Tuesday, 2019-12-10

*** ociuhandu has quit IRC00:01
*** rcernin_ has joined #openstack-cinder00:04
*** rcernin has quit IRC00:07
*** ociuhandu has joined #openstack-cinder00:10
*** ociuhandu has quit IRC00:14
*** ociuhandu has joined #openstack-cinder00:17
*** rcernin_ has quit IRC00:19
*** ociuhandu has quit IRC00:21
*** ociuhandu has joined #openstack-cinder00:23
*** rcernin has joined #openstack-cinder00:27
*** ociuhandu has quit IRC00:28
*** awalende has joined #openstack-cinder00:28
*** dave-mccowan has joined #openstack-cinder00:29
*** ociuhandu has joined #openstack-cinder00:31
*** dave-mccowan has quit IRC00:33
*** awalende has quit IRC00:33
*** ociuhandu has quit IRC00:36
*** dave-mccowan has joined #openstack-cinder00:36
*** ociuhandu has joined #openstack-cinder00:41
*** ociuhandu has quit IRC00:46
*** zhanglong has joined #openstack-cinder00:47
*** brinzhang has joined #openstack-cinder00:50
*** ociuhandu has joined #openstack-cinder00:51
*** dave-mccowan has quit IRC00:52
*** ociuhandu has quit IRC00:57
*** ociuhandu has joined #openstack-cinder01:02
*** ociuhandu has quit IRC01:06
*** ircuser-1 has joined #openstack-cinder01:07
*** ociuhandu has joined #openstack-cinder01:12
*** enriquetaso has quit IRC01:12
*** ociuhandu has quit IRC01:16
*** Liang__ has joined #openstack-cinder01:18
*** ociuhandu has joined #openstack-cinder01:21
*** zhanglong has quit IRC01:25
*** ociuhandu has quit IRC01:26
*** zhanglong has joined #openstack-cinder01:27
*** ociuhandu has joined #openstack-cinder01:29
*** ociuhandu has quit IRC01:37
*** ociuhandu has joined #openstack-cinder01:42
*** ociuhandu has quit IRC01:47
*** ociuhandu has joined #openstack-cinder01:49
*** whfnst has joined #openstack-cinder01:52
*** ociuhandu has quit IRC01:57
*** ociuhandu has joined #openstack-cinder02:08
*** Xuchu has joined #openstack-cinder02:11
*** ociuhandu has quit IRC02:13
*** ociuhandu has joined #openstack-cinder02:14
*** brinzhang_ has joined #openstack-cinder02:22
*** ociuhandu has quit IRC02:22
*** brinzhang has quit IRC02:25
*** zhanglong has quit IRC02:27
*** zhanglong has joined #openstack-cinder02:32
*** lvbin01 has quit IRC02:36
*** ociuhandu has joined #openstack-cinder02:38
*** vesper11 has quit IRC02:45
*** tridde has joined #openstack-cinder02:45
*** trident has quit IRC02:46
*** vesper11 has joined #openstack-cinder02:46
*** ociuhandu has quit IRC02:47
*** lvbin01 has joined #openstack-cinder02:50
*** brinzhang has joined #openstack-cinder02:56
openstackgerritPeng Wang proposed openstack/cinder master: DS8k Cinder Driver support Python3  https://review.opendev.org/69796002:57
*** brinzhang_ has quit IRC02:59
*** brinzhang_ has joined #openstack-cinder03:00
*** brinzhang has quit IRC03:03
*** brinzhang has joined #openstack-cinder03:12
*** brinzhang_ has quit IRC03:15
*** Xuchu_ has joined #openstack-cinder03:20
*** Xuchu has quit IRC03:23
*** Xuchu_ is now known as Xuchu03:23
*** psachin has joined #openstack-cinder03:31
*** awalende has joined #openstack-cinder03:39
*** brinzhang_ has joined #openstack-cinder03:42
*** awalende has quit IRC03:44
*** brinzhang has quit IRC03:46
*** ociuhandu has joined #openstack-cinder03:47
*** ociuhandu has quit IRC03:51
*** bhagyashris has joined #openstack-cinder04:23
*** rishabhhpe has joined #openstack-cinder04:30
*** ociuhandu has joined #openstack-cinder04:41
*** udesale has joined #openstack-cinder04:41
*** brinzhang_ has quit IRC04:44
*** brinzhang_ has joined #openstack-cinder04:45
*** ociuhandu has quit IRC04:46
*** brinzhang_ has quit IRC04:46
*** brinzhang_ has joined #openstack-cinder04:47
*** brinzhang has joined #openstack-cinder04:52
*** brinzhang has quit IRC04:53
*** brinzhang has joined #openstack-cinder04:54
*** brinzhang_ has quit IRC04:55
*** brinzhang_ has joined #openstack-cinder05:12
*** brinzhang_ has quit IRC05:12
*** brinzhang_ has joined #openstack-cinder05:13
*** brinzhang has quit IRC05:15
*** bhagyashris has quit IRC05:20
*** ociuhandu has joined #openstack-cinder05:35
