Wednesday, 2019-03-13

*** enriquetaso has joined #openstack-cinder00:13
*** enriquetaso has quit IRC00:20
openstackgerritMerged openstack/cinder master: Use excutils.save_and_reraise_exception instead of reraise
*** markvoelker has joined #openstack-cinder00:43
*** whoami-rajat has joined #openstack-cinder00:51
*** zhubx has quit IRC00:58
*** zhubx has joined #openstack-cinder00:58
*** tkajinam_ has joined #openstack-cinder00:58
*** tkajinam has quit IRC01:01
*** markvoelker has quit IRC01:21
*** Nel1x has joined #openstack-cinder01:25
*** dviroel_ has quit IRC01:47
*** rosmaita has left #openstack-cinder02:11
*** adriant has joined #openstack-cinder02:35
adriantOut of curiosity, is there any reason why Gunicorn wouldn't work as a server for Cinder?02:37
adriantI've added "application = initialize_application()" to the wsgi file, and was trying to run it as: "gunicorn cinder.wsgi.wsgi:application" in a docker container, but it just doesn't work.02:38
adriantI get "gunicorn: error: unrecognized arguments: cinder.wsgi.wsgi:application" because it seems gunicorn can't quite use the application variable Cinder makes.02:38
adriantuWSGI works fine, but I can't for the life of my understand why it would work, but Gunicorn won't.02:39
adriantbut even for uWSGI we have to add: "application = initialize_application()" so I am curious why that isn't there in the first place02:45
adriantor potentially should I be doing it via:02:59
*** psachin has joined #openstack-cinder03:03
*** Nel1x has quit IRC03:05
whoami-rajatadriant: could you provide a error traceback?03:13
adriantwhoami-rajat: When running the gunicorn command I didn't get any errors other than that, and my colleague who was trying to get it running ran into much the same thing03:16
adriantpotentially we configured the log output of Cinder wrong, but we didnt' see anything. it mostly just failed to start with gunicorn as the server03:17
adriantand as I look at it, that gunicorn paste option doesn't actually at all seem related to paste
*** itlinux has quit IRC03:41
*** udesale has joined #openstack-cinder04:17
whoami-rajatadriant: sorry i went off. could you paste your wsgi file. i might be wrong but application should be a function right? 'application=initialize_application'04:23
adriantwhoami-rajat: a WSGI application object is a callable, but it expects 2 arguments04:30
adriantinitialize_application() isn't valid04:30
adriantand for some reason the object it returns does work for Gunicorn04:30
adriantwhoami-rajat: and as for my wsgi file, literally all I did is append: "application = initialize_application()" to the one in cinder04:31
adriantso that it calls that function to save it's result as a variable I can give gunicorn04:31
*** zhubx has quit IRC04:34
whoami-rajatadriant:  yeah it's returning the object to my point shouldn't matter.04:34
adriantoh typo.. "and for some reason the object it returns *doesn't* work for Gunicorn"04:34
*** zhubx has joined #openstack-cinder04:35
adriantIt's just weird that the same exactly code and installed versions work for UWSGI with a uwsgi file, but not gunicorn04:35
*** zhubx has quit IRC04:36
*** zhubx has joined #openstack-cinder04:36
adriantIf I'm just doing something wrong I'm happy because it's easy to fix, but if there is a bug that probably should be looked at.04:37
*** zhubx has quit IRC04:37
adriantI don't even mind using uWSGI, I like prefer Gunicorn and was hoping to maintain a standard :P04:38
*** zhubx has joined #openstack-cinder04:38
adriantand it's just weird it doesn't work04:38
adriantChances are Cinder isn't the only service with the issue, I just happened to start with Cinder.'04:39
whoami-rajatadriant:  are the 2 arguments mandatory? i haven't really tried deploying it with any other server since uWSGI is default and works well. i don't think cinder has some specific settings for uWSGI and should work for other servers as well.04:39
adriantwhoami-rajat: the wsgi interface expects to be called with "(environ, start_response)"04:41
adriantso the value that initialize_application returns is what uWSGI or Gunicorn would be using as the app callable04:42
*** zhubx has quit IRC04:44
*** zhubx has joined #openstack-cinder04:45
adriantanyway, I need to run, but my IRC bouncer will still keep me logged in. whoami-rajat: if you have any ideas of advice I'd be happy to test things out tomorrow04:46
