumeng | Do you guys want to bring up anything else? | |
| 03:46:44 | xinranwang__ | Just noticed that cyborg-tempest-plugins has no doc | |
| 03:46:44 | brinzhang_ | nothings by me | |
| 03:47:08 | xinranwang__ | shall we add some docs for it? | |
| 03:48:08 | brinzhang_ | if necessagr we can | |
| 03:48:26 | Yumeng | not sure if we need. | |
| 03:48:35 | brinzhang_ | May there will be add in the future | |
| 03:48:48 | Yumeng | I can take a look, we can decide later.:) | |
| 03:48:58 | brinzhang_ | yeah | |
| 03:49:18 | Yumeng | ok. sounds good. | |
| 03:49:41 | Yumeng | let's wrap up this meeting and have a good lunch. | |
| 03:49:46 | xinranwang__ | Ok, it's not urgent. Just mention it here.lol | |
| 03:50:20 | xinranwang__ | Ok, thanks. Bye | |
| 03:50:22 | Yumeng | see you next week. Thank you for attending! bye | |
| 03:50:43 | songwenping__ | bye | |
| 03:51:00 | Yumeng | #endmeeting | |
| 03:51:02 | openstack | Meeting ended Thu Jul 23 03:51:00 2020 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 03:51:03 | openstack | Minutes: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-07-23-03.06.html | |
| 03:51:04 | openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-07-23-03.06.txt | |
| 03:51:05 | openstack | Log: http://eavesdrop.openstack.org/meetings/openstack_cyborg/2020/openstack_cyborg.2020-07-23-03.06.log.html | |
| 10:35:17 | openstackgerrit | Xinran WANG proposed openstack/cyborg-specs master: Add Intel QAT driver specification. https://review.opendev.org/728014 | |
| #openstack-cyborg - 2020-07-24 | |||
| 02:49:30 | openstackgerrit | YumengBao proposed openstack/cyborg master: P0: Refresh cyborg basic_policies and device_profile_policies https://review.opendev.org/740542 | |
| 06:52:43 | openstackgerrit | YumengBao proposed openstack/cyborg master: P0: Refresh cyborg basic_policies and device_profile_policies https://review.opendev.org/740542 | |
| 06:53:12 | openstackgerrit | YumengBao proposed openstack/cyborg master: P0: Refresh cyborg basic_policies and device_profile_policies https://review.opendev.org/740542 | |
| 09:18:40 | openstackgerrit | YumengBao proposed openstack/cyborg master: P0: Refresh cyborg basic_policies and device_profile_policies https://review.opendev.org/740542 | |
| 13:14:41 | openstackgerrit | Raildo Mascena proposed openstack/cyborg master: Add new default policy https://review.opendev.org/699102 | |
| #openstack-cyborg - 2020-07-27 | |||
| 01:41:01 | openstackgerrit | Merged openstack/python-cyborgclient master: Fix hacking min version to 3.0.1 https://review.opendev.org/727531 | |
| 11:26:47 | openstackgerrit | YumengBao proposed openstack/cyborg-specs master: Add Inspur FPGA driver specification. https://review.opendev.org/730760 | |