*** ociuhandu has quit IRC05:40
*** bhagyashris has joined #openstack-cinder06:00
*** sapd1 has joined #openstack-cinder06:04
*** Luzi has joined #openstack-cinder06:04
*** e0ne has joined #openstack-cinder06:09
*** dpawlik has joined #openstack-cinder06:12
*** dpawlik has quit IRC06:18
*** e0ne has quit IRC06:19
*** dpawlik has joined #openstack-cinder06:20
*** whoami-rajat has joined #openstack-cinder06:21
*** dpawlik has quit IRC06:27
*** ociuhandu has joined #openstack-cinder06:29
*** ociuhandu has quit IRC06:34
*** psachin has quit IRC06:42
*** dpawlik has joined #openstack-cinder06:43
*** psachin has joined #openstack-cinder06:49
*** brinzhang has joined #openstack-cinder06:50
openstackgerritXuan Yandong proposed openstack/python-cinderclient master: Drop support for python 2  https://review.opendev.org/69146706:53
*** brinzhang_ has quit IRC06:54
*** lpetrut has joined #openstack-cinder07:00
*** ociuhandu has joined #openstack-cinder07:02
*** ociuhandu has quit IRC07:06
*** sapd1 has quit IRC07:14
*** brinzhang_ has joined #openstack-cinder07:24
*** brinzhang_ has quit IRC07:26
*** ociuhandu has joined #openstack-cinder07:26
*** brinzhang_ has joined #openstack-cinder07:26
*** brinzhang has quit IRC07:27
*** brinzhang_ has quit IRC07:28
*** brinzhang_ has joined #openstack-cinder07:28
*** zhanglong has quit IRC07:29
*** brinzhang_ has quit IRC07:30
*** brinzhang_ has joined #openstack-cinder07:30
*** zhanglong has joined #openstack-cinder07:30
*** brinzhang_ has quit IRC07:31
*** ociuhandu has quit IRC07:32
*** pcaruana has joined #openstack-cinder07:39
*** ociuhandu has joined #openstack-cinder07:46
*** sahid has joined #openstack-cinder07:51
*** rcernin has quit IRC07:54
*** rishabhhpe has quit IRC07:55
*** bhagyashris has quit IRC08:05
*** tkajinam has quit IRC08:06
*** tosky has joined #openstack-cinder08:11
*** rishabhhpe has joined #openstack-cinder08:15
*** awalende has joined #openstack-cinder08:18
*** tesseract has joined #openstack-cinder08:27
*** RyanLiang has joined #openstack-cinder08:42
*** ociuhandu has quit IRC08:48
*** raghavendrat has quit IRC09:00
*** Luzi has quit IRC09:11
*** sam_wan has quit IRC09:14
*** ociuhandu has joined #openstack-cinder09:17
*** martinkennelly has joined #openstack-cinder09:19
*** ociuhandu has quit IRC09:22
*** spsurya has joined #openstack-cinder09:25
*** Luzi has joined #openstack-cinder09:26
*** ociuhandu has joined #openstack-cinder09:42
openstackgerritLenny Verkhovsky proposed openstack/cinder master: Fix open tempfile.NamedTemporaryFile as string for Python3  https://review.opendev.org/69820209:44
*** davidsha has joined #openstack-cinder09:46
*** martinkennelly has quit IRC09:46
*** e0ne has joined #openstack-cinder09:54
*** bhagyashris has joined #openstack-cinder10:22
*** Liang__ has quit IRC10:26
*** martinkennelly has joined #openstack-cinder10:28
*** bhagyashris has quit IRC10:50
*** zhanglong has quit IRC10:53
*** bhagyashris has joined #openstack-cinder10:58
openstackgerritxinding proposed openstack/cinder master: Snapshot does not expand after expanding volume  https://review.opendev.org/69821711:04
*** kambiz has quit IRC11:09
*** ociuhandu has quit IRC11:13
*** ociuhandu has joined #openstack-cinder11:13
*** kambiz has joined #openstack-cinder11:14
*** ociuhandu has quit IRC11:28
*** rishabhhpe has quit IRC11:34
*** tommylikehu has joined #openstack-cinder11:37
*** bhagyashris has quit IRC11:42
*** bhagyashris has joined #openstack-cinder11:42
*** udesale has quit IRC12:01
*** lvbin01 has quit IRC12:02
*** lvbin01 has joined #openstack-cinder12:17
*** ociuhandu has joined #openstack-cinder12:23
*** ociuhandu has quit IRC12:27
*** spsurya has quit IRC12:35
*** Xuchu has quit IRC12:51
*** bhagyashris has quit IRC12:58
*** dpawlik has quit IRC13:01
*** enriquetaso has joined #openstack-cinder13:04
*** jgriffith has joined #openstack-cinder13:29
*** whoami-rajat has quit IRC13:31
*** yaawang has quit IRC13:39
*** yaawang has joined #openstack-cinder13:39
*** Xuchu has joined #openstack-cinder13:56
*** Liang__ has joined #openstack-cinder14:00
*** tkajinam has joined #openstack-cinder14:01
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Add missing release notes  https://review.opendev.org/69825114:07