whoami-rajatadriant: hmm. so currently what we can do is report a bug against it, i can query for the same in cinder meeting tonight if anyone has tried the same.04:46
adriantIf this is entirely PEBKAC I'm fine with that, I just would like the option of using gunicorn if I can for consistency :)04:46
whoami-rajatadriant: also if you're willing to try it for other modules, would be great to see if this is cinder specific or on openstack level.04:47
whoami-rajatadriant: Thanks for mentioning this scenario.04:47
adriantAs I have time I'll try a few others. I know Keystone recently switched to using Flask so I think they are safe04:48
adriantbut I'll have a look at Glance api soon as well04:48
adriantit's possible the old paste + oslo.wsgi service stuff is a little weird04:49
whoami-rajatadriant: did they? wow, i thought Flask wasn't much suited for production env.04:49
adriantwhoami-rajat: really? I know quite a lot of projects outside of openstack using Flask04:50
adriantwe tend to use it for a lot of microservices, but I know some people building reasonably bigger apps with it.04:50
whoami-rajatadriant: yeah i read it in a few articles, maybe things have changed recently in newer versions.04:52
*** udesale has quit IRC05:16
*** anks2k has joined #openstack-cinder05:17
*** udesale has joined #openstack-cinder05:17
*** anks2k has quit IRC05:17
*** zhubx has quit IRC05:43
*** zhubx has joined #openstack-cinder05:44
*** markvoelker has joined #openstack-cinder05:47
*** udesale has quit IRC05:51
*** udesale has joined #openstack-cinder06:11
*** pcaruana has joined #openstack-cinder06:19
*** jdillaman has quit IRC06:30
*** avishay has joined #openstack-cinder06:33
*** jdillaman has joined #openstack-cinder06:35
*** e0ne has joined #openstack-cinder06:40
*** Luzi has joined #openstack-cinder06:48
*** tonyb is now known as tonyb_gone07:20
*** tonyb_gone is now known as tonyb07:21
*** dpawlik has joined #openstack-cinder07:23
*** e0ne has quit IRC07:31
*** gkadam has joined #openstack-cinder07:48
*** e0ne has joined #openstack-cinder07:51
*** dpawlik has quit IRC07:54
*** sahid has joined #openstack-cinder08:05
*** lennyb has joined #openstack-cinder08:05
*** dpawlik has joined #openstack-cinder08:06
*** rcernin has quit IRC08:09
*** zhubx has quit IRC08:17
*** zhubx has joined #openstack-cinder08:17
*** Liang__ has joined #openstack-cinder08:21
*** zhubx has quit IRC08:21
*** Liang__ is now known as LiangFang08:22
*** zhubx has joined #openstack-cinder08:23
*** zhubx has quit IRC08:23
*** zhubx has joined #openstack-cinder08:23
*** zhubx has quit IRC08:25
*** zhubx has joined #openstack-cinder08:26
avishaysahid: Hey, do you have a few minutes to discuss the reverted WWN verification commit?08:35
*** e0ne has quit IRC08:36
sahidavishay: sure08:45
sahidbtw i saw an other revert related to the rename of scaleio, i'm wondering whether that have already been discussed? because that looks tricky since we already release a version with that change in-tree08:48
avishayI don't know about scaleio08:53
avishaysahid: When I try the WWN verification on my setup it works almost always. I have trouble reproducing the case where no WWN is found. I was thinking that maybe it happens when some device is leftover from a previous attachment?08:54
sahidavishay: you mean with the first patch that you have landed?08:55
*** mszwed has quit IRC09:03
whoami-rajatsahid:  we discussed the revert change, the renaming isn't merged in cinder and nova so changes in os-brick seems unrequired currently. also we don't need to rebrand os-brick due to a driver renaming, we can just use the same protocol 'scaleio' and same connection properties since os-brick changes doesn't reflect to the end user.09:13
avishaysahid: Yes, the first patch (the one that was reverted)09:14
avishayI am running a setup with multipath, and it works fine (usually)09:15
*** kukacz has quit IRC09:15
avishaySometimes it fails to find the WWN but I have no idea why09:15
sahidavishay: are you using multipah in your setup?09:15
avishaysahid: Yes09:16
sahidwhoami-rajat: ack09:16
