| *** ociuhandu has quit IRC | 00:01 | |
| *** rcernin_ has joined #openstack-cinder | 00:04 | |
| *** rcernin has quit IRC | 00:07 | |
| *** ociuhandu has joined #openstack-cinder | 00:10 | |
| *** ociuhandu has quit IRC | 00:14 | |
| *** ociuhandu has joined #openstack-cinder | 00:17 | |
| *** rcernin_ has quit IRC | 00:19 | |
| *** ociuhandu has quit IRC | 00:21 | |
| *** ociuhandu has joined #openstack-cinder | 00:23 | |
| *** rcernin has joined #openstack-cinder | 00:27 | |
| *** ociuhandu has quit IRC | 00:28 | |
| *** awalende has joined #openstack-cinder | 00:28 | |
| *** dave-mccowan has joined #openstack-cinder | 00:29 | |
| *** ociuhandu has joined #openstack-cinder | 00:31 | |
| *** dave-mccowan has quit IRC | 00:33 | |
| *** awalende has quit IRC | 00:33 | |
| *** ociuhandu has quit IRC | 00:36 | |
| *** dave-mccowan has joined #openstack-cinder | 00:36 | |
| *** ociuhandu has joined #openstack-cinder | 00:41 | |
| *** ociuhandu has quit IRC | 00:46 | |
| *** zhanglong has joined #openstack-cinder | 00:47 | |
| *** brinzhang has joined #openstack-cinder | 00:50 | |
| *** ociuhandu has joined #openstack-cinder | 00:51 | |
| *** dave-mccowan has quit IRC | 00:52 | |
| *** ociuhandu has quit IRC | 00:57 | |
| *** ociuhandu has joined #openstack-cinder | 01:02 | |
| *** ociuhandu has quit IRC | 01:06 | |
| *** ircuser-1 has joined #openstack-cinder | 01:07 | |
| *** ociuhandu has joined #openstack-cinder | 01:12 | |
| *** enriquetaso has quit IRC | 01:12 | |
| *** ociuhandu has quit IRC | 01:16 | |
| *** Liang__ has joined #openstack-cinder | 01:18 | |
| *** ociuhandu has joined #openstack-cinder | 01:21 | |
| *** zhanglong has quit IRC | 01:25 | |
| *** ociuhandu has quit IRC | 01:26 | |
| *** zhanglong has joined #openstack-cinder | 01:27 | |
| *** ociuhandu has joined #openstack-cinder | 01:29 | |
| *** ociuhandu has quit IRC | 01:37 | |
| *** ociuhandu has joined #openstack-cinder | 01:42 | |
| *** ociuhandu has quit IRC | 01:47 | |
| *** ociuhandu has joined #openstack-cinder | 01:49 | |
| *** whfnst has joined #openstack-cinder | 01:52 | |
| *** ociuhandu has quit IRC | 01:57 | |
| *** ociuhandu has joined #openstack-cinder | 02:08 | |
| *** Xuchu has joined #openstack-cinder | 02:11 | |
| *** ociuhandu has quit IRC | 02:13 | |
| *** ociuhandu has joined #openstack-cinder | 02:14 | |
| *** brinzhang_ has joined #openstack-cinder | 02:22 | |
| *** ociuhandu has quit IRC | 02:22 | |
| *** brinzhang has quit IRC | 02:25 | |
| *** zhanglong has quit IRC | 02:27 | |
| *** zhanglong has joined #openstack-cinder | 02:32 | |
| *** lvbin01 has quit IRC | 02:36 | |
| *** ociuhandu has joined #openstack-cinder | 02:38 | |
| *** vesper11 has quit IRC | 02:45 | |
| *** tridde has joined #openstack-cinder | 02:45 | |
| *** trident has quit IRC | 02:46 | |
| *** vesper11 has joined #openstack-cinder | 02:46 | |
| *** ociuhandu has quit IRC | 02:47 | |
| *** lvbin01 has joined #openstack-cinder | 02:50 | |
| *** brinzhang has joined #openstack-cinder | 02:56 | |
| openstackgerrit | Peng Wang proposed openstack/cinder master: DS8k Cinder Driver support Python3 https://review.opendev.org/697960 | 02:57 |
|---|---|---|
| *** brinzhang_ has quit IRC | 02:59 | |
| *** brinzhang_ has joined #openstack-cinder | 03:00 | |
| *** brinzhang has quit IRC | 03:03 | |
