| *** yamamoto has joined #openstack-meeting-5 | 00:02 | |
| *** sshank has quit IRC | 00:39 | |
| *** yamamoto has quit IRC | 00:45 | |
| *** yamamoto has joined #openstack-meeting-5 | 01:24 | |
| *** spzala has joined #openstack-meeting-5 | 02:01 | |
| *** yamamoto has quit IRC | 02:22 | |
| *** yamamoto has joined #openstack-meeting-5 | 02:24 | |
| *** yamamoto has quit IRC | 02:40 | |
| *** ricolin_ has joined #openstack-meeting-5 | 02:52 | |
| *** yamahata has quit IRC | 02:54 | |
| *** iyamahat has quit IRC | 02:54 | |
| *** yamamoto has joined #openstack-meeting-5 | 02:54 | |
| *** ricolin_ has quit IRC | 02:59 | |
| *** yamamoto has quit IRC | 03:05 | |
| *** ricolin_ has joined #openstack-meeting-5 | 03:38 | |
| *** iyamahat has joined #openstack-meeting-5 | 04:10 | |
| *** yamahata has joined #openstack-meeting-5 | 04:26 | |
| *** lukepatrick has joined #openstack-meeting-5 | 04:51 | |
| *** lukepatrick has quit IRC | 05:24 | |
| *** kei-ichi has quit IRC | 05:31 | |
| *** kei-ichi has joined #openstack-meeting-5 | 05:34 | |
| *** yamamoto has joined #openstack-meeting-5 | 05:39 | |
| *** yamamoto has quit IRC | 05:43 | |
| *** yamamoto has joined #openstack-meeting-5 | 05:53 | |
| *** yamamoto has quit IRC | 05:57 | |
| *** yamamoto has joined #openstack-meeting-5 | 06:31 | |
| *** yamamoto has quit IRC | 06:43 | |
| *** yamamoto has joined #openstack-meeting-5 | 06:59 | |
| *** yamamoto has quit IRC | 07:06 | |
| *** markvoelker has quit IRC | 07:37 | |
| *** yamahata has quit IRC | 07:38 | |
| *** yamamoto has joined #openstack-meeting-5 | 07:49 | |
| *** yamamoto has quit IRC | 07:51 | |
| *** yamamoto has joined #openstack-meeting-5 | 07:57 | |
| *** TuanLA has joined #openstack-meeting-5 | 08:07 | |
| *** makowals has quit IRC | 08:16 | |
| *** yamamoto has quit IRC | 08:33 | |
| *** makowals has joined #openstack-meeting-5 | 08:36 | |
| *** markvoelker has joined #openstack-meeting-5 | 08:38 | |
| *** yamamoto has joined #openstack-meeting-5 | 08:40 | |
| *** yamamoto has quit IRC | 08:44 | |
| *** iyamahat has quit IRC | 08:52 | |
| *** yamamoto has joined #openstack-meeting-5 | 09:04 | |
| *** yamamoto has quit IRC | 09:19 | |
| *** TuanLA has quit IRC | 09:58 | |
| *** xek has quit IRC | 10:09 | |
| *** xek has joined #openstack-meeting-5 | 10:15 | |
| *** derekh has joined #openstack-meeting-5 | 10:45 | |
| *** zaneb has quit IRC | 11:09 | |
| *** zaneb has joined #openstack-meeting-5 | 11:15 | |
| *** yamamoto has joined #openstack-meeting-5 | 12:37 | |
| *** yamamoto has quit IRC | 13:04 | |
| *** yamamoto has joined #openstack-meeting-5 | 13:05 | |
| *** markvoelker has quit IRC | 13:23 | |
| *** markvoelker has joined #openstack-meeting-5 | 13:24 | |
| *** lujinluo has joined #openstack-meeting-5 | 13:54 | |
| *** hoangcx_ has joined #openstack-meeting-5 | 13:54 | |
| *** annp has joined #openstack-meeting-5 | 13:55 | |
| *** TuanVu has joined #openstack-meeting-5 | 13:55 | |
| *** ihrachys has joined #openstack-meeting-5 | 14:00 | |
| ihrachys | #startmeeting neutron_upgrades | 14:01 |
|---|---|---|
| openstack | Meeting started Thu Nov 30 14:01:31 2017 UTC and is due to finish in 60 minutes. The chair is ihrachys. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:01 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:01 |
| *** openstack changes topic to " (Meeting topic: neutron_upgrades)" | 14:01 | |
| openstack | The meeting name has been set to 'neutron_upgrades' | 14:01 |
| annp | ihrachys, Hi | 14:01 |
| lujinluo | o/ | 14:01 |
| TuanVu | Hi everybody | 14:01 |
| ihrachys | hi!! | 14:02 |
| *** hungpv has joined #openstack-meeting-5 | 14:02 | |
