Wednesday, 2019-02-20

« Tuesday, 2019-02-19 Index Thursday, 2019-02-21 »
*** ganso has quit IRC00:07
*** markvoelker has quit IRC00:15
*** itlinux has joined #openstack-cinder00:21
*** lixiaoy1 has joined #openstack-cinder00:25
*** Florian has joined #openstack-cinder00:33
*** lemko has quit IRC00:39
*** phuongnh has joined #openstack-cinder00:58
*** sdake has joined #openstack-cinder00:59
*** dave-mccowan has joined #openstack-cinder01:08
*** markvoelker has joined #openstack-cinder01:12
*** zhubx has joined #openstack-cinder01:16
*** itlinux has quit IRC01:28
*** Dinesh_Bhor has joined #openstack-cinder01:41
*** markvoelker has quit IRC01:44
*** bhagyashris has joined #openstack-cinder01:54
*** eharney has quit IRC02:08
*** itlinux has joined #openstack-cinder02:09
*** doubletao has joined #openstack-cinder02:18
*** whoami-rajat has joined #openstack-cinder02:31
*** sdake has quit IRC02:36
*** rosmaita has left #openstack-cinder02:39
*** markvoelker has joined #openstack-cinder02:41
*** gnufied has quit IRC02:46
*** psachin has joined #openstack-cinder03:04
*** markvoelker has quit IRC03:15
*** itlinux has quit IRC03:45
openstackgerritzengyingzhe proposed openstack/cinder master: Huawei driver refactor(3/10)  https://review.openstack.org/63807003:48
*** udesale has joined #openstack-cinder03:56
*** itlinux has joined #openstack-cinder04:02
*** itlinux has quit IRC04:10
whoami-rajatxyang:  i've fixed the group_snapshot case too. PTAL https://review.openstack.org/#/c/627359/04:10
*** itlinux has joined #openstack-cinder04:12
*** markvoelker has joined #openstack-cinder04:12
xyangwhoami-rajat: Sure.  I still prefer to set 0 as default, but since Eric has already given +2, I'm not going to block this04:12
whoami-rajatxyang: actually it was Eric's idea to use None.04:15
*** itlinux has quit IRC04:18
*** itlinux has joined #openstack-cinder04:23
*** itlinux has quit IRC04:28
*** dave-mccowan has quit IRC04:31
*** itlinux has joined #openstack-cinder04:33
*** markvoelker has quit IRC04:44
openstackgerritMerged openstack/cinder master: Include .inc files in doc8 linting  https://review.openstack.org/63794904:53
*** itlinux_ has joined #openstack-cinder05:24
*** itlinux has quit IRC05:27
*** ociuhandu_ has quit IRC05:30
*** ociuhandu has joined #openstack-cinder05:30
*** itlinux has joined #openstack-cinder05:40
*** markvoelker has joined #openstack-cinder05:42
*** itlinux_ has quit IRC05:43
*** itlinux has quit IRC05:56
*** tkajinam_ has joined #openstack-cinder05:57
*** tkajinam has quit IRC05:59
*** markvoelker has quit IRC06:15
*** itlinux has joined #openstack-cinder06:23
*** sdake has joined #openstack-cinder06:25
*** sdake has quit IRC06:27
*** itlinux has quit IRC06:28
*** itlinux has joined #openstack-cinder06:33
*** lemko has joined #openstack-cinder06:47
*** itlinux has quit IRC06:50
*** Luzi has joined #openstack-cinder06:52
*** markvoelker has joined #openstack-cinder07:01
*** itlinux has joined #openstack-cinder07:02
*** avishay has joined #openstack-cinder07:04
*** itlinux has quit IRC07:07
avishaygeguileo: Hey, thank you for your review.  When you have a minute, I replied to your comments: https://review.openstack.org/#/c/635457/07:11
*** e0ne has joined #openstack-cinder07:25
*** itlinux has joined #openstack-cinder07:26
*** udesale has quit IRC07:34
*** udesale has joined #openstack-cinder07:35
*** itlinux has quit IRC07:36
*** pcaruana has joined #openstack-cinder07:59
openstackgerritQian Mingyue proposed openstack/cinder master: Update Fujitsu ETERNUS DX driver  https://review.openstack.org/60964508:01
*** itlinux has joined #openstack-cinder08:02
*** tkajinam_ has quit IRC08:14
*** itlinux has quit IRC08:16
*** itlinux has joined #openstack-cinder08:23
*** itlinux has quit IRC08:28
*** rcernin has quit IRC08:29
*** lixiaoy1 has quit IRC08:54
*** ociuhandu has quit IRC09:00
*** ociuhandu has joined #openstack-cinder09:24
*** ociuhandu has quit IRC09:29
*** davidsha has joined #openstack-cinder09:30
*** e0ne has quit IRC09:33
geguileoavishay: done09:37
*** Dinesh_Bhor has quit IRC09:41
avishaygeguileo: Great, thanks!09:41
*** bhagyashris has quit IRC09:50
*** ociuhandu has joined #openstack-cinder09:57
*** markvoelker has quit IRC10:01
*** markvoelker has joined #openstack-cinder10:02
*** markvoelker has quit IRC10:06
*** phuongnh has quit IRC10:12
*** Florian has quit IRC10:39
*** ganso has joined #openstack-cinder10:47
*** jgriffith has quit IRC10:49
*** erlon has joined #openstack-cinder10:50
*** ianychoi has quit IRC10:59
openstackgerritRajat Dhasmana proposed openstack/cinder master: Add online migrations for untyped volumes  https://review.openstack.org/63813011:00
*** markvoelker has joined #openstack-cinder11:03
*** doubletao has quit IRC11:07
*** walshh__ has joined #openstack-cinder11:07
walshh__Is anyone available to look at https://review.openstack.org/#/c/635546/3.  Thank you11:10