| *** brinzhang has joined #openstack-cinder | 03:12 | |
| *** brinzhang_ has quit IRC | 03:15 | |
| *** Xuchu_ has joined #openstack-cinder | 03:20 | |
| *** Xuchu has quit IRC | 03:23 | |
| *** Xuchu_ is now known as Xuchu | 03:23 | |
| *** psachin has joined #openstack-cinder | 03:31 | |
| *** awalende has joined #openstack-cinder | 03:39 | |
| *** brinzhang_ has joined #openstack-cinder | 03:42 | |
| *** awalende has quit IRC | 03:44 | |
| *** brinzhang has quit IRC | 03:46 | |
| *** ociuhandu has joined #openstack-cinder | 03:47 | |
| *** ociuhandu has quit IRC | 03:51 | |
| *** bhagyashris has joined #openstack-cinder | 04:23 | |
| *** rishabhhpe has joined #openstack-cinder | 04:30 | |
| *** ociuhandu has joined #openstack-cinder | 04:41 | |
| *** udesale has joined #openstack-cinder | 04:41 | |
| *** brinzhang_ has quit IRC | 04:44 | |
| *** brinzhang_ has joined #openstack-cinder | 04:45 | |
| *** ociuhandu has quit IRC | 04:46 | |
| *** brinzhang_ has quit IRC | 04:46 | |
| *** brinzhang_ has joined #openstack-cinder | 04:47 | |
| *** brinzhang has joined #openstack-cinder | 04:52 | |
| *** brinzhang has quit IRC | 04:53 | |
| *** brinzhang has joined #openstack-cinder | 04:54 | |
| *** brinzhang_ has quit IRC | 04:55 | |
| *** brinzhang_ has joined #openstack-cinder | 05:12 | |
| *** brinzhang_ has quit IRC | 05:12 | |
| *** brinzhang_ has joined #openstack-cinder | 05:13 | |
| *** brinzhang has quit IRC | 05:15 | |
| *** bhagyashris has quit IRC | 05:20 | |
| *** ociuhandu has joined #openstack-cinder | 05:35 | |
| *** ociuhandu has quit IRC | 05:40 | |
| *** bhagyashris has joined #openstack-cinder | 06:00 | |
| *** sapd1 has joined #openstack-cinder | 06:04 | |
| *** Luzi has joined #openstack-cinder | 06:04 | |
| *** e0ne has joined #openstack-cinder | 06:09 | |
| *** dpawlik has joined #openstack-cinder | 06:12 | |
| *** dpawlik has quit IRC | 06:18 | |
| *** e0ne has quit IRC | 06:19 | |
| *** dpawlik has joined #openstack-cinder | 06:20 | |
| *** whoami-rajat has joined #openstack-cinder | 06:21 | |
| *** dpawlik has quit IRC | 06:27 | |
| *** ociuhandu has joined #openstack-cinder | 06:29 | |
| *** ociuhandu has quit IRC | 06:34 | |
| *** psachin has quit IRC | 06:42 | |
| *** dpawlik has joined #openstack-cinder | 06:43 | |
| *** psachin has joined #openstack-cinder | 06:49 | |
| *** brinzhang has joined #openstack-cinder | 06:50 | |
| openstackgerrit | Xuan Yandong proposed openstack/python-cinderclient master: Drop support for python 2 https://review.opendev.org/691467 | 06:53 |
| *** brinzhang_ has quit IRC | 06:54 | |
| *** lpetrut has joined #openstack-cinder | 07:00 | |
| *** ociuhandu has joined #openstack-cinder | 07:02 | |
| *** ociuhandu has quit IRC | 07:06 | |
| *** sapd1 has quit IRC | 07:14 | |
| *** brinzhang_ has joined #openstack-cinder | 07:24 | |
| *** brinzhang_ has quit IRC | 07:26 | |
| *** ociuhandu has joined #openstack-cinder | 07:26 | |
| *** brinzhang_ has joined #openstack-cinder | 07:26 | |
| *** brinzhang has quit IRC | 07:27 | |
| *** brinzhang_ has quit IRC | 07:28 | |
| *** brinzhang_ has joined #openstack-cinder | 07:28 | |
| *** zhanglong has quit IRC | 07:29 | |
| *** brinzhang_ has quit IRC | 07:30 | |
| *** brinzhang_ has joined #openstack-cinder | 07:30 | |
| *** zhanglong has joined #openstack-cinder | 07:30 | |
