Tuesday, 2018-12-18

*** itlinux has quit IRC00:02
*** dave-mccowan has joined #openstack-cinder00:12
*** dave-mccowan has quit IRC00:18
openstackgerritSean McGinnis proposed openstack/cinder master: Revert "Synchronize all LVM operations"  https://review.openstack.org/62573800:18
lixiaoy1jungleboyj, thank you00:33
*** Liang__ has joined #openstack-cinder00:37
*** Liang__ is now known as LiangFang00:40
*** erlon_ has joined #openstack-cinder00:48
*** gouthamr_ has joined #openstack-cinder00:49
openstackgerritLisaLi proposed openstack/cinder-specs master: Driver reinitialization after failure  https://review.openstack.org/62485301:04
*** yikun has joined #openstack-cinder01:27
*** Dinesh_Bhor has joined #openstack-cinder01:48
*** Dinesh_Bhor has quit IRC01:55
*** N3l1x_ has quit IRC02:02
*** N3l1x_ has joined #openstack-cinder02:04
*** mhen has quit IRC02:22
*** mhen has joined #openstack-cinder02:25
*** psachin has joined #openstack-cinder02:41
*** hedvig_01 has quit IRC02:42
openstackgerritfutaotao proposed openstack/cinder master: Update FusionStorage Cinder Driver by using REST API  https://review.openstack.org/61995202:48
openstackgerritwanghao proposed openstack/cinder-specs master: Update query cinder resources filter by changes-since to new  https://review.openstack.org/62315402:53
*** Dinesh_Bhor has joined #openstack-cinder02:57
*** erlon_ has quit IRC02:57
*** dpawlik has joined #openstack-cinder03:06
*** dpawlik has quit IRC03:10
*** Dinesh_Bhor has quit IRC03:10
*** Dinesh_Bhor has joined #openstack-cinder03:18
*** rcernin has quit IRC03:25
*** rcernin has joined #openstack-cinder03:27
*** rcernin has quit IRC03:28
*** rcernin has joined #openstack-cinder03:28
*** Dinesh_Bhor has quit IRC03:43
*** mriedem has joined #openstack-cinder03:49
*** Dinesh_Bhor has joined #openstack-cinder04:10
*** udesale has joined #openstack-cinder04:11
*** Bhujay has joined #openstack-cinder04:20
*** Bhujay has quit IRC04:21
*** Bhujay has joined #openstack-cinder04:21
*** Bhujay has quit IRC04:22
*** Bhujay has joined #openstack-cinder04:23
*** mriedem has quit IRC04:31
*** kmalloc has quit IRC05:14
*** sdinescu has quit IRC05:39
*** itlinux has joined #openstack-cinder05:45
*** sapd1_x has joined #openstack-cinder05:58
*** avishay has joined #openstack-cinder06:28
*** bhagyashris has joined #openstack-cinder06:28
openstackgerritLisaLi proposed openstack/cinder-specs master: Driver reinitialization after failure  https://review.openstack.org/62485306:30
*** mgagne_ has quit IRC06:35
*** mgagne has joined #openstack-cinder06:39
openstackgerritYingxin Cheng proposed openstack/cinder master: [wip] Cinder RSD Driver  https://review.openstack.org/62146506:43
openstackgerritDinesh Bhor proposed openstack/cinder master: Correct the use of assertTrue  https://review.openstack.org/62580806:48
*** gouthamr has quit IRC06:55
*** Dinesh_Bhor has quit IRC07:01
*** Dinesh_Bhor has joined #openstack-cinder07:03
openstackgerritLisaLi proposed openstack/cinder master: Driver reinitialization after failure  https://review.openstack.org/61870207:03
*** Luzi has joined #openstack-cinder07:03
*** sapd1_x has quit IRC07:10
*** e0ne has joined #openstack-cinder07:11
*** alkhodos has quit IRC07:15
*** alkhodos has joined #openstack-cinder07:16
openstackgerritcaixiaoyu proposed openstack/cinder master: Report msg error when cannot revert to snapshot  https://review.openstack.org/62581807:20
*** sdinescu has joined #openstack-cinder07:24
*** gkadam has joined #openstack-cinder07:28
*** imacdonn has quit IRC07:29
*** imacdonn has joined #openstack-cinder07:29
*** avishay has quit IRC07:31
openstackgerritcaixiaoyu proposed openstack/cinder master: Reset snapshot status error  https://review.openstack.org/62582207:34
*** dpawlik has joined #openstack-cinder07:35
*** dpawlik has quit IRC07:42
*** dpawlik has joined #openstack-cinder07:56
*** noonedeadpunk[h] has joined #openstack-cinder08:07
*** noonedeadpunk has quit IRC08:09
*** Dinesh_Bhor has quit IRC08:21
*** Dinesh_Bhor has joined #openstack-cinder08:27
openstackgerritSven Wegener proposed openstack/cinder master: backup, ceph: Use subprocess from eventlet  https://review.openstack.org/61901508:36
openstackgerritRyan Liang proposed openstack/cinder stable/ocata: Unity driver: check array OE version  https://review.openstack.org/62583208:36