*** udesale has quit IRC11:14
*** thgcorrea has joined #openstack-cinder11:27
*** markvoelker has quit IRC11:36
*** sapd1 has quit IRC11:38
openstackgerritLee Yarwood proposed openstack/cinder master: docs: Add os-migrate_volume_completion api-ref  https://review.openstack.org/63784711:40
*** e0ne has joined #openstack-cinder11:54
*** sapd1 has joined #openstack-cinder11:55
*** erlon has quit IRC12:04
*** tmm[m] has joined #openstack-cinder12:05
tmm[m]hello all12:05
tmm[m]I have a bit of a weird problem. I've got a cinder setup with ceph backends. One cinder-volume instance per backend, with cinder-backup on both. Cinder backup is configured to send backups to another ceph cluster. All of this works a treat12:06
*** erlon has joined #openstack-cinder12:06
tmm[m]however cinder backup makes snaps on the source ceph volume when creating a backup and those apparently do not get deleted12:06
tmm[m]This is causing some issues, is there a way to tell cinder backup to delete the snaps it no longer needs?12:07
tmm[m](I imagine this is for incremental backups, but it doesn't need to keep all old snaps does it?)12:07
*** erlon has quit IRC12:07
*** erlon_ has joined #openstack-cinder12:07
*** erlon__ has joined #openstack-cinder12:09
*** sapd1 has quit IRC12:10
*** erlon_ has quit IRC12:13
*** walshh__ has quit IRC12:18
*** zhubx has quit IRC12:19
*** zhubx has joined #openstack-cinder12:19
*** sapd1 has joined #openstack-cinder12:22
*** markvoelker has joined #openstack-cinder12:33
smcginnistmm[m]: Yeah, that doesn't sound right.12:33
smcginnistmm[m]: Hopefully someone comes along that knows how that works better than me, but in the meantime, could you file a bug to capture some of the things you're seeing? https://bugs.launchpad.net/cinder/+bugs12:34
*** psachin has quit IRC13:01
tmm[m]Hmm, the backup script (someone else wrote! :)) didn't use --incremental, adding that got rid of the issue13:01
*** avishay has quit IRC13:01
tmm[m]but it still seems somewhat unexpected and undesirable behavior13:01
*** psachin has joined #openstack-cinder13:04
*** markvoelker has quit IRC13:05
*** rosmaita has joined #openstack-cinder13:07
*** enriquetaso has joined #openstack-cinder13:14
enriquetasojungleboyj, whoami-rajat good morning! o/ Would you mind reviewing this backport to queens https://review.openstack.org/#/c/637941/ ?13:18
*** ianychoi has joined #openstack-cinder13:22
*** mriedem has joined #openstack-cinder13:28
whoami-rajatenriquetaso:  I'm not a stable branch core :)13:39
e0neenriquetaso: approved13:42
*** spsurya has joined #openstack-cinder13:44
enriquetasoThanks e0ne :D13:45
enriquetasothanks anyway whoami-rajat :D13:46
*** jmlowe has quit IRC13:47
openstackgerritMichal Arbet proposed openstack/cinder master: Fix python3 compatibility of rbd get_fsid  https://review.openstack.org/63519513:52
openstackgerritIvan Kolodyazhny proposed openstack/cinder master: Publish backup capabilities to a scheduler  https://review.openstack.org/62088113:53
*** markvoelker has joined #openstack-cinder14:02
*** sdake has joined #openstack-cinder14:10
*** dave-mccowan has joined #openstack-cinder14:15
*** sdake has quit IRC14:18
openstackgerritIvan Kolodyazhny proposed openstack/cinder master: Create backups via scheduler  https://review.openstack.org/63030514:19
jungleboyje0ne:  Thanks.14:21
e0nejungleboyj: you're welcome14:21
*** srini__ has joined #openstack-cinder14:28
*** s10 has joined #openstack-cinder14:28
*** srini_ has quit IRC14:30
*** jmlowe has joined #openstack-cinder14:30
*** eharney has joined #openstack-cinder14:32
*** sdake has joined #openstack-cinder14:35
*** markvoelker has quit IRC14:36
whoami-rajatjungleboyj  smcginnis e0ne Hi, the horizon bp now needs the group_snapshot project_id also, can i make changes in the previously merged "add project_id response to groups" or do we need another mv bump to add this functionality?14:48
whoami-rajathttps://review.openstack.org/#/c/627549/14:48
*** gnufied has joined #openstack-cinder14:48
smcginniswhoami-rajat: My opinion in the past has been that since we have not released that yet we can still change it, but there have been others with differing opinions on that.14:49
*** srini__ has quit IRC14:50
whoami-rajatsmcginnis:  Since these are almost similar functionalities, i thought adding new mv wouldn't be a good approach.14:53
whoami-rajatwould it be better to discuss it in meeting to have everyone's opinion?14:53
*** jgriffith has joined #openstack-cinder14:55
*** jgriffith has quit IRC14:56
e0newhoami-rajat: I agree with smcginnis. less micrioversions is better14:56
*** jgriffith has joined #openstack-cinder14:57
*** sdake has quit IRC14:58
smcginnisMaybe we can get input from the big boss (jungleboyj).14:58
jungleboyjHuh?14:59
smcginnisI would just say, let's make sure we have all related cases covered so we don't need to do it again.14:59
smcginnisWho?14:59
smcginnisWHere?15:00
smcginnisWhy?15:00
smcginnis:P :D15:00