| *** brinzhang_ has quit IRC | 07:31 | |
| *** ociuhandu has quit IRC | 07:32 | |
| *** pcaruana has joined #openstack-cinder | 07:39 | |
| *** ociuhandu has joined #openstack-cinder | 07:46 | |
| *** sahid has joined #openstack-cinder | 07:51 | |
| *** rcernin has quit IRC | 07:54 | |
| *** rishabhhpe has quit IRC | 07:55 | |
| *** bhagyashris has quit IRC | 08:05 | |
| *** tkajinam has quit IRC | 08:06 | |
| *** tosky has joined #openstack-cinder | 08:11 | |
| *** rishabhhpe has joined #openstack-cinder | 08:15 | |
| *** awalende has joined #openstack-cinder | 08:18 | |
| *** tesseract has joined #openstack-cinder | 08:27 | |
| *** RyanLiang has joined #openstack-cinder | 08:42 | |
| *** ociuhandu has quit IRC | 08:48 | |
| *** raghavendrat has quit IRC | 09:00 | |
| *** Luzi has quit IRC | 09:11 | |
| *** sam_wan has quit IRC | 09:14 | |
| *** ociuhandu has joined #openstack-cinder | 09:17 | |
| *** martinkennelly has joined #openstack-cinder | 09:19 | |
| *** ociuhandu has quit IRC | 09:22 | |
| *** spsurya has joined #openstack-cinder | 09:25 | |
| *** Luzi has joined #openstack-cinder | 09:26 | |
| *** ociuhandu has joined #openstack-cinder | 09:42 | |
| openstackgerrit | Lenny Verkhovsky proposed openstack/cinder master: Fix open tempfile.NamedTemporaryFile as string for Python3 https://review.opendev.org/698202 | 09:44 |
| *** davidsha has joined #openstack-cinder | 09:46 | |
| *** martinkennelly has quit IRC | 09:46 | |
| *** e0ne has joined #openstack-cinder | 09:54 | |
| *** bhagyashris has joined #openstack-cinder | 10:22 | |
| *** Liang__ has quit IRC | 10:26 | |
| *** martinkennelly has joined #openstack-cinder | 10:28 | |
| *** bhagyashris has quit IRC | 10:50 | |
| *** zhanglong has quit IRC | 10:53 | |
| *** bhagyashris has joined #openstack-cinder | 10:58 | |
| openstackgerrit | xinding proposed openstack/cinder master: Snapshot does not expand after expanding volume https://review.opendev.org/698217 | 11:04 |
| *** kambiz has quit IRC | 11:09 | |
| *** ociuhandu has quit IRC | 11:13 | |
| *** ociuhandu has joined #openstack-cinder | 11:13 | |
| *** kambiz has joined #openstack-cinder | 11:14 | |
| *** ociuhandu has quit IRC | 11:28 | |
| *** rishabhhpe has quit IRC | 11:34 | |
| *** tommylikehu has joined #openstack-cinder | 11:37 | |
| *** bhagyashris has quit IRC | 11:42 | |
| *** bhagyashris has joined #openstack-cinder | 11:42 | |
| *** udesale has quit IRC | 12:01 | |
| *** lvbin01 has quit IRC | 12:02 | |
| *** lvbin01 has joined #openstack-cinder | 12:17 | |
| *** ociuhandu has joined #openstack-cinder | 12:23 | |
| *** ociuhandu has quit IRC | 12:27 | |
| *** spsurya has quit IRC | 12:35 | |
| *** Xuchu has quit IRC | 12:51 | |
| *** bhagyashris has quit IRC | 12:58 | |
| *** dpawlik has quit IRC | 13:01 | |
| *** enriquetaso has joined #openstack-cinder | 13:04 | |
| *** jgriffith has joined #openstack-cinder | 13:29 | |
| *** whoami-rajat has quit IRC | 13:31 | |
| *** yaawang has quit IRC | 13:39 | |
| *** yaawang has joined #openstack-cinder | 13:39 | |
| *** Xuchu has joined #openstack-cinder | 13:56 | |
| *** Liang__ has joined #openstack-cinder | 14:00 | |
| *** tkajinam has joined #openstack-cinder | 14:01 | |
| openstackgerrit | Gorka Eguileor proposed openstack/cinderlib master: Add missing release notes https://review.opendev.org/698251 | 14:07 |
