14:00:06 #startmeeting neutron_upgrades 14:00:07 Meeting started Thu Oct 18 14:00:06 2018 UTC and is due to finish in 60 minutes. The chair is lujinluo. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:11 The meeting name has been set to 'neutron_upgrades' 14:00:16 o/ 14:01:53 o/ 14:02:12 o/ 14:02:24 hi 14:02:36 hi njohnston ralonsoh slaweq ! 14:02:43 then let's get started 14:03:04 last week we have one action item: njohnston to enabl new_face in Agent OVO 14:03:13 #link https://review.openstack.org/#/c/609749/ 14:03:19 Yes, I tried that, and it had some issues 14:03:22 *new_facade 14:03:31 not as straightforward as I would have hoped 14:03:42 given that the agent part of the spreadsheet was completed 14:04:28 yeah, I downloaded the patch locally and tried to test it. I think we still miss something 14:05:05 I worry that it is some of the "common" code that we can't convert until all of the code has been changed to OVO 14:05:12 i.e. the stuff marked as "blocked 14:05:21 " on the OVO tracking spreadsheet 14:07:03 i see. 14:07:26 the failures in unit tests: http://logs.openstack.org/49/609749/1/check/openstack-tox-py36/a962088/testr_results.html.gz 14:07:32 and in functional: http://logs.openstack.org/49/609749/1/check/neutron-functional/d69aeae/job-output.txt.gz#_2018-10-11_19_14_37_954448 14:07:51 that "common" thing is just a hypothesis, I have no data to back it up yet 14:08:16 the failures are most related to router and l3 agent though 14:08:39 it could be router OVO that is blocking 14:08:49 so my thought was to pick a specific test and step through with pdb and see if I can track the problem 14:09:54 i think that is what we should try at least 14:10:48 i will continue testing that patch on my side as well. i will ping you if i find anything interesting njohnston 14:11:03 #action njohnston pick tests from https://review.openstack.org/609749 and step through with pdb to detect failure 14:11:24 will do, thanks very much for your help lujinluo! 14:11:38 no problem! 14:11:47 then let's move to OVO patches 14:11:50 #topic OVO 14:12:02 #link https://review.openstack.org/#/q/status:open+topic:bp/adopt-oslo-versioned-objects-for-db 14:12:29 #link https://review.openstack.org/#/c/608198/ 14:12:39 Use NetworkRBAC OVO in neutron/db/db_base_plugin_v2.py 14:13:40 i just have one small question in https://review.openstack.org/#/c/608198/7/neutron/db/db_base_plugin_v2.py@437 14:13:51 other changes look good to me 14:14:22 lujinluo: I'll answer this question today 14:14:55 thanks! ralonsoh 14:15:31 next #link https://review.openstack.org/#/c/608638/ 14:15:40 Use NetworkRBAC OVO in tests/unit/db/test_db_base_plugin_v2.py 14:16:01 it has got 3 +1 14:16:14 slaweq: could you please take a look at it when you are available? 14:16:24 it is changes to a test file 14:16:29 lujinluo: sure 14:16:33 I will review it today 14:16:43 thank you slaweq ! 14:17:23 #link https://review.openstack.org/#/c/611461/ 14:17:36 WIP: Fix port status for router ha_migration 14:17:41 it is still a WIP 14:20:17 it is not a OVO transition but related to it, so i guess we can wait until CI is cool with it before we review 14:20:28 +1 14:20:59 #link https://review.openstack.org/#/c/610280/ 14:21:10 Use RBACBaseObject OVO in neutron/db/rbac_db_mixin.py 14:21:39 I still need to fix the UTs 14:21:47 I need more time 14:21:59 got it. yeah, take your time :) 14:22:46 #link https://review.openstack.org/#/c/608628/ 14:22:56 Use NetworkRBAC OVO in neutron/db/external_net_db.py 14:23:40 slawek suggeted that we wait for #link https://review.openstack.org/#/c/609692/ to be merged first 14:23:50 but the changes look good to most of us :) 14:24:56 #link https://review.openstack.org/#/c/544206/ 14:25:05 Integration of (Distributed) Port Binding OVO 14:26:10 last week, i said that i was trying to justify the code change mentioned by hongbin, but it turns that the change was not necessary, so i pushed a new patch set 14:26:48 i still have one last comment related to load relationship that i have not tested yet. 14:27:33 i will remove my -1 if i finished the testing, then reviews are much welcomed 14:28:32 #link https://review.openstack.org/#/c/602497/ 14:28:43 use ovo for L3HARouterAgentPortBinding in l3_hamode. 