03:04:38 #startmeeting openstack-cyborg 03:04:39 Meeting started Thu May 14 03:04:38 2020 UTC and is due to finish in 60 minutes. The chair is Sundar. Information about MeetBot at http://wiki.debian.org/MeetBot. 03:04:40 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 03:04:42 The meeting name has been set to 'openstack_cyborg' 03:05:24 We haven't seen any RC patches for Cyborg yet, right? 03:05:24 hi 03:05:53 By now, I would have expected a release patch 03:06:21 yes, we need a release patch 03:06:52 and recently we have a lot patches need to be deal, https://review.opendev.org/#/q/project:openstack/cyborg+status:open 03:09:40 Hmm, Ussuri is already released: http://lists.openstack.org/pipermail/openstack-discuss/2020-May/014847.html 03:10:37 Sundar: I think we ignored this, I will check for cyborg and cyborgclient 03:11:25 The release patch was this: https://review.opendev.org/#/c/721400/ 03:11:32 Yumeng signed off. 03:12:32 This was the last commit: https://opendev.org/openstack/cyborg/commit/973716aa23d033df3a23aa2b34ce533f3f320c8f 03:13:24 Only 5 or so patches after that, so we didn't miss much 03:14:09 Ok, so we are already in Victoria cycle 03:14:41 I had submitted the Cyborg highlights for Ussuri back in early April. So, we are good there. 03:15:14 Sundar: OK, I will push a commit for https://opendev.org/openstack/cyborg/commit/973716aa23d033df3a23aa2b34ce533f3f320c8f 03:15:16 Ussuri highlights puts Cyborg up there: https://releases.openstack.org/ussuri/highlights.html 03:16:20 There are three patch belongs to Ussuri, so I would like to merge that to U https://review.opendev.org/#/c/723517/ 03:17:50 brinzhang_: It is already committed, right? 03:18:38 Sundar: yes, these patch already committed by Sean McGinnis 03:18:40 I didn't fllow why the revert patches were submitted. Does anybody know the context? 03:18:46 *follow 03:19:05 I pushed that in recently IRC meeting 03:19:42 You said you can talk to Sean McGinnis by irc, maybe you forget Sundar 03:20:16 I thought Yumeng said she'd follow up in that discussion? 03:20:46 I remember that another, said to gibi in nova channel 03:20:58 Anyway, if it is still unclear, I should ask Sean. 03:21:37 Sundar, ok, after you talked him, then you can +A or -2 for this change, thanks 03:22:10 I put a reminder for myself. Will follow up. 03:22:28 Sundar: thanks 03:22:40 NP. Anything specific for today? 03:23:43 I think there more open status patches need to consider, whether in Cyborg, cyborgclient or cyborg-manila-plugin, such as bump hacking version patch 03:24:43 Could you give a pointer to this cyborg-manila-plugin? 03:25:38 sorry.. cyborg-tempest-plugin ^^ 03:25:55 spelling error :( 03:26:00 Oh ok, sure, np :) 03:26:11 I thought I missed something 03:27:15 Yup, there are quite a few patches. If you LMK the top patches to look at, I'll try to catch up after this meeting. 03:27:45 I think you mentioned the hacking patch: https://review.opendev.org/727483 03:28:28 right 03:28:48 Done. What else? 03:29:09 Looking at https://review.opendev.org/#/c/727603/ 03:29:23 Sundar: So far, do we only support launch one instance with accelerators? 03:30:02 Sundar: I think https://review.opendev.org/#/c/727482/ double with https://review.opendev.org/#/c/727483 03:30:10 No, we cna launch up to 10 instances together. Only the fake driver supports that so far. 03:31:07 If we launch more than one instances with accelerators, there may be problems in filter scheduler. pls see my patch:https://review.opendev.org/#/c/722651/ 03:32:55 Sundar: yes https://review.opendev.org/#/c/727603/ reported by zuul check, thanks for +2 03:33:10 songwenping_: I have launched 10 instances with fake driver in one host, using my Nova patches. This may be a concurrency issue. 03:36:48 Merged openstack/cyborg master: Fix hacking min version to 3.0.1 https://review.opendev.org/727483 03:39:13 Sundar: In my environment, it does not look like a concurrency issue. Are your rps in db is same as mine as description at:https://etherpad.opendev.org/p/filter_scheduler_issue_with_accelerators 03:39:50 Sundar: https://opendev.org/openstack/cyborg/commit/973716aa23d033df3a23aa2b34ce533f3f320c8f this should 4.0.0.0rc2, right? 03:41:01 songwenping_: With the fake driver, it is one RP with 16 accelerator resources. In your case, you have two RPs with one resource each. May be that accounts for the difference. 03:42:01 brinzhang_: Does it matter? It got merged anyway, right? 03:42:47 Sundar, yes, I will keeprc1 03:43:50 songwenping_: Thanks for https://review.opendev.org/723865 . I gave a comment. 03:43:52 Sundar: ths. i will check the fake driver. 03:45:15 Sundar: the release change, please check https://review.opendev.org/#/c/727916/ 03:47:20 Sundar: Thanks for your comment, i will add the 423 case. 03:48:55 Sundar: does need I add this to stable/ussuri? now in master branch 03:50:05 brinzhang_: It is already in stable/ussuri, I think: https://opendev.org/openstack/cyborg/src/branch/stable/ussuri/.zuul.yaml 03:50:54 Sundar: I know, I mean, the release patach https://review.opendev.org/#/c/727916/ 03:51:44 Now, I submit in master branch, I am not sure whether need move to stable/ussuri in openstack/releases project 03:51:52 I don't understand why we need this patch. It was already merged into Ussuri. 03:52:16 That is, tox-bandit is a voting job in Ussuri 03:53:05 Sundar: ok, I think I was wrong 03:53:28 I will abandon this patch, Yumeng updated in https://review.opendev.org/#/c/721400/ 03:53:58 Yes 03:54:05 I have nothing to talk 03:54:17 Anything else, folks? 03:54:50 Nothing to talk. 03:56:07 Thanks again for a successful Ussuri release. It has been my pleasure to server as PTL. From the next meeting, I hope Yumeng will drive the meeting. 03:56:22 I think it is lunch time for most of you. Enjoy your lunch, and have a good day! 03:56:38 #endmeeting