| *** eharney has quit IRC | 14:07 | |
| *** geguileo has joined #openstack-cinder | 14:08 | |
| geguileo | rosmaita: Promised missing cinderlib release notes: https://review.opendev.org/698251 | 14:08 |
| rosmaita | geguileo: ty! | 14:09 |
| geguileo | np | 14:09 |
| rosmaita | did the config patch get merged yet? | 14:09 |
| geguileo | rosmaita: not yet | 14:10 |
| rosmaita | hmmmm ... we will have to twist some arms | 14:11 |
| rosmaita | cinder cores: need someone other than me or geguileo to review https://review.opendev.org/#/c/696375/ -- needs to land before we can cut the train cinderlib release | 14:11 |
| rosmaita | jungleboyj: e0ne: whoami-rajat: smcginnis: eharney: ^^ | 14:12 |
| rosmaita | hemna_: ^^ (sorry, did not mean to leave you out) | 14:12 |
| *** awalende has quit IRC | 14:13 | |
| *** awalende has joined #openstack-cinder | 14:14 | |
| *** awalende has quit IRC | 14:14 | |
| *** awalende has joined #openstack-cinder | 14:14 | |
| *** Xuchu has quit IRC | 14:23 | |
| e0ne | rosmaita: looking on it | 14:24 |
| rosmaita | e0ne: ty! | 14:24 |
| e0ne | rosmaita: +2'ed. | 14:29 |
| *** Luzi has quit IRC | 14:29 | |
| rosmaita | thanks! appreciate it | 14:29 |
| e0ne | I need to do more reviews to cinderlib. I'm still feeling uncomfortable to approve patches | 14:29 |
| e0ne | rosmaita: feel free to approve this patch now :) | 14:33 |
| rosmaita | ok | 14:34 |
| *** ociuhandu has joined #openstack-cinder | 14:35 | |
| *** mmethot has joined #openstack-cinder | 14:35 | |
| *** mmethot is now known as mmethot|conferen | 14:35 | |
| *** pcaruana has quit IRC | 14:36 | |
| hemna_ | mornin | 14:42 |
| rosmaita | good morning | 14:42 |
| hemna_ | https://review.opendev.org/#/c/696857/2 | 14:44 |
| hemna_ | so, is it really the plan to remove that driver in the next release? | 14:44 |
| hemna_ | can we just leave it as unsupported? | 14:44 |
| rosmaita | well, it has never been tested with py3 and we are py3 only in ussuri | 14:45 |
| rosmaita | and the vendor really seems not interested in it at all | 14:45 |
| hemna_ | that sucks | 14:45 |
| rosmaita | so i'd really liike to remove it in ussuri | 14:45 |
| rosmaita | agree that it sucks | 14:46 |
| hemna_ | this is not good news for openstack/cinder really | 14:46 |
| hemna_ | that would leave us with only 1 fczm driver | 14:46 |
| hemna_ | and I think most folks use brocade, or they had | 14:46 |
| hemna_ | do the unit tests pass for py3 ? | 14:46 |
| rosmaita | yes | 14:46 |
| hemna_ | I think the HPE team still uses brocade | 14:46 |
| hemna_ | this would be a disaster for FC | 14:47 |
| rosmaita | i don't know what the stats are on usage, maybe jungleboyj knows | 14:47 |
| *** eharney has joined #openstack-cinder | 14:47 | |
| rosmaita | but customers haven't convinced brocade that it's worth maintaining | 14:47 |
| rosmaita | that email to customers was pretty definitive that support ends with train | 14:47 |
| hemna_ | I doubt most customers are being vocal about it | 14:47 |
| hemna_ | we might as well rip out al fibre channel support then | 14:48 |
| hemna_ | without a functional fczm, then it's pretty pointless | 14:48 |
| rosmaita | we'd have to wait for V to do that | 14:49 |
| jungleboyj | Yeah. It is not a good thing that they are removing it. | 14:49 |
| rosmaita | but if you are serious, we need to deprecate it in U | 14:49 |