whoami-rajatsmcginnis: yes, i confirmed this is the last case. they've completed the bp just this detail is missing.15:00
smcginniswhoami-rajat: Probably good to bring it up in the meeting. Can you add it to the agenda?15:00
jungleboyjwhoami-rajat:  I agree with smcginnis and e0ne .  Don't think we need another MV bump if we haven't released yet.15:00
jungleboyjIf there is a spec associated we need to update that as well.15:01
whoami-rajatsmcginnis:  should i add it now too? jungleboyj agreed to the same.15:02
whoami-rajatjungleboyj: yes, will update the bp, spec and code respectively.15:02
jungleboyjwhoami-rajat:  ++ Sounds good.15:02
smcginniswhoami-rajat: Up to jungleboyj15:03
whoami-rajatjungleboyj:  should i add this topic to the meeting agenda?15:04
jungleboyjwhoami-rajat:  What additional discussion is required?15:06
whoami-rajatjungleboyj:  Nothing from my side. Sean just asked to confirm. I'm happy with all the agreement. I think i can start with the code now. Thanks jungleboyj  smcginnis  e0ne15:08
*** enriquetaso has quit IRC15:08
*** zhubx has quit IRC15:08
jungleboyjwhoami-rajat:  You can add it to the agenda just to make everyone aware.15:09
smcginnis++, if we're all in agreement, then I agree no need for further discussion in the meeting.15:09
whoami-rajatjungleboyj: sure. thanks.15:11
*** Luzi has quit IRC15:11
geguileojungleboyj: any news on the storyboard front?15:13
jungleboyjgeguileo:  He he.  Well, I was about to add that to the meeting conversation as I think as a team we need to come to an agreement.  Will you be available to talk?15:15
geguileojungleboyj: yes15:15
jungleboyjCool.  Thanks.15:16
geguileojungleboyj: thank you  :-)15:16
openstackgerritGorka Eguileor proposed openstack/os-brick master: Fix FC case sensitive scanning  https://review.openstack.org/63817815:18
openstackgerritEric Harney proposed openstack/python-cinderclient master: Fix bash_completion cache path  https://review.openstack.org/63817915:18
whoami-rajatjungleboyj: i've updated 2 sections of spec tracking, volume re-image and untyped volumes. updated the parts i was working on. Hope it's not a bad practice so informed :)15:21
whoami-rajat https://etherpad.openstack.org/p/cinder-spec-review-tracking15:21
jungleboyjwhoami-rajat:  Ah, Thank you.15:21
eharneyoh, i didn't know you were looking at the untyped volumes stuff, neat15:22
*** sapd1_x has joined #openstack-cinder15:23
eharneyi think i still have to a todo to write a spec for that...15:23
whoami-rajatjungleboyj:  thanks :)15:23
*** sapd1 has quit IRC15:23
eharneywhoami-rajat: do you want to write a spec for that?15:24
whoami-rajateharney:  yes. just pushed the online migrations patch today. was thinking of informing you after some good progress.15:24
whoami-rajateharney: sure :)15:24
*** markvoelker has joined #openstack-cinder15:33
s10Hello. I've found a problem with the version discovery in Cinder. This problem is caused by urllib.parse.urljoin in https://github.com/openstack/cinder/blob/master/cinder/api/views/versions.py#L7215:40
s10This function doesn't concatenate properly, for example, 'https://api.example.com/volume' (no trailing slash) with 'v3'. Result is 'https://api.example.com/v3' instead of 'https://api.example.com/volume/v3'15:40
s10What could be done with this bug?15:40
openstackgerritChuck Short proposed openstack/cinder master: Remove 'rm' from rootwrap filters  https://review.openstack.org/63818415:42
smcginniss10: Looks like someone else noticed that - https://review.openstack.org/#/c/630085/15:44
whoami-rajatsmcginnis: i knew i reviewed similar issue, was searching for it. :)15:45
smcginnisHaha, yep. That one rang a bell right away.15:45
jungleboyjsmcginnis:  Thanks.  That sounded familiar.15:46
smcginnisNow if we can actually get tests to pass.15:46
whoami-rajatsmcginnis: yep. passed the first time now failing continiously.15:49
*** jgriffith has quit IRC15:49
smcginnisI've had a few patches lately that have needed several rechecks to get by various random failures.15:50
eharneywhoami-rajat:  just found another instance of the all_tenants validation causing problems for us: https://bugs.launchpad.net/cinder/+bug/1816792   :/15:52
openstackLaunchpad bug 1816792 in Cinder "revert-to-snapshot fails w/ 400 when snapshot does not exist" [Undecided,New]15:52
whoami-rajatsmcginnis:  yes. specially this one https://review.openstack.org/#/c/635319/ . failing due to backup.createprogress!=volume.retype .15:52
whoami-rajateharney: this is the case with non-admin user right?15:54
eharneywhoami-rajat: yes15:54
whoami-rajateharney: this is really strange. it is treating the snapshot name as a filter.15:56
eharneythis seems to be more evidence that when we added the filtering on the server, it flags things as invalid that it shouldn't15:57
eharneyi assume it's failing because the client decided to add all_tenants=1 in the query15:57
eharneybut this one is messier than the group commands, because that isn't even an argument on this command15:58