*** kukacz has joined #openstack-cinder09:17
sahidavishay: so my thinking it's that it's a timeing issue. the fix you have proposed is not right since at some point the devices are not registered anymore in /dev/disk/by-id/09:18
sahidand that method you are useing in your patch is using this path09:18
avishaysahid: There is in fact only one device in /dev/disk/by-id/ that points to the multipath device, but the link is called /dev/disk/by-id/scsi-<wwid> so it still works.09:21
avishayThere is a comment there in get_sysfs_wwn() about multiple designators, not sure what that means09:21
*** tkajinam__ has joined #openstack-cinder09:22
*** FlorianFa has joined #openstack-cinder09:23
*** tkajinam_ has quit IRC09:25
sahidavishay: hum my thinkin is, the wwid which will be returned by get_sysfs_wwn() will be for example based on sda, so then the condition fails since it should not have a ref like /dev/disk/by-id/scsi-<sda-wwid>, then in the second condition we go through all the devices [sda, sdb, ...] and check if there are in /dev/disk/by-id/ but they are not anymore09:32
avishaysahid: But in the correct case, the devices for all paths will have the same wwid, so the symlink /dev/disk/by-id/scsi-<wwid> is valid for all of them.09:33
sahidavishay: have you try to log the return of get_sysfs_wwid?, i just checked in my env and basically i don't have /sys/block/sdb/device/wwid09:41
sahidavishay: basically you have the mapth so what about to use get_scsi_wwn(mpath) when use_multipath is configured to do your check09:43
avishaysahid: Yes, I thought to use get_scsi_wwn for all cases, instead of get_sysfs_wwn.  I don't know if there are any downsides to it though.09:44
sahidavishay: i have no idea, basically when your patch is ready i could apply it in my env to double check, it's using a purestorage array09:45
avishaysahid: Mine is pure storage as well :)09:46
sahidahaha :) ok fair-enough so09:47
sahidavishay: btw i would be happy if you can have a look at this at some point
avishaysahid: I was thinking though, in case the CI failures were really from leftover devices, maybe the best thing isn't to fail the operation? I had another patch that tried to clean up those devices at the start of the connect but it had issues as well :(09:48
avishaywill review your patch09:49
*** zhubx has quit IRC10:19
*** zhubx has joined #openstack-cinder10:20
*** avishay has quit IRC10:21
*** avishay has joined #openstack-cinder10:24
*** carlos_silva has joined #openstack-cinder10:38
*** luizbag has joined #openstack-cinder10:39
*** dave-mccowan has joined #openstack-cinder10:40
*** dave-mccowan has quit IRC10:45
*** bal has joined #openstack-cinder11:09
*** avishay has quit IRC11:20
*** gkadam_ has joined #openstack-cinder11:24
*** gkadam has quit IRC11:25
*** mriedem has joined #openstack-cinder11:26
*** Alon_KS has joined #openstack-cinder11:29
*** thgcorrea has joined #openstack-cinder11:30
*** avishay has joined #openstack-cinder11:30
*** e0ne has joined #openstack-cinder11:36
*** avishay has quit IRC11:37
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Enable publish-to-pypi Zuul job
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Add DevStack plugin
*** avishay has joined #openstack-cinder11:39
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Fix RBD connector
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Support setting attach_mode
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Fix error_deleting status in vol and snap
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Fix Volume snapshots and connections on deletion
*** luizbag_ has joined #openstack-cinder11:54
*** thgcorrea has quit IRC11:55
*** luizbag has quit IRC11:55
*** zhubx has quit IRC12:07
*** markvoelker has quit IRC12:12
*** markvoelker has joined #openstack-cinder12:13
*** abishop has joined #openstack-cinder12:25
*** rosmaita has joined #openstack-cinder12:32
*** mchlumsky has joined #openstack-cinder12:38
*** lseki has joined #openstack-cinder12:44
*** lseki has quit IRC12:45
*** lseki has joined #openstack-cinder12:45
openstackgerritMatt Riedemann proposed openstack/os-brick stable/rocky: 'iscsiadm -m session' failure handling