| hemna_ | no, it's a disaster for FC | 14:49 |
| rosmaita | are you east or west coast this week? | 14:49 |
| hemna_ | east coast | 14:49 |
| rosmaita | ok, maybe we should discuss this at the meeting tomorrow | 14:49 |
| jungleboyj | The sad thing is that it is going to be quite a while before customers get to the point it is removed and tehy realize the mess. | 14:49 |
| jungleboyj | rosmaita: ++ | 14:49 |
| hemna_ | yup | 14:50 |
| hemna_ | most deployments are 2 or 3 releases old | 14:50 |
| rosmaita | jungleboyj: let me know what you think about my comment on https://review.opendev.org/#/c/696857/2 | 14:50 |
| rosmaita | hemna_: at least -- i know there are a lot of people still on newton | 14:51 |
| *** tkajinam has quit IRC | 14:52 | |
| hemna_ | looks like the HPE FC CI isn't setup to use fczm | 14:52 |
| hemna_ | but I know their setup has a brocade switch | 14:53 |
| hemna_ | maybe we can see if they can enable it | 14:53 |
| hemna_ | since they really should be doing CI with FCZM anyway | 14:53 |
| *** lbragsta_ has joined #openstack-cinder | 15:00 | |
| *** pcaruana has joined #openstack-cinder | 15:01 | |
| *** rosmaita has quit IRC | 15:06 | |
| *** senrique_ has joined #openstack-cinder | 15:08 | |
| *** rosmaita has joined #openstack-cinder | 15:09 | |
| *** enriquetaso has quit IRC | 15:10 | |
| jungleboyj | hemna_: ++ | 15:17 |
| jungleboyj | rosmaita: So, I looked at your comment. | 15:17 |
| rosmaita | ok | 15:18 |
| jungleboyj | Not sure the right answer there. | 15:18 |
| jungleboyj | Lets talk in the meeting tomorrow. | 15:19 |
| rosmaita | sounds good | 15:19 |
| jungleboyj | hemna_: Can you try reaching out to your contacts at HPE and see if they can maybe run things using the FCZM? | 15:19 |
| *** lbragsta_ has quit IRC | 15:36 | |
| *** lbragsta_ has joined #openstack-cinder | 15:37 | |
| *** spatel has joined #openstack-cinder | 15:50 | |
| *** lbragsta_ has quit IRC | 15:55 | |
| *** jmlowe has joined #openstack-cinder | 16:00 | |
| *** ociuhandu has quit IRC | 16:02 | |
| *** awalende has quit IRC | 16:08 | |
| *** awalende has joined #openstack-cinder | 16:08 | |
| *** awalende has quit IRC | 16:13 | |
| *** lpetrut has quit IRC | 16:29 | |
| *** Liang__ has quit IRC | 16:33 | |
| *** lpetrut has joined #openstack-cinder | 16:39 | |
| *** awalende has joined #openstack-cinder | 16:45 | |
| *** mmethot|conferen is now known as mmethot | 16:48 | |
| *** jmlowe has quit IRC | 16:50 | |
| *** awalende has quit IRC | 16:50 | |
| *** lpetrut has quit IRC | 16:50 | |
| *** senrique__ has joined #openstack-cinder | 16:54 | |
| *** senrique_ has quit IRC | 16:57 | |
| *** RyanLiang has quit IRC | 16:58 | |
| *** tesseract has quit IRC | 16:58 | |
| *** ociuhandu has joined #openstack-cinder | 17:25 | |
| *** tosky has quit IRC | 17:27 | |
| *** ociuhandu has quit IRC | 17:29 | |
| openstackgerrit | Merged openstack/cinderlib master: Fix complex configuration options https://review.opendev.org/696375 | 17:42 |
| openstackgerrit | Merged openstack/cinderlib master: Add missing release notes https://review.opendev.org/698251 | 17:42 |
| *** davidsha has quit IRC | 17:45 | |
| *** senrique__ is now known as enriquetaso | 17:50 | |
| rosmaita | geguileo: cinderlib release: https://review.opendev.org/698306 | 17:51 |