| * ihrachys waves at hungpv | 14:02 | |
| ihrachys | #topic Action items from prev meeting | 14:03 |
| *** openstack changes topic to "Action items from prev meeting (Meeting topic: neutron_upgrades)" | 14:03 | |
| ihrachys | there was a single one on me | 14:03 |
| ihrachys | "ihrachys to categorize remaining models imports" | 14:03 |
| ihrachys | I totally forgot about it :-x | 14:03 |
| ihrachys | I will take it to the next time | 14:03 |
| ihrachys | #action ihrachys to categorize remaining models imports | 14:03 |
| ihrachys | sorry for that | 14:03 |
| ihrachys | #topic OVO patches | 14:04 |
| *** openstack changes topic to "OVO patches (Meeting topic: neutron_upgrades)" | 14:04 | |
| lujinluo | never mind. ;) | 14:04 |
| ihrachys | https://review.openstack.org/#/q/status:open+project:openstack/neutron+branch:master+topic:bp/adopt-oslo-versioned-objects-for-db | 14:04 |
| ihrachys | before going through the list, there is another OVO patch here: https://review.openstack.org/#/c/519762/ | 14:04 |
| ihrachys | the one that reverts the revert of subnet patch | 14:04 |
| ihrachys | I think we are waiting for garyk to confirm he is ok with it now, giving him several more days to reply | 14:05 |
| ihrachys | ok, as for the list... | 14:05 |
| ihrachys | https://review.openstack.org/#/c/521797/ "Use Router OVO in external_net_db" | 14:05 |
| ihrachys | hungpv, there are logical issues with the code | 14:06 |
| ihrachys | or better, there WERE | 14:06 |
| ihrachys | I haven't checked latest version | 14:06 |
| hungpv | Yes, I see | 14:07 |
| hungpv | I'm trying to figure out on that | 14:07 |
| ihrachys | I think there are still issues. you use ~ there on the result of get_objects which is weird. | 14:07 |
| ihrachys | hungpv, what surprises me is that no unit tests fail because of the patch | 14:07 |
| ihrachys | we probably have bad coverage for the code | 14:07 |
| lujinluo | agree | 14:08 |
| ihrachys | there are some tempest api test failures here: http://logs.openstack.org/97/521797/5/check/neutron-tempest-plugin-api/7841216/logs/testr_results.html.gz but it would be cool if we could trigger the code with unit tests | 14:08 |
| *** kiennt26 has joined #openstack-meeting-5 | 14:09 | |
| ihrachys | on related note, there is a patch currently in review for a bug that affects the code that adds some unit tests that should trigger the code: https://review.openstack.org/#/c/512484/19/neutron/tests/unit/db/test_rbac_db_mixin.py | 14:11 |
| ihrachys | you could try to check whether new tests would trigger a failure with your (broken) patch. if they do, we could then rely on them to prove it works. | 14:11 |
| *** lukepatrick has joined #openstack-meeting-5 | 14:12 | |
| hungpv | Yes, I see. I'll check it to see if it happens | 14:12 |
| ihrachys | but if not, it would be nice to see some unit tests contributed as a separate preparatory patch that would cover the code | 14:12 |
| ihrachys | so that we are more sure the change doesn't break something | 14:13 |
| ihrachys | next is https://review.openstack.org/#/c/407868/ "[WIP] Integration of (Distributed) Port Binding OVO" | 14:13 |
| ihrachys | lujinluo, I believe it's still wip? | 14:14 |
| lujinluo | yes | 14:14 |
| ihrachys | anything we could help or advise? | 14:14 |
| lujinluo | i think it is something wrong in db side. it would be appreciated if you could check the ERROR in neutron-server logs | 14:15 |
| lujinluo | http://logs.openstack.org/68/407868/75/check/legacy-tempest-dsvm-neutron-dvr/3b5f642/logs/screen-q-svc.txt.gz?level=ERROR | 14:15 |