smcginnisI thought we decided tests should mock out those notification calls. Getting really sick of seeing failures from that.15:58
eharneysmcginnis: we did decide that -- but i think there was still at least one patch in flight to do it?15:59
s10smcginnis: Ok, thank you. I've filled a bug report https://bugs.launchpad.net/cinder/+bug/181679315:59
openstackLaunchpad bug 1816793 in Cinder "Incorrect concatenation in the version discovery" [Undecided,New]15:59
*** enriquetaso has joined #openstack-cinder15:59
smcginniss10: Thanks. If I find they already filed a bug I will mark it as a duplicate. Otherwise, we can update that commit to close that bug so we have tracking on it.15:59
*** e0ne has quit IRC16:01
*** e0ne has joined #openstack-cinder16:02
*** markvoelker has quit IRC16:06
*** phuongnh has joined #openstack-cinder16:07
whoami-rajateharney:  found the problem https://github.com/openstack/python-cinderclient/blob/master/cinderclient/base.py#L40016:10
eharneywhoami-rajat: i think the problem is that the server shouldn't be throwing an http 400 for that..16:11
whoami-rajateharney: now the question is why revert_to_snapshot is using this flow of functions when all_tenants can't be a part of it's request.16:11
smcginnisAnother reason we should fix that on the server side.16:12
*** phuongnh has quit IRC16:12
eharneywhoami-rajat: why is all_tenants=1 a bad request?  the server side validation is just broken16:12
smcginnis++16:12
eharneyat best it would be a 403 but probably it should just be ignored when it isn't allowed in cases like this so that things work like they used to16:12
whoami-rajateharney:  but is it a good thing to send it in a request where it shouldn't be allowed ? it won't be serving any purpose anyway. Agree on the part that server should be handling this more appropriately. IMO both places needs to be fixed.16:14
eharneywhoami-rajat: we don't know if it's "allowed" or not because we haven't designed and documented how it's supposed to work in the server.  i'm going with "it should work like it did before we accidentally broke it" for now16:14
smcginniseharney: Totally agree. We've broken existing behavior in too many places.16:15
whoami-rajateharney: hmm. makes sense. looks like only few non admins test these commands or don't report it since its been a long time that the filtering change merged.16:20
whoami-rajateharney: any ideas on resolving this issue?16:21
eharneywhoami-rajat: my idea currently is we should go fix the filtering to allow "all_tenants" through instead of rejecting it, and verify that other applicable rules on "all_tenants" re: allowing users to see things they don't own are properly handled at a lower layer16:23
whoami-rajateharney: allowing it meaning adding it to resource_filters file?16:25
eharneywhoami-rajat: not sure16:25
eharneymaybe that's the right way but i'd have to research some more to know if that's the right way to do it16:25
whoami-rajateharney:  allowing a filter for non-admins that shouldn't be allowed to fix things it broke kind of defeats the purpose of the filtering api change. if we could create and handle a generic case at the filtering level rather than allowing it then it would be better IMO (if its feasible).16:31
eharneywhoami-rajat: it only defeats the purpose if you assume that the code we landed for filtering perfectly matches a well-defined purpose.  what was the purpose?  and does the code really match it?16:33
openstackgerritMerged openstack/os-brick master: VMware: Detach backing vmdk during disconnect  https://review.openstack.org/63450816:33
*** s10 has quit IRC16:34
whoami-rajateharney: don't think so. causing more problems than giving benefits currently.16:36
whoami-rajateharney:  agree that breaking things is much worse than keeping a feature intact when it isn't serving the intended purpose properly.16:37
*** larsks has joined #openstack-cinder16:39
larsksWhen we reboot our controller, the iscsi target configuration is lost.  What is supposed to refresh the target configuration when the system reboots? Or is this supposed to be persistent without help?  We're using a tripleo standalone deployment.16:41
eharneylarsks: for the LVM driver, right?16:42
larskseharney: right.16:42
larsks(We're using this w/ ironic to boot baremetal nodes from iSCSI luns)16:43
eharneylarsks: is this w/ containerized cinder-volume?16:43
smcginniseharney: Ironically since you've been the most vocal about using git.o.o links instead of github in case things change with that mirring, the git.o.o are now the ones that will all be broken. :]16:43
larskseharney: Yes.  After a reboot, `docker exec -u root -it cinder_volume targetcli` shows an empty configuration.16:43
larsksWe're running from a recent delorea, so cinder is at commit 5656aa9.16:45
larsks*delorean16:45
openstackgerritMerged openstack/os-brick master: Handle None value 'inititator_target_map'  https://review.openstack.org/53918416:45