| geguileo | rosmaita: thanks, LGTM, though I don't know much about delivery, so it dones't count for much | 17:54 |
| *** e0ne has quit IRC | 17:55 | |
| rosmaita | geguileo: key thing to check is to make sure the hash is the one currently at the point we want to cut the release (so HEAD right now in master) | 17:55 |
| geguileo | rosmaita: that one I did check XD | 17:56 |
| rosmaita | great! | 17:57 |
| *** martinkennelly has quit IRC | 18:12 | |
| *** sahid has quit IRC | 18:17 | |
| *** ociuhandu has joined #openstack-cinder | 18:39 | |
| *** ociuhandu has quit IRC | 18:44 | |
| *** mmethot has quit IRC | 18:46 | |
| *** mmethot has joined #openstack-cinder | 18:46 | |
| *** psachin has quit IRC | 19:02 | |
| *** efried is now known as efried_afk | 19:06 | |
| *** e0ne has joined #openstack-cinder | 19:07 | |
| *** spatel has quit IRC | 19:16 | |
| *** spatel has joined #openstack-cinder | 19:33 | |
| *** gmann is now known as gmann_afk | 19:40 | |
| *** eharney has quit IRC | 19:55 | |
| *** awalende has joined #openstack-cinder | 20:09 | |
| *** awalende has quit IRC | 20:14 | |
| *** whfnst has quit IRC | 20:15 | |
| *** rcernin has joined #openstack-cinder | 20:16 | |
| *** pcaruana has quit IRC | 20:16 | |
| *** e0ne has quit IRC | 20:17 | |
| *** sorrison has quit IRC | 20:24 | |
| *** jcosmao has joined #openstack-cinder | 20:40 | |
| *** jcosmao has left #openstack-cinder | 20:40 | |
| *** jcosmao has joined #openstack-cinder | 20:42 | |
| *** eharney has joined #openstack-cinder | 21:02 | |
| *** smcginnis has joined #openstack-cinder | 21:05 | |
| *** ChanServ sets mode: +o smcginnis | 21:05 | |
| smcginnis | hemna_, rosmaita: Reading eavedrop for what I missed - Angela from Brocade reached out directly to jungleboyj and I and asked that we remove the driver for her since she has no time to work on it. | 21:08 |
| smcginnis | We do still have the Cisco driver, since the odd twist of fate is that now Cisco is keeping up their CI but Brocade cannot. | 21:08 |
| smcginnis | rosmaita: To you comment on there, I do still think we need to mark it unsupported in U first and remove in V. | 21:09 |
| smcginnis | There may be users that do not have current support contracts and therefore would not have been notified that Brocade is no longer supporting it. | 21:09 |
| rosmaita | smcginnis: i am worried about it not being tested with py3 | 21:09 |
| smcginnis | Yeah, me too. | 21:09 |
| smcginnis | But I think if it's there and flagged unsupported, that covers two things. | 21:10 |
| smcginnis | a) It's clear to the end user and they need to go out of their way to run it. | 21:10 |
| smcginnis | And b) it's still there, so if they want to hack some temporary py3 fixes, they can a little more easily than needing to pull down an old copy of the code. | 21:11 |
| smcginnis | My 2 cents anyway. | 21:11 |
| rosmaita | ok, it's on the agenda for tomorrow, so we can discuss some more | 21:11 |
| smcginnis | ++ | 21:11 |
| rosmaita | i wonder if we should go ahead and backport the supported=false to train even if we keep the driver in u | 21:12 |
| smcginnis | One other food for thought ahead of that - great if HPE wants to run CI, but we would also need them to commit to also maintaining the code too, not just testing it. | 21:12 |
| smcginnis | I don't think so. We really should never backport deprecations. | 21:12 |