*** eharney has quit IRC14:07
*** geguileo has joined #openstack-cinder14:08
geguileorosmaita: Promised missing cinderlib release notes: https://review.opendev.org/69825114:08
rosmaitageguileo: ty!14:09
rosmaitadid the config patch get merged yet?14:09
geguileorosmaita: not yet14:10
rosmaitahmmmm ... we will have to twist some arms14:11
rosmaitacinder 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 release14:11
rosmaitajungleboyj: e0ne: whoami-rajat: smcginnis: eharney: ^^14:12
rosmaitahemna_: ^^ (sorry, did not mean to leave you out)14:12
*** awalende has quit IRC14:13
*** awalende has joined #openstack-cinder14:14
*** awalende has quit IRC14:14
*** awalende has joined #openstack-cinder14:14
*** Xuchu has quit IRC14:23
e0nerosmaita: looking on it14:24
rosmaitae0ne: ty!14:24
e0nerosmaita: +2'ed.14:29
*** Luzi has quit IRC14:29
rosmaitathanks! appreciate it14:29
e0neI need to do more reviews to cinderlib. I'm still feeling uncomfortable to approve patches14:29
e0nerosmaita: feel free to approve this patch now :)14:33
*** ociuhandu has joined #openstack-cinder14:35
*** mmethot has joined #openstack-cinder14:35
*** mmethot is now known as mmethot|conferen14:35
*** pcaruana has quit IRC14:36
rosmaitagood morning14:42
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
rosmaitawell, it has never been tested with py3 and we are py3 only in ussuri14:45
rosmaitaand the vendor really seems not interested in it at all14:45
hemna_that sucks14:45
rosmaitaso i'd really liike to remove it in ussuri14:45
rosmaitaagree that it sucks14:46
hemna_this is not good news for openstack/cinder really14:46
hemna_that would leave us with only 1 fczm driver14:46
hemna_and I think most folks use brocade, or they had14:46
hemna_do the unit tests pass for py3 ?14:46
hemna_I think the HPE team still uses brocade14:46
hemna_this would be a disaster for FC14:47
rosmaitai don't know what the stats are on usage, maybe jungleboyj knows14:47
*** eharney has joined #openstack-cinder14:47
rosmaitabut customers haven't convinced brocade that it's worth maintaining14:47
rosmaitathat email to customers was pretty definitive that support ends with train14:47
hemna_I doubt most customers are being vocal about it14:47
hemna_we might as well rip out al fibre channel support then14:48
hemna_without a functional fczm, then it's pretty pointless14:48
rosmaitawe'd have to wait for V to do that14:49
jungleboyjYeah.  It is not a good thing that they are removing it.14:49
rosmaitabut if you are serious, we need to deprecate it in U14:49
hemna_no, it's a disaster for FC14:49
rosmaitaare you east or west coast this week?14:49
hemna_east coast14:49
rosmaitaok, maybe we should discuss this at the meeting tomorrow14:49
jungleboyjThe 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
jungleboyjrosmaita:  ++14:49
hemna_most deployments are 2 or 3 releases old14:50
rosmaitajungleboyj: let me know what you think about my comment on https://review.opendev.org/#/c/696857/214:50
rosmaitahemna_: at least -- i know there are a lot of people still on newton14:51
*** tkajinam has quit IRC14:52
hemna_looks like the HPE FC CI isn't setup to use fczm14:52
hemna_but I know their setup has a brocade switch14:53
hemna_maybe we can see if they can enable it14:53
hemna_since they really should be doing CI with FCZM anyway14:53
*** lbragsta_ has joined #openstack-cinder15:00
*** pcaruana has joined #openstack-cinder15:01
*** rosmaita has quit IRC15:06
*** senrique_ has joined #openstack-cinder15:08
*** rosmaita has joined #openstack-cinder15:09
*** enriquetaso has quit IRC15:10
jungleboyjhemna_:  ++15:17
jungleboyjrosmaita:  So, I looked at your comment.15:17
jungleboyjNot sure the right answer there.15:18
jungleboyjLets talk in the meeting tomorrow.15:19