*** avishay has joined #openstack-cinder08:38
openstackgerritDinesh Bhor proposed openstack/cinder master: Correct the use of assertTrue  https://review.openstack.org/62580808:39
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239308:46
*** Dinesh_Bhor has quit IRC09:05
*** rcernin has quit IRC09:18
*** gkadam has quit IRC09:23
*** Bhujay has quit IRC09:24
*** Dinesh_Bhor has joined #openstack-cinder09:25
openstackgerritRyan Liang proposed openstack/cinder stable/ocata: Unity: fail to detach lun when auto zone enabled  https://review.openstack.org/62585509:29
openstackgerritYikun Jiang proposed openstack/cinder master: Move 'attach.end' notify to attachment_complete  https://review.openstack.org/62585609:30
*** arne_wiebalck has quit IRC09:36
*** arne_wiebalck has joined #openstack-cinder09:38
*** LiangFang has quit IRC09:45
*** ganso has joined #openstack-cinder09:50
*** erlon_ has joined #openstack-cinder09:51
*** sdin has joined #openstack-cinder10:00
*** Dinesh_Bhor has quit IRC10:02
*** Bhujay has joined #openstack-cinder10:02
*** sdinescu has quit IRC10:03
*** Bhujay has quit IRC10:03
*** Bhujay has joined #openstack-cinder10:04
*** bhagyashris has quit IRC10:04
*** Bhujay has quit IRC10:05
*** Bhujay has joined #openstack-cinder10:05
*** Bhujay has quit IRC10:06
*** Bhujay has joined #openstack-cinder10:07
*** erlon_ has quit IRC10:08
*** dpawlik has quit IRC10:30
*** dpawlik_ has joined #openstack-cinder10:30
*** Dinesh_Bhor has joined #openstack-cinder10:34
*** Dinesh_Bhor has quit IRC10:36
*** udesale has quit IRC10:40
*** sapd1_x has joined #openstack-cinder10:47
*** sapd1_x has quit IRC11:15
*** sapd1 has joined #openstack-cinder11:18
*** sapd1 has quit IRC11:21
*** sapd1 has joined #openstack-cinder11:22
openstackgerritcaixiaoyu proposed openstack/cinder master: Reset snapshot status available  https://review.openstack.org/62582211:27
*** erlon_ has joined #openstack-cinder11:32
*** luizbag has joined #openstack-cinder11:34
*** jmlowe has quit IRC11:37
*** jmlowe has joined #openstack-cinder11:39
*** sapd1 has quit IRC11:44
*** sapd1 has joined #openstack-cinder11:45
*** Bhujay has quit IRC11:46
*** sapd1 has quit IRC11:47
*** sapd1 has joined #openstack-cinder11:48
*** baojg has joined #openstack-cinder11:52
*** Bhujay has joined #openstack-cinder11:52
*** Bhujay has quit IRC11:53
*** Bhujay has joined #openstack-cinder11:53
*** Bhujay has quit IRC11:54
*** Bhujay has joined #openstack-cinder11:55
*** Bhujay has quit IRC11:56
*** Bhujay has joined #openstack-cinder11:56
*** Bhujay has quit IRC11:57
*** Bhujay has joined #openstack-cinder11:58
*** Bhujay has quit IRC11:59
*** Bhujay has joined #openstack-cinder11:59
*** Bhujay has quit IRC12:00
*** Bhujay has joined #openstack-cinder12:01
*** tpsilva has joined #openstack-cinder12:06
*** thgcorrea has joined #openstack-cinder12:11
*** carlos_silva has joined #openstack-cinder12:12
*** avishay has quit IRC12:14
*** sapd1 has quit IRC12:17
*** udesale has joined #openstack-cinder12:27
*** eharney has joined #openstack-cinder12:33
*** e0ne has quit IRC12:53
*** abishop has joined #openstack-cinder12:53
*** luizbag has quit IRC13:08
*** sapd1 has joined #openstack-cinder13:18
*** luizbag has joined #openstack-cinder13:20
geguileoerlon_: ping - https://review.openstack.org/#/c/621575/213:51
lixiaoy1jungleboyj, about Our RSD driver, you said we don't say anything about Host assisted volume migration and it should addressed. What do you mean?13:56
*** e0ne has joined #openstack-cinder13:57
lixiaoy1jungleboyj, currently our plan is to use generic volume migration in driver.py13:57
jungleboyjlixiaoy1:  Ok, if the generic migration code works with your driver then you should be fine.  You just didn't have it listed in the commit message of supported functionality.13:58
lixiaoy1jungleboyj, ok, thank you14:00
jungleboyjlixiaoy1:  You are welcome.14:00
lixiaoy1jungleboyj, when is the deadline for commits in os-brick?14:00
lixiaoy1jungleboyj, smcginnis told me it should be a week before stein milestone-2, and would like to confirm with you14:01
jungleboyjlixiaoy1: I concur with smcginnis given that your os-brick change is needed for your driver so it would need to be in place for your driver to merge.14:02