eharneylarsks: i have a suspicion that those used to get saved on reboot and restored by a service that used "targetcli save" etc but i'm having trouble finding that at the moment...16:48
eharneywondering if the mechanics of that changed with the container work16:48
larskseharney: when existing targetcli, I see that it says "Configuration saved to /etc/target/saveconfig.json"...but that file is empty right now.16:48
larsksWell, let me re-create a target configuration and see (a) if it gets written out to that file *at all*, and (b) if it survives a container restart.16:50
eharneylarsks: cinder also has some code that calls restore_export in the driver that is supposed to recreate these targets upon startup16:52
*** Florian has joined #openstack-cinder16:56
smcginnisrosmaita, geguileo, jungleboyj: That's kind if why I'm thinking either wait completely, or just bite the bullet and do a full cinder team migration.17:02
smcginnisI think have some here, some there, will just cause confusion and increase the odds we miss something.17:02
rosmaitasmcginnis: i agree with you, we want everything in the same place17:02
jungleboyjsmcginnis:  Ugh.  Ok.17:03
rosmaitawe can split the difference with nova and say we'll move after glance moves17:03
*** markvoelker has joined #openstack-cinder17:03
rosmaita:)17:03
jungleboyjSo we either put cinderlib in launchpad for now and move it in the future or we move everything now?17:03
geguileosmcginnis: ++17:04
jungleboyjrosmaita:  Ha!  That is even less likely.17:04
geguileoeverything in the same place sounds best17:04
rosmaita(that is my hidden agenda)17:04
rosmaitayes, i think put cinderlib into launchpad sounds best17:04
*** e0ne has quit IRC17:04
*** psachin has quit IRC17:04
jungleboyjTo be honest I am not interested in doing this either but I am trying to make it not be a panicked last minute move.17:05
whoami-rajatjungleboyj: won't moving everything together can cause issues (if there are any with automated migrations, i'm not sure). can we confirm with the projects that have migrated before like ironic?17:05
geguileojungleboyj: yeah, doing a migration now sounds too rushed17:05
smcginnisjungleboyj: What's the panic though? There is nothing really compelling us to move to storyboard other than that team would like us to.17:06
openstackgerritMerged openstack/os-brick master: Update hacking version  https://review.openstack.org/62859617:06
jungleboyjwhoami-rajat:  Ironic's migration went fine but they did it quickly and didn't think it through as much as they should have so they aren't really leveraging it well.17:06
rosmaitayeah, i have sympathy for the storyboard team, they need real projects in there so they can make sure it's feature-complete, but i have sympathy for us not wanting to move to a product that isn't completed yet17:06
jungleboyjrosmaita:  They have gotten better at addressing issues.  I think.17:06
geguileojungleboyj: smcginnis I'll go change the cinderlib project patch so it doesn't go into storyboard17:06
jungleboyjgeguileo:  Ok.17:07
jungleboyjI will continue to try and get the design we talked about at the mid-cycle documented and we will discuss it at the PTG in the hopes that we will move in Train.17:07
smcginnisgeguileo: Cool, thanks!17:07
rosmaitajungleboyj: that sounds good, if we can decide exactly what workflows we need in storyboard, it will give us more confidence to migrate17:08
rosmaitabut i think that will entail locking some of us in a room for a few hours :(17:09
jungleboyjI am fine going with the safe approach as long as we are not just going to avoid everything.17:09
geguileosmcginnis: thanks for insisting, otherwise it would have been painfull for all of us  XD17:09
jungleboyjBecause I want us to improve our processes.  Need to leave something for my legacy here.17:09
jungleboyj@!h geguileo17:09
geguileorofl17:10
jungleboyjUgh.17:10
jungleboyj(/ .□.) ︵╰(゜Д゜)╯︵ /(.□. )17:10
* jungleboyj gives up for today.17:10
rosmaitajungleboyj: maybe we can schedule an evening out-of-band session at the forum ... would be good to create some bugs in launchpad and storyboard simultaneously and see where the workflow breaks down/needs to be adjusted17:11
jungleboyjrosmaita: You mean to create them in both locations and go through the workflow for comparison.17:12
rosmaitaexactly, i think that's the only way to catch everything17:12
*** pcaruana has quit IRC17:13
rosmaitai will volunteer to help out since i am dragging my feet on the change17:13
*** enriquetaso has quit IRC17:13
*** carlos_silva has joined #openstack-cinder17:13
*** lseki has joined #openstack-cinder17:14
rosmaitawe can warn the storyboard team in advance so that the dev instance is up to date17:14
rosmaitaor i guess just use production storyboard and then delete the projects we create17:14
whoami-rajatsmcginnis:  eharney  : https://git.openstack.org/cgit/openstack/cinder/tree/doc/source/contributor/api_microversion_dev.rst#n1017:14