rosmaitasounds good15:19
jungleboyjhemna_: 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 IRC15:36
*** lbragsta_ has joined #openstack-cinder15:37
*** spatel has joined #openstack-cinder15:50
*** lbragsta_ has quit IRC15:55
*** jmlowe has joined #openstack-cinder16:00
*** ociuhandu has quit IRC16:02
*** awalende has quit IRC16:08
*** awalende has joined #openstack-cinder16:08
*** awalende has quit IRC16:13
*** lpetrut has quit IRC16:29
*** Liang__ has quit IRC16:33
*** lpetrut has joined #openstack-cinder16:39
*** awalende has joined #openstack-cinder16:45
*** mmethot|conferen is now known as mmethot16:48
*** jmlowe has quit IRC16:50
*** awalende has quit IRC16:50
*** lpetrut has quit IRC16:50
*** senrique__ has joined #openstack-cinder16:54
*** senrique_ has quit IRC16:57
*** RyanLiang has quit IRC16:58
*** tesseract has quit IRC16:58
*** ociuhandu has joined #openstack-cinder17:25
*** tosky has quit IRC17:27
*** ociuhandu has quit IRC17:29
openstackgerritMerged openstack/cinderlib master: Fix complex configuration options  https://review.opendev.org/69637517:42
openstackgerritMerged openstack/cinderlib master: Add missing release notes  https://review.opendev.org/69825117:42
*** davidsha has quit IRC17:45
*** senrique__ is now known as enriquetaso17:50
rosmaitageguileo: cinderlib release: https://review.opendev.org/69830617:51
geguileorosmaita: thanks, LGTM, though I don't know much about delivery, so it dones't count for much17:54
*** e0ne has quit IRC17:55
rosmaitageguileo: 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
geguileorosmaita: that one I did check XD17:56
*** martinkennelly has quit IRC18:12
*** sahid has quit IRC18:17
*** ociuhandu has joined #openstack-cinder18:39
*** ociuhandu has quit IRC18:44
*** mmethot has quit IRC18:46
*** mmethot has joined #openstack-cinder18:46
*** psachin has quit IRC19:02
*** efried is now known as efried_afk19:06
*** e0ne has joined #openstack-cinder19:07
*** spatel has quit IRC19:16
*** spatel has joined #openstack-cinder19:33
*** gmann is now known as gmann_afk19:40
*** eharney has quit IRC19:55
*** awalende has joined #openstack-cinder20:09
*** awalende has quit IRC20:14
*** whfnst has quit IRC20:15
*** rcernin has joined #openstack-cinder20:16
*** pcaruana has quit IRC20:16
*** e0ne has quit IRC20:17
*** sorrison has quit IRC20:24
*** jcosmao has joined #openstack-cinder20:40
*** jcosmao has left #openstack-cinder20:40
*** jcosmao has joined #openstack-cinder20:42
*** eharney has joined #openstack-cinder21:02
*** smcginnis has joined #openstack-cinder21:05
*** ChanServ sets mode: +o smcginnis21:05
smcginnishemna_, 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
smcginnisWe 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
smcginnisrosmaita: To you comment on there, I do still think we need to mark it unsupported in U first and remove in V.21:09
smcginnisThere 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
rosmaitasmcginnis: i am worried about it not being tested with py321:09
smcginnisYeah, me too.21:09
smcginnisBut I think if it's there and flagged unsupported, that covers two things.21:10
smcginnisa) It's clear to the end user and they need to go out of their way to run it.21:10
smcginnisAnd 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
smcginnisMy 2 cents anyway.21:11
rosmaitaok, it's on the agenda for tomorrow, so we can discuss some more21:11
rosmaitai wonder if we should go ahead and backport the supported=false to train even if we keep the driver in u21:12
smcginnisOne 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
smcginnisI don't think so. We really should never backport deprecations.21:12
smcginnisNasty surprise if someone upgrades a point release and suddenly something stops working.21:12
smcginnisYeah, 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
rosmaitatoo bad brocade didn't make that decision a few weeks earlier21:13
smcginnisYeah, totally agree.21:13