lixiaoy1jungleboyj, yes, the os-brick change is needed for our driver.14:04
lixiaoy1jungleboyj, we will have a CI and it will test both the os-brick and our driver14:04
jungleboyjOk.  So it will need to be in by milestone-2 so that your driver can merge.14:05
jungleboyjlixiaoy1:  Good.14:05
lixiaoy1jungleboyj, so once our CI is ready and approves the os-brick change is good, we can merge the os-brick change. right?14:06
lixiaoy1jungleboyj, or can we merge it before our CI? as it is enhancement for current implementation14:06
*** e0ne has quit IRC14:07
jungleboyjlixiaoy1: Given that we have agreement from you that your CI will test it also I think we can merge that sooner to make sure that things are available for your driver.14:08
jungleboyjlixiaoy1:  The big thing is if you are going to be able to close the required functionality gap in your driver before milestone-214:08
lixiaoy1jungleboyj, ok, got it. thanks. We will try to make our codes and CI ready as soon as possible14:10
jungleboyjlixiaoy1:  Great.  Thank you.14:10
lixiaoy1jungleboyj, :) thank you for your help14:11
*** psachin has quit IRC14:12
arne_wiebalcklixiaoy1: You have a moment?14:16
*** egonzalez has joined #openstack-cinder14:20
egonzalezHi, is there any progress on this bug? https://bugs.launchpad.net/cinder/+bug/1808598 kolla-ansible have all cinder jobs broken14:21
openstackLaunchpad bug 1808598 in Cinder "db migration issue on CentOS" [High,Confirmed]14:21
*** mriedem has joined #openstack-cinder14:27
jungleboyjegonzalez:  Not yet.  I was trying to get an environment up yesterday to confirm it and try some possible fixes but kept hitting other issues.14:33
thgcorrea@geguileo, hi Gorka I'm working on this bug https://bugs.launchpad.net/cinder/+bug/1797181. I saw that nova checks the volume status for 'retyping' or 'migrating' [1], so I traced it back to cinder and I realized that the status is changed when retyping [2] but not when migrating [3]. Since you have committed these changes in cinder I was wondering whether you have some information about it I must consider or I can14:40
thgcorreaadd the status change.14:40
thgcorrea[1] https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L577314:40
thgcorrea[2] https://github.com/openstack/cinder/blob/master/cinder/volume/api.py#L168114:40
thgcorrea[3] https://github.com/openstack/cinder/blob/master/cinder/volume/api.py#L150214:40
openstackLaunchpad bug 1797181 in Cinder "Migration of attached volumes does not delete old volume" [High,New]14:40
*** e0ne has joined #openstack-cinder14:43
*** dpawlik_ has quit IRC14:47
*** Luzi has quit IRC14:50
*** Bhujay has quit IRC14:53
lixiaoy1arne_wiebalck, hi14:54
arne_wiebalcklixiaoy1: Hi.14:54
arne_wiebalckIt’s about the deferred deletion patch.14:55
arne_wiebalcklixiaoy1: Actually, I just posted a response.14:55
lixiaoy1andreaf, let me check14:56
arne_wiebalcklixiaoy1: The main point is about the support for the trash API in older versions of Ceph.14:57
arne_wiebalcklixiaoy1: Or better the lack of support, and how to handle it.14:57
lixiaoy1lixiaoy1, old version of Ceph, can we just check the version of librbd?14:59
*** itlinux has quit IRC15:00
arne_wiebalcklixiaoy1: I guess we could. Technically.15:00
arne_wiebalcklixiaoy1: Probably, a hasatttr() call would do.15:01
geguileothgcorrea: I'm looking at the bug now15:01
arne_wiebalcklixiaoy1: I was planning to do sth like that now when starting the periodic purge.15:01
arne_wiebalcklixiaoy1: And to log an error like “Ceph version does not support trash”.15:02
arne_wiebalcklixiaoy1: I would then also fail the start of the corresponding daemone for that backend.15:03
arne_wiebalcklixiaoy1: This should make it bvious that something is wrong :)15:03
arne_wiebalcklixiaoy1: s/bvious/obvious/15:03
*** xyang has joined #openstack-cinder15:04
arne_wiebalcklixiaoy1: Does thts sound sensible?15:05
lixiaoy1arne_wiebalck, do you know which error shows "ceph version doesn't support trash"15:06
arne_wiebalcklixiaoy1: What do you mean? I’d add code to do that when the admin has configured deferred deletion, but the the Ceph version on the controller does not support it.15:07
lixiaoy1andreaf, let us start from the beginning. My understanding is When initializing rbd driver, if users enable deferred deletion, it checks whether librbd supports this feature15:11