whoami-rajatthis doesn't seem true now since it won't break changes introduced BEFORE that feature mv but will break everything after that. seems like a negative point,  do we need to modify it?17:14
geguileojungleboyj: guess what...17:15
geguileojungleboyj: it got merged it storyboard17:15
geguileosmcginnis: rosmaita ^17:15
rosmaitacan we revert?17:15
whoami-rajatgeguileo: lol, all this discussion and this.17:15
jungleboyjgeguileo:  Yeah?17:16
geguileosince you +1 it, they went ahead and merged it17:16
smcginnisShoot, that's right/17:16
* jungleboyj sighs17:16
*** dviroel has joined #openstack-cinder17:16
jungleboyjSorry.17:16
smcginnisMight have to ask in infra if there's a way to switch that over to launchpad now.17:16
rosmaitathey're going to love that :D17:17
jungleboyjrosmaita:  Yeah, they are going to resist that.17:17
*** davidsha has quit IRC17:17
smcginnisWhy?17:17
rosmaitai think we have legitimate technical reasons to keep all cinder stuff in the same bug tracker17:18
jungleboyjBecause of the push to have all new projects in Storyboard.  But, lets ask.17:18
smcginnisLaunchpad is still a valid option. There's no mandate that projects HAVE to go to storyboard.17:18
jungleboyjrosmaita:  Yeah.17:18
smcginnisrosmaita: ++17:18
jungleboyjgeguileo:  Can you approach them and see if they will change it?17:19
potsjungleboyj: Hi Jay, could you take a look at https://review.openstack.org/#/c/631702/1 when you get a chance?17:22
rosmaitageguileo: it is after end of work in your TZ, want me to follow up on this?17:22
geguileojungleboyj: rosmaita I just dumped my question to the people that +2 the patch17:22
rosmaitacool17:23
jungleboyjgeguileo:  Cool.  Thanks.17:23
geguileorosmaita: I may take you up on that offer, depending on whether they answer in a few minutes or not, thanks! :-)17:23
rosmaitaok17:23
jungleboyjrosmaita:  Thanks.17:23
*** ircuser-1 has quit IRC17:26
*** mauricio_ps has joined #openstack-cinder17:27
rosmaitageguileo: do you have time to put up the patch fungi mentioned, or do you want me to do it?17:29
geguileorosmaita: what patch? r:-??17:29
geguileoI missed that17:29
rosmaita(12:24:21 PM) fungi: geguileo: you'll want to update gerrit/projects.yaml to remove the use-storyboard:true line, for a start17:29
rosmaita(12:24:51 PM) fungi: though storyboard doesn't have a way to delete a project as far as i'm aware, so we may need to get creative there17:29
fungigeguileo: just need to remove the use-storyboard:true line your original change added to gerrit/projects.yaml17:29
fungionce that gets merged i can manually delete the project from storyboard's backend db17:30
geguileofungi: sorry, too many notifications at the same time and I missed your comment, sorry17:30
fungino worries! i know exactly what that's like ;)17:31
geguileofungi: and it will create the launchpad automatically when I remove it?17:31
fungiit doesn't create anything in launchpad. what do you want created in launchpad?17:32
fungii guess let's have this discussion in #openstack-infra so we don't keep saying the same thing in both channels ;)17:32
geguileofungi: +117:33
*** enriquetaso has joined #openstack-cinder17:35
*** markvoelker has quit IRC17:36
enriquetasoLooking for Outreachy mentors: As we discussed in the Stein mid-cycle-planning, It will be nice if we have an intern working in the integrations with the OpenStack client.17:38
enriquetaso(- check the things that are supported - give full support to the OpenStack cli - remove the cinder cli - what to do with os-brick). If anybody likes the idea, this could be the moment :P17:38
*** ociuhandu_ has joined #openstack-cinder17:39
whoami-rajatenriquetaso: the context isn't clear to me. is the OSC migration official now?17:40
*** ociuhandu has quit IRC17:42
larskseharney: the configuration definitely gets written to /etc/target/saveconfig.json when its created.17:43
*** ociuhandu_ has quit IRC17:43
enriquetasowhoami-rajat, for what I understood we already migrate almost everything, we need to deprecated cindercli and check for the commands that are not supported in OPS but there are in the cindercli. http://paste.openstack.org/show/745488/17:45
*** ociuhandu has joined #openstack-cinder17:54
*** jgriffith has joined #openstack-cinder17:55
*** ociuhandu has quit IRC17:58
*** mauricio_ps has quit IRC17:59
*** thgcorrea has quit IRC17:59
geguileojungleboyj: smcginnis rosmaita OK, launchpad for cinderlib is ready18:00
*** lseki has quit IRC18:02
rosmaitageguileo: ty!18:04
*** ircuser-1 has joined #openstack-cinder18:04
*** sdake has joined #openstack-cinder18:04
*** jgriffith has quit IRC18:08
whoami-rajatenriquetaso: oh, i missed the outreachy part. i was planning to implement the remaining OSC commands but currently busy with some priority items.18:14
whoami-rajatso will an intern be assigned this task?18:14