| smcginnis | Nasty surprise if someone upgrades a point release and suddenly something stops working. | 21:12 |
| smcginnis | Yeah, they can set the flag in cinder.conf, but very risky that by some fluke they wouldn't actually carefully read the documentation. ;) | 21:13 |
| rosmaita | yeah | 21:13 |
| rosmaita | too bad brocade didn't make that decision a few weeks earlier | 21:13 |
| smcginnis | Yeah, totally agree. | 21:13 |
| smcginnis | They probably did, but just didn't get to actually making that change upstream. | 21:14 |
| *** spatel has quit IRC | 22:09 | |
| *** awalende has joined #openstack-cinder | 22:10 | |
| *** awalende has quit IRC | 22:15 | |
| *** sorrison has joined #openstack-cinder | 22:17 | |
| *** efried_afk is now known as efried | 22:25 | |
| *** awalende has joined #openstack-cinder | 22:30 | |
| *** awalende has quit IRC | 22:35 | |
| efried | o/ cinder, anyone about? | 22:50 |
| efried | jungleboyj, rosmaita, smcginnis: ? | 22:51 |
| jungleboyj | Yep. | 22:51 |
| rosmaita | efried: what's up? | 22:51 |
| efried | I draw your attention to: https://opendev.org/openstack/cinder/src/branch/master/cinder/compute/nova.py#L145 | 22:51 |
| efried | afaict this is used by cinder to send events to nova when volumes are extended. | 22:51 |
| efried | I believe there is a bug in there: come with me... | 22:52 |
| jungleboyj | Yes, that looks familiar. | 22:52 |
| * jungleboyj picks up a torch | 22:52 | |
| efried | I believe the loop on L161 triggers if the response code is 207, whereupon the events in the response payload are enumerated with a `code` that indicates their individual success or failure. | 22:53 |
| efried | the result of returning False from this method is that you will send a | 22:54 |
| efried | NOTIFY_COMPUTE_SERVICE_FAILED = ( | 22:54 |
| efried | '009', | 22:54 |
| efried | _("Compute service failed to extend volume.")) | 22:54 |
| efried | message over your message API | 22:54 |
| efried | is this just a "let the user know something went wrong" thing, or does it cause something more involved to happen? | 22:54 |
| efried | (I haven't gotten to the bug yet...) | 22:54 |
| efried | I believe the bug occurs if a subset of the events "fail" (e.g. the instance is gone, or in the process of being scheduled an not yet associated with a host, that kind of thing): in that case you'll happily log the "failed status" message, but you *won't* set response_error to True, so you won't send that message over the message API. | 22:56 |
| efried | I scared you off | 22:59 |
| jungleboyj | No. Just thinking. | 22:59 |
| efried | This is probably a) pretty low probability, since it would have to be multiple instances where a subset failed; and b) not very visible -- are users likely to complain about not getting a message informing them of a failure by Nova to extend the volume? Wouldn't that failure manifest in some other way? Or if they don't notice, then it doesn't matter? | 23:00 |
| jungleboyj | So, I don't know that code path very well. | 23:01 |
| jungleboyj | I was just thinking. If we have gotten to that point it means the volume should have been extended. | 23:01 |
| efried | By cinder, yes. But it would mean that the nova side punted. | 23:01 |
| jungleboyj | The impact would be that Nova wouldn't cause the scsi bus to be rescanned and the user wouldn't see the new size. | 23:02 |