| lujinluo | No details.: InvalidRequestError: Can't attach instance <DistributedPortBinding at 0x7f6a0c3fe6d0>; another instance with key (<class 'neutron.plugins.ml2.models.DistributedPortBinding'>, (u'32b7f57a-6b27-476b-be60-cb6eae576fe9', u'ubuntu-xenial-rax-dfw-0001136185')) is already present in this session. | 14:16 |
| lujinluo | i root caused to update_port() in plugin/ml2/plugin.py | 14:16 |
| lujinluo | it seems to be the problem of mixing usage of distributed port binding in db and ovo | 14:16 |
| ihrachys | it seems to happen on delete_port only | 14:20 |
| ihrachys | so for what I understand, you can get this error message when you try to add a model state to the session twice | 14:23 |
| lujinluo | eh, right. | 14:24 |
| ihrachys | here, since it's delete, it would probably be "a model marked as deleted" | 14:24 |
| ihrachys | so there should be some other code somewhere on the code path that also modifies the model, maybe in a slightly different way | 14:24 |
| ihrachys | and also adds it to the session | 14:24 |
| lujinluo | hmm, so I am confused here. In the code path, we delete port and this triggers the cascade deletion of distributed_port_binding | 14:26 |
| ihrachys | here in https://review.openstack.org/#/c/407868/75/neutron/plugins/ml2/plugin.py@1446 | 14:26 |
| ihrachys | I wonder if this persist_state_to_session is somehow related | 14:27 |
| ihrachys | lujinluo, I was thinking, maybe delete_port happens after some other action on the binding model (in the same session / request) that would touch the binding model | 14:27 |
| ihrachys | so you first touch the model, e.g. update it somehow, and then delete_port happens a tad later in the code path. then sqlalchemy may be puzzled whether it should persist first or second state. | 14:28 |
| ihrachys | I suspect this persist_state_to_session logic is somehow related to the failure. | 14:29 |
| lujinluo | hmm, i will check that. So it is possible that we need to add some guards to sessions? | 14:30 |
| ihrachys | it's probably yet another kludge we needed to work around sqlalchemy | 14:30 |
| ihrachys | lujinluo, we may try to pull kevinbenton on that one since he introduced the persistence "snapshot" concept there | 14:31 |
| ihrachys | of course he is not working on openstack but probably could give an advice | 14:31 |
| lujinluo | ok. i will try to catch him somehow | 14:32 |
| ihrachys | maybe let's describe the issue in an email and send it to him | 14:32 |
| ihrachys | we could capture as many details in there | 14:32 |
| *** TuanVu has quit IRC | 14:32 | |
| lujinluo | sure. this would be a good idea | 14:32 |
| ihrachys | let's draft one in an etherpad (I will send a link to you later) | 14:32 |
| lujinluo | thanks | 14:33 |
| *** TuanVu has joined #openstack-meeting-5 | 14:33 | |
| ihrachys | ok moving on | 14:33 |
| ihrachys | next is https://review.openstack.org/#/c/396351/ "Integration of Floating IP OVO" | 14:33 |
| ihrachys | lujinluo, you left a comment there about qos_policy_binding | 14:34 |
| lujinluo | yes | 14:34 |
| ihrachys | lujinluo, I am not sure what's that about? | 14:34 |
| ihrachys | oh I think I follow now | 14:35 |
| ihrachys | you were just exploring whether another field is needed | 14:35 |
| ihrachys | so no issue in that version of the code | 14:35 |
| ihrachys | I see slaweq left some nits in the code | 14:35 |
| lujinluo | i editted a little | 14:35 |
| lujinluo | cause qos_policy_binding (a newly added table) has added relationship with floatingip table | 14:36 |
| lujinluo | i was checking if we need to make qos_policy_binding a synthetic field | 14:36 |
| lujinluo | in SQLA side, some codes like floatingip_db.qos_policy_binding appeared | 14:36 |
| lujinluo | i modified them to floatingip_obj.db_obj.qos_policy_binding | 14:37 |