arne_wiebalcklixiaoy1: No.15:11
lixiaoy1andreaf, if it supports, it starts periodic task to do purge; if it doesn't it follows current flow15:11
arne_wiebalcklixiaoy1: Not at the moment.15:11
lixiaoy1arne_wiebalck, what is your plan?15:12
lixiaoy1andreaf, yes, I know. Current codes is not in this way15:12
arne_wiebalcklixiaoy1: ok :)15:12
arne_wiebalcklixiaoy1: What I suggest is to add a check in check_setup() or whatever the function is called and not start the backend when deferred deletion is configured, but not dupported.15:13
arne_wiebalcklixiaoy1: s/dupported/supported/15:13
arne_wiebalcklixiaoy1: I don’t think it’s a good idea to fall back to something else automatically.15:14
arne_wiebalcklixiaoy1: If the admin configured this non-default behavior we should not fall back to the default, but rather signal that the config does not work.15:14
lixiaoy1arne_wiebalck, you mean we don't set rbd as initialized, right?15:15
arne_wiebalcklixiaoy1: I mean check_for_setup_error().15:15
openstackgerritHelen Walsh proposed openstack/cinder stable/queens: VMAX Driver - VMAX OS Upgrade Bug  https://review.openstack.org/62593915:15
arne_wiebalcklixiaoy1: And raise an exception in there.15:16
arne_wiebalcklixiaoy1: Correct, rbd would not be initialised.15:16
*** dpawlik has joined #openstack-cinder15:16
lixiaoy1arne_wiebalck, I see. if you raise an exception here, rbd driver will be uninitialized. and it can't be used15:16
arne_wiebalcklixiaoy1: Correct.15:16
arne_wiebalcklixiaoy1: I think this is fair, if combined with a clear error message to tell the admin why the initialisation failed.15:17
geguileothgcorrea: I didn't remove the 'migrating' status when I touched that piece of code (https://github.com/openstack/cinder/commit/135e094f33e39e4c80e32b0237910c355513a66f) let me check where we are setting the status15:17
arne_wiebalcklixiaoy1: And the try/catch code in try_remove_volume() would not be needed any more.15:18
lixiaoy1arne_wiebalck, I think it is ok to raise an exception15:19
arne_wiebalcklixiaoy1: What about catching the AttributeError when removing the volume?15:20
lixiaoy1arne_wiebalck, in try_remove_volume, if it raises an exception by other reason, if we don't handle, I am not sure the result15:20
*** dpawlik has quit IRC15:21
lixiaoy1arne_wiebalck, AttributeError, what does the error mean by?15:21
*** mchlumsky has joined #openstack-cinder15:22
arne_wiebalcklixiaoy1: I think this is raised when you try to access sth that does not exist.15:22
*** eharney has quit IRC15:22
arne_wiebalcklixiaoy1: This is raised when you try to call trash_* on an older version of Ceph.15:22
smcginnis"string object".test15:23
lixiaoy1arne_wiebalck, got it15:23
arne_wiebalcksmcginnis: thx :)15:23
geguileothgcorrea: We change the migration_status to migrating on the manager https://github.com/openstack/cinder/blob/master/cinder/volume/manager.py#L237315:23
arne_wiebalcklixiaoy1: If we do the check in check_for_setup_erorr() we should be ok is my understanding.15:24
geguileothgcorrea: and it changes to completing here https://github.com/openstack/cinder/blob/master/cinder/volume/manager.py#L224315:24
arne_wiebalcklixiaoy1: Sounds like we’d be using the function for what it was made :)15:24
arne_wiebalcklixiaoy1: And as the dirver would not be initialised, we would never try a trash_* call down the line.15:25
lixiaoy1arne_wiebalck, yes, it won't be used if not initialized.15:26
arne_wiebalcklixiaoy1: I think this is clearly better than what I have now to handle this case.15:26
lixiaoy1arne_wiebalck, I am just thinking about another point: users enable deferred deletion, and driver is initialized and periodic task is running15:26
thgcorrea@geguileo that is the 'migration_status' being changed. The nova is checking the volume 'status' value, which I guess is never being set as 'migrating'15:26
lixiaoy1arne_wiebalck, but sometimes for some reason, trash_purge has errors and throw errors, how to handle15:27
geguileothgcorrea: it doesn't look like we change it to migrating15:28
geguileothgcorrea: though I remember we did at some point...15:28
lixiaoy1arne_wiebalck, I think we catch such errors, log errors, and move on15:28
arne_wiebalcklixiaoy1: It’s a fair point.15:29
lixiaoy1arne_wiebalck, that means let period task try next time15:29