14:29:23 Zuul is still not happy about it 14:31:42 o/ 14:31:47 just rechecked and hope it will make it this time 14:31:50 hi hongbin ! 14:31:58 * hongbin is being pulled from fwaas meeting 14:32:05 lujinluo: hi 14:32:23 aha, busy with fwaas these days? hongbin 14:32:40 next ps is from hongbin #link https://review.openstack.org/#/c/610182/ 14:32:46 lujinluo: a bit, but my topic is finished there now 14:32:54 Support fetching specific db column in OVO 14:33:26 for this one, i am figuring out how to pass the unit tests 14:33:29 this patch is -W by hongbin, so let's give him more time :) 14:33:45 yes, more time is always needed :) 14:34:32 #link https://review.openstack.org/#/c/507772/ Use Network OVO in db_base_plugin 14:35:11 Tuan did not update this patch since last week, and boden left a comment about the compatibility layer which we once discussed before 14:35:34 I will try to catch Tuan this week to see if he has any updates on the comments 14:35:54 #action lujin to keep an eye on Network OVO in db_base_plugin 14:36:32 #link https://review.openstack.org/#/c/609847/ Support filter with "not in" criterion 14:37:00 i think this patch would benefit us a lot, thanks hongbin! 14:37:04 +1 14:37:15 np 14:37:17 agree 14:37:28 will address the comments later, good comments though 14:37:46 excellent! 14:38:28 and i expect there might be a bit more criterion to add (other than "not in") 14:39:53 yes, this builds a good base for those future operators to be developed on later 14:40:13 +1 14:41:49 The next two patches seem to be WIP #link https://review.openstack.org/#/c/610184/ and #link https://review.openstack.org/#/c/609848/ 14:42:12 Convert Subnet to OVO in ipam_pluggable_backend.py 14:42:21 Convert Port to OVO in db/ipam_pluggable_backend.py 14:43:10 the first one is for testing the feature for fetching specific column 14:43:24 the second one is for testing the feature of "not in" filter 14:44:22 both are small patches, i will remove the WIP marker once the depending features are landed 14:44:34 understood. 14:45:20 the next one about retry decorator, which i want to discuss about 14:45:23 #Link https://review.openstack.org/#/c/602362/ 14:45:43 and this one in neutron-lib #link https://review.openstack.org/#/c/605365/ 14:46:14 So michel would like us not to refactor old engine facade 14:46:45 and suggested to "unpack the context and call a helper function" 14:47:30 yes, it seems like an interesting idea, although I have not given it enough thought to know how to go about it 14:47:40 my concern is that, if we follow this way, do we need to do it in every methods that we use retry decorator 14:47:56 Lajos said he/she would push a patch set to do it 14:48:13 but only the Tag ovo 14:48:56 ok, i think it is an interesting idea too. just want to bring it on the table to make us aware of its status :) 14:50:24 we can continue the discussion in Gerrit after Lajos pushes another ps in #link https://review.openstack.org/#/c/602362/ 14:50:54 the rest of the ovo patches are not updated since last week 14:51:01 #topic new engine facade 14:51:17 #link https://review.openstack.org/#/q/status:open+project:openstack/neutron+branch:master+topic:bp/enginefacade-switch 14:51:43 interesting enough, i was trying to switch to new engine facade following #Link https://review.openstack.org/#/c/545501/ 14:52:06 so i pushed #link https://review.openstack.org/611462 14:52:42 but it turns out it is some cleaning work of agent ovo 14:53:03 "Remove context guard of Agent OVO" 14:54:02 I'll try rebasing my agent ovo engine facade adoption change on top of this 14:54:33 yeah, that worth a try 14:54:45 done 14:55:06 great! 14:55:19 ok, that's all for the patches. 14:55:24 #topic open discussion 14:55:38 Does anyone have anything else wants to share? 14:56:40 if not, let's call it a day 14:56:44 thanks all 14:56:51 have a great day/night everyone 14:56:55 bye! 14:56:57 see you all next week 14:57:10 #endmeeting