*** ociuhandu has joined #openstack-cinder18:16
enriquetaso whoami-rajat, only if I found an mentor first hahaha, the task is free. I will be co-mentoring for the Manila project. It's my first time, so I'm not sure if I could do it for two persons. I have received a few emails from possible interns, but I'm only looking for only one.18:16
*** jgriffith has joined #openstack-cinder18:21
whoami-rajatenriquetaso: sounds like a good way to help new contributors. I'm not exactly familiar with the process of outreachy.18:21
*** enriquetaso has quit IRC18:21
rosmaitawhoami-rajat: https://wiki.openstack.org/wiki/Outreachy18:22
whoami-rajatthanks Brain!18:22
whoami-rajatBrian*18:22
rosmaitanp18:22
openstackgerritMerged openstack/os-brick master: Make sure looping calls are properly mocked  https://review.openstack.org/63563618:25
openstackgerritChuck Short proposed openstack/cinder master: Remove 'rm' from rootwrap filters  https://review.openstack.org/63818418:33
*** markvoelker has joined #openstack-cinder18:33
openstackgerritErlon R. Cruz proposed openstack/cinder-tempest-plugin master: Adds new scenario test for revert to snapshot  https://review.openstack.org/62057718:43
*** e0ne has joined #openstack-cinder18:50
*** sdake has quit IRC18:58
*** markvoelker has quit IRC19:06
*** e0ne has quit IRC19:10
openstackgerritRajat Dhasmana proposed openstack/cinder master: Update guidelines of 'Adding a microversion'  https://review.openstack.org/63824119:12
*** sdake has joined #openstack-cinder19:13
whoami-rajat^ smcginnis jungleboyj  updated the microversion doc, kindly take a look. Thanks!19:13
* eharney takes a look19:15
* jungleboyj opens a tab.19:15
*** jmlowe has quit IRC19:19
whoami-rajateharney: should i include it here [1] before the footnotes ?19:24
whoami-rajat[1] https://docs.openstack.org/cinder/rocky/contributor/api_microversion_dev.html#when-do-i-need-a-new-microversion19:24
eharneywhoami-rajat: yeah i think so19:24
*** jmlowe has joined #openstack-cinder19:30
*** e0ne has joined #openstack-cinder19:30
*** enriquetaso has joined #openstack-cinder19:33
openstackgerritRajat Dhasmana proposed openstack/cinder master: Update section 'When do i need a new Microversion?'  https://review.openstack.org/63824119:40
*** e0ne has quit IRC19:41
openstackgerritRajat Dhasmana proposed openstack/cinder master: Update section 'When do i need a new Microversion?'  https://review.openstack.org/63824119:45
enriquetasowhoami-rajat, sorry i disconnected19:48
*** angela-s has joined #openstack-cinder19:49
whoami-rajatenriquetaso: np!19:50
*** sdake has quit IRC19:50
*** e0ne has joined #openstack-cinder19:51
*** markvoelker has joined #openstack-cinder20:03
*** jmlowe has quit IRC20:06
*** carlos_silva has quit IRC20:08
*** jmlowe has joined #openstack-cinder20:08
openstackgerritSean McGinnis proposed openstack/cinder master: Fix version return incorrect when endpoint url end without /  https://review.openstack.org/63008520:30
*** markvoelker has quit IRC20:36
openstackgerritEric Harney proposed openstack/cinder master: Use native python truncate for privsep  https://review.openstack.org/63826220:40
openstackgerritEric Harney proposed openstack/cinder master: Ignore hacking rule H101  https://review.openstack.org/63826320:40
*** spsurya has quit IRC20:42
*** eharney has quit IRC20:46
*** jmlowe has quit IRC20:53
*** enriquetaso has quit IRC20:58
*** senrique_ has joined #openstack-cinder20:58
*** eharney has joined #openstack-cinder20:59
*** erlon__ has quit IRC21:01
*** senrique_ has quit IRC21:03
openstackgerritMerged openstack/cinder stable/rocky: VNX Driver: delete_hba() instead of remove_hba()  https://review.openstack.org/63621221:04
e0nesmcginnis: Sean, I'm going to block this chain https://review.openstack.org/#/c/635546/ because there is no CI for that driver21:07
smcginnise0ne: It's the VMAX one.21:07
smcginnisThey've rebranded the storage, and the driver, but didn't want to go through the effort of redoing the CI account.21:08
smcginnisWhich I can totally sympathise with.21:08
e0nesmcginnis: here is a first patch with rebranding https://review.openstack.org/#/c/620659/2021:10
e0nesmcginnis: I checked CI logs and cinder conf: http://publiclogs.emc.com/620659/20/EMC_VMAX_iSCSI/763/logs/etc/cinder/cinder.conf.txt.gz21:11
e0nesmcginnis: and grep by old driver name: http://paste.openstack.org/show/745519/21:11
e0nesmcginnis: I don't see running CI with new code21:12
smcginnisIs there an issue?21:12
*** sdake has joined #openstack-cinder21:12
smcginnise0ne: I'm not sure what you mean.21:12
e0neEMC_VMAX_iSCSI CI has this line 'volume_driver = cinder.volume.drivers.dell_emc.powermax.iscsi.PowerMaxISCSIDriver' in the cinder.conf21:13
e0neit's an old driver name21:13
smcginnisThat's the new driver name, but OK.21:13
e0neoh.. I'm sorry21:14
e0nehere is it: volume_driver = cinder.volume.drivers.dell_emc.vmax.iscsi.VMAXISCSIDriver21:14
smcginnisYeah, not sure why they have both in there, but doesn't appear to matter.21:15