| lujinluo | https://review.openstack.org/#/c/396351/46/neutron/db/l3_fip_qos.py | 14:37 |
| lujinluo | ^ here | 14:37 |
| lujinluo | this is the patch adds the new table https://review.openstack.org/#/c/424466/ | 14:38 |
| ihrachys | yeah. I think we may eventually want to make it similar to how we handle bindings for networks/ports | 14:38 |
| ihrachys | https://github.com/openstack/neutron/blob/master/neutron/objects/qos/policy.py#L208-L258 | 14:39 |
| ihrachys | actually, we already have attach/detach for FIPs there | 14:40 |
| ihrachys | then why don't we use them in l3_fip_qos? | 14:40 |
| ihrachys | oh we do | 14:41 |
| ihrachys | it just updates attributes in db model to reflect changes to bindings table | 14:42 |
| lujinluo | yes | 14:42 |
| ihrachys | I am not even sure why bother updating the attribute there. we don't seem to use it in caller? | 14:43 |
| lujinluo | i am not 100% sure. but before I made those changes, some tests in https://review.openstack.org/#/c/424466/31/neutron/tests/unit/extensions/test_qos_fip.py were failing | 14:45 |
| ihrachys | ok I will need some more time on that one | 14:45 |
| ihrachys | but there is a definite bug in the new dns code (the new version of it), I posted a comment | 14:46 |
| ihrachys | but there may be some more. I need to understand why we need to update the model. | 14:46 |
| ihrachys | ok those were all patches that are actionable | 14:47 |
| lujinluo | ok. let's continue on gerrit | 14:47 |
| *** lukepatrick has quit IRC | 14:47 | |
| ihrachys | #topic Work planning | 14:47 |
| *** openstack changes topic to "Work planning (Meeting topic: neutron_upgrades)" | 14:47 | |
| ihrachys | I (of course) haven't done remaining model usage classification | 14:48 |
| ihrachys | but I would nevertheless like to check if everyone involved has items to work on | 14:48 |
| ihrachys | hungpv and lujinluo work on patches already in gerrit | 14:49 |
| ihrachys | TuanVu, annp do you have things to work on? | 14:49 |
| TuanVu | Hi Ihar | 14:49 |
| TuanVu | regarding to this patch | 14:49 |
| ihrachys | "this" == ? | 14:50 |
| TuanVu | https://review.openstack.org/#/c/507772/ | 14:50 |
| TuanVu | We’re still working on this patch, there’s 1 blocking point at this moment. | 14:50 |
| TuanVu | Here’s the error: http://paste.openstack.org/show/627868/ | 14:50 |
| TuanVu | webob.exc.HTTPClientError: The server could not comply with the request since it is either malformed or otherwise incorrect. | 14:51 |
| lujinluo | this is not the root error | 14:51 |
| ihrachys | yeah it's just indication of error bubbled up to api layer | 14:52 |
| lujinluo | InvalidRequestError: Instance '<Network at 0x7f6c67b2e250>' is not persistent within this Session | 14:52 |
| lujinluo | ^ this is | 14:52 |
| ihrachys | + | 14:52 |
| lujinluo | i have not checked the patch but i suppose you used something like "network_obj.db_obj"? | 14:52 |
| TuanVu | thank you, Luo | 14:53 |
| TuanVu | yes, correct | 14:53 |
| annp | lujinluo, ihrachys, maybe this related to attach_rbac function | 14:53 |
| annp | when i check code, i saw self.obj_context.tenant_id is None. | 14:53 |
| annp | But i'm not sure. I need to dig more. | 14:54 |
| lujinluo | so from my personal experience, obj.db_obj is not persistent to any session. it cannot be used to be updated, etc. if you do, you will see that error. | 14:55 |
| ihrachys | yeah, it's by design | 14:56 |
| ihrachys | we detach it after fetching | 14:56 |
| lujinluo | exactly. | 14:56 |
| ihrachys | it fixes conflicts with other code that may want to work with models directly, and also that we don't fetch additional attributes in extension calls after data is persisted in db | 14:57 |