| efried | Right, but the only effect of the "bug" is that cinder doesn't alert the user of that. | 23:02 |
| jungleboyj | Which isn't a super unusual situation. The users sees that gets irritated and rescans the bus or unmounts and remounts and goes on with life. | 23:03 |
| efried | Okay. | 23:03 |
| efried | "How did you get here in the first place efried?" | 23:03 |
| efried | I'm glad you asked. | 23:03 |
| efried | I recently "fixed" the external events API: https://review.opendev.org/#/c/698037/ | 23:03 |
| efried | What this means to you is that this particular corner case becomes higher probability. | 23:03 |
| efried | Previously it would only happen if a) multiple instances and b) a *subset* had failures. | 23:04 |
| efried | Now it will happen any time *all* events fail -- whether multiple instances or one instance. | 23:04 |
| jungleboyj | Ah, interesting. | 23:04 |
| efried | But perhaps this was worth fixing in any case | 23:05 |
| *** tkajinam has joined #openstack-cinder | 23:05 | |
| efried | Let me WIP a patch right quick. | 23:05 |
| efried | ...though I may need to pass the torch to someone cinder-savvy cause I have no idea how to test it. | 23:05 |
| *** gmann_afk is now known as gmann | 23:06 | |
| efried | this is a case of *one* volume attached to *multiple* instances, right? | 23:06 |
| jungleboyj | The extend case? No, that could be one volume to one instance. | 23:07 |
| efried | Sorry, I mean it's always just one volume though? | 23:08 |
| efried | jungleboyj: does cinder storyboard or launchpad these days? | 23:10 |
| jungleboyj | Right. Just one volume. | 23:11 |
| jungleboyj | We still Launchpad. | 23:11 |
| * jungleboyj hides from the sb police | 23:11 | |
| efried | okay, I'll open a bug, since this might be backport-worthy | 23:11 |
| jungleboyj | Cool. Sounds good. | 23:11 |
| efried | Worth clarifying I suppose: the error from nova means it didn't even accept the event. Which means it didn't even try to extend the volume. It's still possible the event is accepted, but nova pukes doing the extend. So there are surely still cases where things go pear-shaped but the user doesn't get a message about it. | 23:13 |
| jungleboyj | Ok. With reporting those kinds of failures it is usually a 'best effort' approach. | 23:15 |
| jungleboyj | smcginnis: ^^^ | 23:15 |
| jungleboyj | Thoughts on Eric's concern here? Looks like you were the last one working in that path? | 23:16 |
| jungleboyj | I need to run for a bit I will follow up with Sean and see what he thinks. | 23:17 |
| efried | https://bugs.launchpad.net/cinder/+bug/1855940 | 23:21 |
| openstack | Launchpad bug 1855940 in Cinder "Messages sometimes not sent when nova fails to accept "extend volume" events" [Undecided,New] | 23:21 |
| openstackgerrit | Eric Fried proposed openstack/cinder master: WIP: Always message on nova volume extend failures https://review.opendev.org/698340 | 23:24 |
| efried | and ^ | 23:24 |
| efried | Will let zuul tell me where the tests are :) | 23:24 |
| efried | I'm going to bail for the day, will check back tomorrow. Thanks for the help Jay. | 23:24 |
| jungleboyj | Ok. Sounds good. Thank you! | 23:26 |
| *** enriquetaso has quit IRC | 23:30 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!