* geguileo goes to check migrating...15:29
arne_wiebalcklixiaoy1: We need to make sure the periodic task is not exiting, though.15:29
arne_wiebalcklixiaoy1: We had this is production initially.15:29
arne_wiebalcklixiaoy1: And the trash filled up ...15:29
lixiaoy1arne_wiebalck, yes, the periodic task should not exit15:30
arne_wiebalcklixiaoy1: We need to make sure the exceptions are all caught.15:30
arne_wiebalcklixiaoy1: I think I added this after we realised we have this issue …15:30
* arne_wiebalck : goes reading his own code15:30
lixiaoy1arne_wiebalck, yeah15:31
arne_wiebalcklixiaoy1: Yes, I catch all exceptions exactly for this reason.15:32
lixiaoy1arne_wiebalck, sorry I am going to leave. any other things about rbd deferred deletion15:32
arne_wiebalcklixiaoy1: So, the periodic task will retry.15:32
geguileothgcorrea: Looking at the code I don't think we ever set it to migrating...15:33
arne_wiebalcklixiaoy1: Now, if trash_remove() fails repeatedly, the log will fill up with errors. At some we have to rely on the admin to realise th is not ok :)15:33
yikunhi, whoami-rajat15:33
*** mchlumsky has quit IRC15:34
lixiaoy1arne_wiebalck, so do you have any better idea?15:34
yikunI want to check in-use and also notify is called, so I use this way instead of self.assert_notify_called~15:34
arne_wiebalcklixiaoy1: This is asynchronous, so there is no direct way to send a notification upstream.15:34
arne_wiebalcklixiaoy1: I think it is sensible to expect that admins keep an eye on their deployment or have alarming in place.15:36
arne_wiebalcklixiaoy1: To notice if the logs fill up with back traces and errors I mean.15:36
lixiaoy1arne_wiebalck, hope so15:37
arne_wiebalcklixiaoy1: :-D15:37
arne_wiebalcklixiaoy1: Ok, so we got a plan? The I’d prepare a new pacth set.15:38
lixiaoy1arne_wiebalck, yeah15:38
arne_wiebalcklixiaoy1: Great. Thanks for your input!15:39
lixiaoy1arne_wiebalck, you are welcome!15:39
*** mchlumsky has joined #openstack-cinder15:47
openstackgerritYikun Jiang proposed openstack/cinder master: Move 'attach.end' notify to attachment_complete  https://review.openstack.org/62585615:52
*** sdin has quit IRC15:55
*** sdinescu has joined #openstack-cinder15:57
*** hedvig_01 has joined #openstack-cinder16:04
*** itlinux has joined #openstack-cinder16:04
arne_wiebalckjungleboyj: Regarding the docs for the deferred deletion feature,  I’m happy to add some docs, but I fail to see where that should be added … seems like none of the other rbd specific options has a doc entry … may this is what yo16:07
hedvig_01@eharney @jungleboyj , the test_volume_boot_pattern is working. After you mentioned about sync and cache, we got some lead and made a fix in our backend.16:07
arne_wiebalckjungleboyj: u meant when you said there is some more stuff to be added :)16:07
*** itlinux_ has joined #openstack-cinder16:08
hedvig_01so 1 of the two failing tests works. Other test is test_shelve_volume_backed_instance. Openstack wiki says snapshot is created when you shelve. However, while analysing the logs. I do not see a snapshot call coming in. Is this expected?16:09
*** itlinux has quit IRC16:12
*** dave-mccowan has joined #openstack-cinder16:14
*** dave-mccowan has quit IRC16:18
*** dave-mccowan has joined #openstack-cinder16:21
*** dave-mccowan has quit IRC16:22
*** dave-mccowan has joined #openstack-cinder16:28
hedvig_01@jungleboyj , @eharney, @smcginnis : test_shelve_volume_backed_instance, writes data, sheleve and unshelve and reads data. What I observe, write data, (extra step added by me : I read data  by calling get_timestamp and I get the time stamp), then shelve unshelve happens and data is missing16:33
hedvig_01Could you please help us track down this issue?16:34
openstackgerritHelen Walsh proposed openstack/cinder stable/rocky: VMAX driver doc - notification of fix on AFA OS upgrade issue  https://review.openstack.org/62595016:34
whoami-rajatyikun: replied on the patch :)16:35
*** noonedeadpunk[h] is now known as noonedeadpunk16:35
jungleboyjarne_wiebalck:  Right.  Did I not send you a pointer to where I thought it should probably go?16:36
arne_wiebalckjungleboyj: That pointer is to a high level overview of Ceph :)16:37
jungleboyjarne_wiebalck: :-)  Give me a minute.16:37
arne_wiebalckjungleboyj: At the very end, it references a table, but I couldn’t find that.16:37