| ihrachys | it's good we can reproduce it with unit test | 14:58 |
| annp | lujinluo, ihrachys, Ok, I will dig more. :) | 14:58 |
| ihrachys | I will try to poke it in parallel | 14:58 |
| TuanVu | awesome! | 14:58 |
| TuanVu | thank you so much, Ihar | 14:58 |
| ihrachys | ok so it seems like everyone has a thing to work on, that's good | 14:58 |
| TuanVu | thank you, too, Luo | 14:58 |
| lujinluo | no problem ;) | 14:59 |
| ihrachys | we are out of time, so I guess we wrap up? | 14:59 |
| lujinluo | yes | 14:59 |
| ihrachys | ok, have a nice rest of the week! | 14:59 |
| TuanVu | yes | 14:59 |
| ihrachys | #endmeeting | 14:59 |
| *** openstack changes topic to "OpenStack Meetings || https://wiki.openstack.org/wiki/Meetings" | 14:59 | |
| openstack | Meeting ended Thu Nov 30 14:59:39 2017 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:59 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/neutron_upgrades/2017/neutron_upgrades.2017-11-30-14.01.html | 14:59 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/neutron_upgrades/2017/neutron_upgrades.2017-11-30-14.01.txt | 14:59 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/neutron_upgrades/2017/neutron_upgrades.2017-11-30-14.01.log.html | 14:59 |
| TuanVu | see you guys soon :) | 14:59 |
| lujinluo | thank you guys! | 14:59 |
| *** iyamahat has joined #openstack-meeting-5 | 15:00 | |
| *** yamahata has joined #openstack-meeting-5 | 15:00 | |
| annp | see you guys, thank you | 15:01 |
| *** TuanVu has quit IRC | 15:04 | |
| *** lujinluo has quit IRC | 15:04 | |
| *** hoangcx_ has quit IRC | 15:05 | |
| *** jrist has quit IRC | 15:08 | |
| *** jrist has joined #openstack-meeting-5 | 15:09 | |
| *** annp has quit IRC | 15:12 | |
| *** marst has joined #openstack-meeting-5 | 15:17 | |
| *** zhipeng has joined #openstack-meeting-5 | 15:19 | |
| *** zhipeng has quit IRC | 15:22 | |
| *** lukepatrick has joined #openstack-meeting-5 | 15:23 | |
| *** lukepatrick has quit IRC | 15:24 | |
| *** kiennt26 has quit IRC | 15:39 | |
| *** yamamoto has quit IRC | 15:49 | |
| *** dtantsur has joined #openstack-meeting-5 | 15:50 | |
| *** yamamoto has joined #openstack-meeting-5 | 15:50 | |
| *** lukepatrick has joined #openstack-meeting-5 | 16:08 | |
| *** phuongnh has joined #openstack-meeting-5 | 16:27 | |
| *** dtantsur has left #openstack-meeting-5 | 16:34 | |
| *** iyamahat has quit IRC | 17:07 | |
| *** yamahata has quit IRC | 17:08 | |
| *** hungpv has quit IRC | 17:12 | |
| *** jlvacation is now known as jlvillal | 17:25 | |
| *** iyamahat has joined #openstack-meeting-5 | 17:33 | |
| *** phuongnh has quit IRC | 17:36 | |
| *** derekh has quit IRC | 17:49 | |
| *** yamahata has joined #openstack-meeting-5 | 17:51 | |
| *** ricolin_ has quit IRC | 18:49 | |
| *** derekh has joined #openstack-meeting-5 | 19:04 | |
| *** felipemonteiro has joined #openstack-meeting-5 | 19:17 | |
| *** derekh has quit IRC | 20:52 | |
| *** sshank has joined #openstack-meeting-5 | 21:29 | |
| *** marst has quit IRC | 22:19 | |
| *** felipemonteiro has quit IRC | 22:44 | |
| *** felipemonteiro has joined #openstack-meeting-5 | 23:02 | |
| *** kei-ichi has quit IRC | 23:02 | |
| *** kei-ichi has joined #openstack-meeting-5 | 23:04 | |
| *** iyamahat_ has joined #openstack-meeting-5 | 23:22 | |
| *** iyamahat has quit IRC | 23:22 | |
| *** felipemonteiro has quit IRC | 23:38 | |
| *** sshank has quit IRC | 23:52 | |
| *** felipemonteiro has joined #openstack-meeting-5 | 23:56 | |
| *** felipemonteiro_ has joined #openstack-meeting-5 | 23:57 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!