e0neCI contains 2 configured backend21:15
smcginnisSorry, I'm not sure I'm following what you're trying to point out.21:15
e0neCI contains two enabled backends: with old and new class name21:17
smcginnisYep, and default is set to the first one. I'm following that so far.21:18
smcginnisI wonder if they are using the two for some replication tests or something.21:18
smcginnisBut anyway...21:19
e0neok, I need to recheck it more carefully21:20
smcginnisSorry, I'm still not sure what needs to be checked. :]21:20
e0nelet's try to do it step by step together21:27
e0nemaybe it's just too late for me and I missed something21:27
e0nehere is a patch https://review.openstack.org/#/c/636118/221:27
smcginnisOK, following.21:28
e0neEMC_VMAX_FCSUCCESS in 7h 20m 18s21:28
e0neEMC_VMAX_iSCSISUCCESS in 7h 21m 33s21:28
smcginnisYep, CI passed.21:28
e0neit is supposed, that these checks verify PowerMax?21:28
e0nehere is cinder conf from EMC_VMAX_iSCSI  logs: http://publiclogs.emc.com/636118/2/EMC_VMAX_iSCSI/915/logs/etc/cinder/cinder.conf.txt.gz21:29
smcginnisYes, like I said and you referred to the specific patch, VMAX was rebranded to PowerMax. They just didn't create a new CI account with the new name. At least not yet.21:29
smcginnisYep, got it.21:30
*** eharney has quit IRC21:30
e0neI don't care about CI name now21:30
e0nevolume_driver = cinder.volume.drivers.dell_emc.vmax.iscsi.VMAXISCSIDriver21:30
*** takamatsu_ has joined #openstack-cinder21:31
e0neit's from cinder.conf on EMC_VMAX_iSCSI for 'PowerMax Driver - Unit Test Refactoring' patch21:31
*** takamatsu has quit IRC21:31
smcginnisYep, I see it.21:31
angela-sSUCCESS in 7h 20m 18s !!! CI run took 7 hours, good grief...21:31
smcginnisHeh, slow one. They must have a concurrency of 1 or something.21:31
*** markvoelker has joined #openstack-cinder21:33
e0neI'm confused21:33
smcginnise0ne: :)21:33
smcginnisI think I've followed so far, so which part?21:33
e0nehow does this CI work if it has .dell_emc.vmax.iscsi.VMAXISCSIDriver in the config?21:34
smcginnisBecause it's the same driver. They just rebranded it.21:35
smcginnisAnd when we rebrand drivers, as part of the "deprecation" transition from the old name to the new name we do this: https://review.openstack.org/#/c/620659/20/cinder/volume/manager.py21:35
smcginnisSo I would guess they are sticking with the old name for now to keep testing on stable branches simpler.21:36
smcginnisBut they will eventually need to switch their cinder.conf over to using the new driver name onces that transition period is over and that mapping is removed.21:36
e0nesmcginnis: oh.. thank you! I missed that part in cinder.volume.manager21:37
smcginnisOK, good. That makes sense then.21:38
*** sdake has quit IRC21:38
smcginnisThat's what we've always done when a driver needs to change its name or is refactored into a different path.21:38
e0nesmcginnis: my bad, absolutely forgot about it:(21:39
e0neit means that I need to stop code review for today and take a rest21:39
smcginnisNo worries! :)21:39
smcginnisYes, it's getting late there. Stop working so late. ;)21:39
*** sdake has joined #openstack-cinder21:40
*** takamatsu_ has quit IRC21:45
*** takamatsu has joined #openstack-cinder21:46
*** sdake has quit IRC21:48
jidaris multipathd required to do volumes from images?  And is multipathd not found in the kolla based cinder images?21:50
*** sdake has joined #openstack-cinder21:50
smcginnisjidar: Not sure about kolla requirements, but no, multipathing isn't needed for anything.21:55
smcginnisjidar: It's recommended to us in general, but nothing requires it.21:55
jidarsmcginnis: actually I'm thinking about the problem a little bit more and the container wouldn't have access to /dev/sdb (it's trying to launch multipath -l /dev/sdb)21:56
jidarand failing obviously21:56
jidarn/m, it maps all of /dev over21:57
* jidar is confused21:57
*** sdake has quit IRC21:59
*** dave-mccowan has quit IRC22:01
*** enriquetaso has joined #openstack-cinder22:01
*** sdake has joined #openstack-cinder22:01
angela-ssmcginnis: can you take a look at this bugfix when you have a chance? https://review.openstack.org/#/c/637297/ ty22:03
*** markvoelker has quit IRC22:07
smcginnisangela-s: Done22:07
*** enriquetaso has quit IRC22:09
angela-sty!22:11
*** mriedem has quit IRC22:25
*** rcernin has joined #openstack-cinder22:27
*** whoami-rajat has quit IRC22:31
*** e0ne has quit IRC22:32
*** lemko has quit IRC22:41
*** jmlowe has joined #openstack-cinder22:47
*** tkajinam has joined #openstack-cinder22:55
*** markvoelker has joined #openstack-cinder23:03
*** rcernin has quit IRC23:06
*** rcernin has joined #openstack-cinder23:08
*** imacdonn has quit IRC23:32
*** imacdonn has joined #openstack-cinder23:33
*** markvoelker has quit IRC23:35
*** rosmaita has left #openstack-cinder23:57
*** mriedem has joined #openstack-cinder23:57
« Tuesday, 2019-02-19 Index Thursday, 2019-02-21 »

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