jungleboyjarne_wiebalck: https://docs.openstack.org/cinder/latest/configuration/block-storage/drivers/ceph-rbd-volume-driver.html16:39
jungleboyjSo, there is that page.  It at least documents the different config options.16:40
arne_wiebalckjungleboyj: Nice!16:40
arne_wiebalckjungleboyj: Now, where is that table? :-D16:40
jungleboyjOther drivers have more fully featured documents like this:  https://docs.openstack.org/cinder/latest/configuration/block-storage/drivers/dell-emc-vmax-driver.html16:40
jungleboyjIf I could get my &()*& ceph cluster to deploy I would work on improving that documentation.  :-)16:41
jungleboyjOk, so that is where your confusion is at.  Give me a minute to find that.16:41
arne_wiebalckjungleboyj: Thx!16:41
jungleboyjarne_wiebalck:  Ah, that is right.  That config table is now dynamically created based on the code.16:43
arne_wiebalckjungleboyj: I think that is autogenerated!16:43
arne_wiebalckjungleboyj: :)16:43
arne_wiebalckjungleboyj: Just grep’ed for the help text ;)16:43
arne_wiebalckjungleboyj: So, the help text from my patch should end up there as well.16:43
jungleboyjI had forgotten that that was now being autogenerated.16:43
jungleboyjOk.  Can you send a link to your patch again?16:44
arne_wiebalckjungleboyj: https://review.openstack.org/60898416:44
smcginnisIsn't that so convenient to have the config table automatically reflect the changes? :P16:45
jungleboyjsmcginnis: :-p16:45
arne_wiebalcksmcginnis: It’s awesome: it automatically answers jungleboyj’s -1 on the patch ;-P16:46
jungleboyjarne_wiebalck:  Ok, so I guess that covers that concern.16:46
arne_wiebalckjungleboyj: Ok, I’ll work on the concerns from lixiaoy1 then, as discussed earleri today.16:46
arne_wiebalckjungleboyj: Thx!16:46
jungleboyjarne_wiebalck:  So I think if you address the one LOG.debug and then anything else from lixiaoy1  it should be good to go.16:47
arne_wiebalckjungleboyj: Great, thx!16:47
*** udesale has quit IRC16:49
noonedeadpunkfolks, can you please give some review for patches https://review.openstack.org/#/q/status:open+project:openstack/cinder+branch:master+topic:fix/qos_per_gb_min ?16:50
*** e0ne has quit IRC16:50
openstackgerritHelen Walsh proposed openstack/cinder stable/rocky: VMAX driver doc - notification of hybrid deprecation  https://review.openstack.org/62595716:57
openstackgerritDirk Mueller proposed openstack/cinder stable/queens: Handle rbd.OSError on broken RBD image  https://review.openstack.org/62596017:05
hedvig_01Review-Priority+2. Could someone explain what does this mean ?17:05
smcginnishedvig_01: I just marked it as being a priority to get that reviewed.17:08
hedvig_01Thanks a lot. I'm yet to work on one of the comments. I need to add strengthen unit tests. rest of the comments are taken care of. I'm trying to get pass the last of 252 tests. It's test_shelve_volume_backed_instance. Hopefully should be done end of today.17:10
*** e0ne has joined #openstack-cinder17:13
openstackgerritDirk Mueller proposed openstack/cinder stable/pike: Handle rbd.OSError on broken RBD image  https://review.openstack.org/62596517:14
*** e0ne has quit IRC17:19
*** eharney has joined #openstack-cinder17:22
noonedeadpunksmcginnis: I've got a question regarding your comment https://review.openstack.org/#/c/625058/6/cinder/volume/manager.py17:32
noonedeadpunkthe point is, that if maximum_value is not set, it will be 0 by default, which is the minimum.17:32
noonedeadpunkSo the whole if/else will be in place, so the only thing, that might be changed, is to replace condition (per_gb_value > maximum_value) with min()17:33
noonedeadpunkare you ok with that?17:33
noonedeadpunkbut I'm not sure that it has a lot of sense, from my humble opinion. Also I may write it in single line condition, but dunno know if it's appropriate17:38
openstackgerritDirk Mueller proposed openstack/cinder stable/pike: RBD: get provisioned capacity using same connection  https://review.openstack.org/62597017:38
smcginnisnoonedeadpunk: I just commented on the patch.17:44
noonedeadpunkmake sense17:45
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Allow to use _max qos option together with per_gb  https://review.openstack.org/62505817:46
smcginnisnoonedeadpunk: Extra file picked up on that last one.17:47