smcginnisThey probably did, but just didn't get to actually making that change upstream.21:14
*** spatel has quit IRC22:09
*** awalende has joined #openstack-cinder22:10
*** awalende has quit IRC22:15
*** sorrison has joined #openstack-cinder22:17
*** efried_afk is now known as efried22:25
*** awalende has joined #openstack-cinder22:30
*** awalende has quit IRC22:35
efriedo/ cinder, anyone about?22:50
efriedjungleboyj, rosmaita, smcginnis: ?22:51
rosmaitaefried: what's up?22:51
efriedI draw your attention to: https://opendev.org/openstack/cinder/src/branch/master/cinder/compute/nova.py#L14522:51
efriedafaict this is used by cinder to send events to nova when volumes are extended.22:51
efriedI believe there is a bug in there: come with me...22:52
jungleboyjYes, that looks familiar.22:52
* jungleboyj picks up a torch22:52
efriedI 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
efriedthe result of returning False from this method is that you will send a22:54
efried        '009',22:54
efried        _("Compute service failed to extend volume."))22:54
efriedmessage over your message API22:54
efriedis 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
efriedI 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
efriedI scared you off22:59
jungleboyjNo.  Just thinking.22:59
efriedThis 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
jungleboyjSo, I don't know that code path very well.23:01
jungleboyjI was just thinking.  If we have gotten to that point it means the volume should have been extended.23:01
efriedBy cinder, yes. But it would mean that the nova side punted.23:01
jungleboyjThe 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
efriedRight, but the only effect of the "bug" is that cinder doesn't alert the user of that.23:02
jungleboyjWhich 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"How did you get here in the first place efried?"23:03
efriedI'm glad you asked.23:03
efriedI recently "fixed" the external events API: https://review.opendev.org/#/c/698037/23:03
efriedWhat this means to you is that this particular corner case becomes higher probability.23:03
efriedPreviously it would only happen if a) multiple instances and b) a *subset* had failures.23:04
efriedNow it will happen any time *all* events fail -- whether multiple instances or one instance.23:04
jungleboyjAh, interesting.23:04
efriedBut perhaps this was worth fixing in any case23:05
*** tkajinam has joined #openstack-cinder23:05
efriedLet 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 gmann23:06
efriedthis is a case of *one* volume attached to *multiple* instances, right?23:06
jungleboyjThe extend case?  No, that could be one volume to one instance.23:07
efriedSorry, I mean it's always just one volume though?23:08
efriedjungleboyj: does cinder storyboard or launchpad these days?23:10
jungleboyjRight.  Just one volume.23:11
jungleboyjWe still Launchpad.23:11
* jungleboyj hides from the sb police23:11
efriedokay, I'll open a bug, since this might be backport-worthy23:11
jungleboyjCool.  Sounds good.23:11
efriedWorth 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
jungleboyjOk.  With reporting those kinds of failures it is usually a 'best effort' approach.23:15
jungleboyjsmcginnis: ^^^23:15
jungleboyjThoughts on Eric's concern here?  Looks like you were the last one working in that path?23:16
jungleboyjI need to run for a bit I will follow up with Sean and see what he thinks.23:17
openstackLaunchpad bug 1855940 in Cinder "Messages sometimes not sent when nova fails to accept "extend volume" events" [Undecided,New]23:21
openstackgerritEric Fried proposed openstack/cinder master: WIP: Always message on nova volume extend failures  https://review.opendev.org/69834023:24
efriedand ^23:24
efriedWill let zuul tell me where the tests are :)23:24
efriedI'm going to bail for the day, will check back tomorrow. Thanks for the help Jay.23:24
jungleboyjOk.  Sounds good.  Thank you!23:26
*** enriquetaso has quit IRC23:30

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!