*** jmlowe has quit IRC17:49
noonedeadpunkyeah, I've already saw it....17:49
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Allow to use _max qos option together with per_gb  https://review.openstack.org/62505817:49
noonedeadpunksorry for that...17:49
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Allow to use _max qos option together with per_gb  https://review.openstack.org/62505817:50
openstackgerritHelen Walsh proposed openstack/cinder master: PowerMax driver - rebranding VMAX to PowerMax  https://review.openstack.org/62065917:52
*** _alastor_ has joined #openstack-cinder17:53
*** _alastor_ has quit IRC17:55
*** _alastor_ has joined #openstack-cinder17:56
*** _alastor_ has quit IRC18:05
*** jmlowe has joined #openstack-cinder18:06
*** sapd1 has quit IRC18:08
*** itlinux_ has quit IRC18:21
*** jmlowe has quit IRC18:21
*** itlinux has joined #openstack-cinder18:21
*** jmlowe has joined #openstack-cinder18:22
*** _alastor_ has joined #openstack-cinder18:25
*** e0ne has joined #openstack-cinder18:34
*** jmlowe has quit IRC18:36
*** e0ne has quit IRC18:40
*** jmlowe has joined #openstack-cinder18:43
geguileoerlon_: ping - https://review.openstack.org/#/c/621575/218:50
*** jmlowe has quit IRC18:53
*** luizbag has quit IRC18:54
enriquetasoHey @all, Does anybody knows if  in stable/queens is available Verify Image Signature?19:02
*** e0ne has joined #openstack-cinder19:05
eharneyenriquetaso: it's not -- if you click on "Included In" on https://review.openstack.org/#/c/554166/ it will show what branches it's in19:07
*** eharney has quit IRC19:08
erlon_geguileo, hey19:11
geguileoerlon_: hi!19:11
geguileoerlon_: I've added a couple of comments on the patch19:11
erlon_will check that out19:11
*** eharney has joined #openstack-cinder19:11
geguileoerlon_: and I was actually able to test it myself  :-)19:11
geguileoerlon_: I got momentarily access to an ONTAP system  XD19:12
openstackgerritSofia Enriquez proposed openstack/cinder stable/queens: Ensure image utils don't block greenthreads  https://review.openstack.org/62599419:20
enriquetasothanks eharney19:21
*** jmlowe has joined #openstack-cinder19:21
*** mriedem is now known as mriedem_afk19:22
erlon_geguileo, about the removal. if the removal patch gets in first you would still need to add a eseries patch to backport it?19:22
geguileoerlon_: are you going to postpone the merge of the patch to get the other one in?  XD19:23
erlon_geguileo, lol, we totally can do that. The patch is ready internally but we still have a couple more steps before its pushed upstream, and we can hold it anyways19:24
geguileoerlon_: how long do you think it'll take before you give a +1 to this patch?  Do you have to test it internally?19:26
geguileoerlon_: I have a customer that is complaining about this...19:26
erlon_geguileo, I have +1 it already. Don't want to hold you back on that. We still will do our internal QA process before the end of the release, so, if we find something we still have time to patch it19:28
geguileoerlon_: if you have issues please feel free to ping me and I'll clean up after myself  ;-)19:28
geguileoerlon_: thanks19:28
erlon_geguileo, sure, no problem. I thank you!19:29
imacdonncinder-stable-maint types; simple backport needs approval, please: https://review.openstack.org/62514819:31
geguileoimacdonn: done19:32
openstackgerritMerged openstack/cinder master: Fix _per_gb_min usage with _per_gb  https://review.openstack.org/62497019:32
imacdonngeguileo: thanks!19:32
imacdonnrelated, this fix (master) seems ready for some +2's also: https://review.openstack.org/61239319:36
*** jmlowe has quit IRC19:53
*** erlon_ has quit IRC20:04
*** jmlowe has joined #openstack-cinder20:11
*** _alastor_ has quit IRC20:29
*** _alastor_ has joined #openstack-cinder20:30
*** erlon has joined #openstack-cinder20:32
*** hedvig_01 has quit IRC20:33
*** mriedem_afk is now known as mriedem20:50
openstackgerritMerged openstack/cinder master: Revert "Synchronize all LVM operations"  https://review.openstack.org/62573820:54
*** mriedem has quit IRC21:02
*** hedvig_01 has joined #openstack-cinder21:03
openstackgerritMerged openstack/cinder stable/rocky: Set message property in ImageDownloadFailed  https://review.openstack.org/62514821:04
*** mriedem has joined #openstack-cinder21:04
hedvig_01@smcginnis  :  We got an email saying Removed Review-Priority+2 . Could you please explain what this means?21:04
*** rosmaita has left #openstack-cinder21:06
smcginnishedvig_01: That +2 was removed?21:09
hedvig_01Yeah I understand that. I was checking why +2 was removed and what changed in 3 hours. why is not  a priority to get reviewed now?21:11
smcginnisIt still is.21:11
*** thingee has joined #openstack-cinder21:11
smcginnishedvig_01: But it's also going on a week without review comments being addressed.21:12
*** dpawlik has joined #openstack-cinder21:12
smcginnishedvig_01: I'll add back the priority.21:13
hedvig_01I've addressed all but strenthen test cases. We offload most of negative tests to our backedn so that our cinder code is simple. I got comments to test rest api. But there is nothing to test there. I populate the dictionary and send the rest api.21:13
smcginnishedvig_01: Helpful hint - if a core leaves comments with a -2 on your patch, at a minimum leave a comment back explaining why you are not able to address their concerns or what you think they missed.21:18
hedvig_01I did leave a comment21:18
hedvig_01there are no responses to that21:19
hedvig_01it's displayed in the link as well21:19
smcginnishedvig_01: The only comment after that one was you rechecking your CI.21:19
hedvig_01Ivan left comments on Dec 11.  I left a comment on Dec 12, 1.52pm21:21
*** dpawlik has quit IRC21:22
hedvig_01Eric left comments on Dec 12 1.19, I replied to comments on Dec 12 2pm21:22
smcginnisIvan left comments on Dec 13. CI was rechecked on Dec 17.21:22
*** rcernin has joined #openstack-cinder21:22
hedvig_01Ivan left the same comment regarding rest unit test case. my response on Dec 12, 1.52pm was for the same comment. I'm still working on those comments21:25
smcginnishedvig_01: I would recommend pointing out the previous response rather than just assuming they will somehow realize that they've overlooked it before and go back and see it now.21:26
smcginnisOr don't and just let it sit. Entirely up to you.21:26
hedvig_01okay. I shall reply to his latest comment again. Thank you for the quick response21:27
*** ganso has quit IRC21:28
*** tpsilva has quit IRC21:31
*** rcernin has quit IRC21:37
*** dpawlik has joined #openstack-cinder21:37
*** eharney has quit IRC21:40
*** dpawlik has quit IRC21:42
*** e0ne has quit IRC21:51
*** thgcorrea has quit IRC21:58
*** abishop has quit IRC22:00
*** rcernin has joined #openstack-cinder22:04
*** N3l1x_ has quit IRC22:14
smcginnisdansmith, mriedem: Was their a query to see if that LVM issue was happening?22:33
smcginnisI haven't noticed anything, but would be great if there was some data to back whether the direct-io change fixed that or not.22:33
*** rcernin has quit IRC22:37
mriedemthere are e-r graphs for the general attachment update timeout and one of the backup test failures,22:37
mriedemnote that if we are still getting hits, they could be due to (1) stable branches or (2) xenial nodes?22:39
smcginnisYeah, I not many, but that last one - http://logs.openstack.org/48/615348/8/check/nova-next/03f940e/logs/screen-c-vol.txt.gz#_Dec_18_04_49_02_80319822:39
smcginnis64 seconds on lvchange.22:40
smcginnisFrom your patch https://review.openstack.org/#/c/615348/22:40
*** rcernin has joined #openstack-cinder22:41
smcginnisOh well, thanks for the links. I'll check back on those later in the week to see if there's a noticeable change.22:41
*** rcernin has quit IRC22:43
openstackgerritMerged openstack/cinder master: Block broken requests 2.20.0  https://review.openstack.org/61734522:43
openstackgerritMerged openstack/cinder master: [api-ref] Fix the value of project_id  https://review.openstack.org/61723622:43
*** rcernin has joined #openstack-cinder22:45
*** itlinux has quit IRC22:56
*** _alastor_ has quit IRC23:37
openstackgerritMerged openstack/cinder master: NetApp SolidFire: Revert volume to snapshot  https://review.openstack.org/61421723:37
openstackgerritMerged openstack/cinder master: Add missing ws seperator between words  https://review.openstack.org/61848823:37
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Allow to use _max qos option together with per_gb  https://review.openstack.org/62505823:56
*** dave-mccowan has quit IRC23:56

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