*** thgcorrea has joined #openstack-cinder13:05
*** e0ne has quit IRC13:05
*** udesale has quit IRC13:16
*** avishay has quit IRC13:16
*** udesale has joined #openstack-cinder13:17
*** yury-kulaz has joined #openstack-cinder13:18
*** raghavendrat has joined #openstack-cinder13:33
*** eharney has joined #openstack-cinder13:37
*** dviroel_ has joined #openstack-cinder13:40
raghavendrathi Eric. would like to discuss about your comment for
raghavendrateharney: do you have few minutes13:40
eharneyraghavendrat: hi, sure13:42
*** thgcorrea has quit IRC13:43
raghavendratIn 3par, there is no way to quickly clone volume13:43
*** thgcorrea has joined #openstack-cinder13:43
*** zhubx has joined #openstack-cinder13:43
raghavendratalso there was discussion between Rajat Dhasmana and Walter Boring13:44
raghavendratwe thought that it would be better to change default value of extra spec to False13:45
eharneywhich means that by default, some 3par volumes will fail to delete, right?13:45
raghavendratRajat set convert_to_base =False as default, tested cinder rally job, specifically create_volume_from_snaphot, and it passed13:47
*** zhubx has quit IRC13:48
whoami-rajat^ snapshots were dependent previously and didn't do much harm (only rally jobs failed) which I tested and passing with default=False now, so hemna_ and I agreed to restore it to previously behaviour.13:48
raghavendratso i guess, there have been some enhancement in rally job which takes care of volume deletion. And now deletion should work fine.13:48
eharneyi'm not sure what rally tests in regard to this, but it is true that this setting will cause some volumes to fail to delete, right?13:49
whoami-rajateharney: snapshots were  created dependent initially, due to a rally job fail the functionality was modified to have independent snaps. Thought to mention it just for info.13:52
raghavendrateharney: sorry for my limited knowledge. may i know test scenarios for which volume delete may fail; i can perform scenarios and validate13:57
eharneyraghavendrat: clone a volume from a snapshot and then try to delete the original source volume13:58
raghavendrati.e set value of extra-spec is false; and v2 is created as child of s1; now try to delete original volume v1.14:00
raghavendratis my understanding correct?14:00
raghavendratok. i will try this and share my finding.14:01
*** gkadam_ has quit IRC14:02
*** thgcorrea has quit IRC14:03
*** luizbag has joined #openstack-cinder14:03
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Fix Volume snapshots and connections on deletion
*** luizbag_ has quit IRC14:05
*** zhubx has joined #openstack-cinder14:07
*** gnufied has joined #openstack-cinder14:07
*** gnufied has quit IRC14:08
*** irclogbot_0 has quit IRC14:09
*** thgcorrea has joined #openstack-cinder14:10
*** e0ne has joined #openstack-cinder14:10
*** irclogbot_0 has joined #openstack-cinder14:12
*** enriquetaso has joined #openstack-cinder14:20
*** eharney has quit IRC14:23
*** zhubx has quit IRC14:24
*** irclogbot_0 has quit IRC14:25
*** eharney has joined #openstack-cinder14:28
*** irclogbot_0 has joined #openstack-cinder14:28
*** efried has joined #openstack-cinder14:29
efriedjungleboyj, whoami-rajat, eharney: I just caught up on (re scaleio rename revert)14:30
efriedI agree that, if the revert is going to happen, it needs to happen right now, before the release, so we never release with the "bad" version14:30
efriedwe also need to blacklist the "bad" release of os-brick14:31
efriedotherwise somebody down the road could pick it up in their project and use it14:31
efriedand be broken when they upgrade.14:31
efriedI believe this would constitute a patch in the requirements project. I don't see one in place currently. I can propose that if you agree it's the right thing to do.14:32
*** gnufied has joined #openstack-cinder14:34
whoami-rajatefried:  ++ smcginnis  ^^14:34
*** thgcorrea has quit IRC14:34
*** luizbag_ has joined #openstack-cinder14:34
*** luizbag has quit IRC14:35
*** mauricio_ps_ has joined #openstack-cinder14:41
*** bal_ has joined #openstack-cinder14:43
*** mauricio_ps_ has quit IRC14:43
openstackgerritGorka Eguileor proposed openstack/cinder master: Fix VolumeAttachment is not bound to a Session
jungleboyjefried:  Yes, I agree that we should mark the releases of os-brick with that code in it as bad so that people don't use it or get into a bad state.14:47
jungleboyjefried:  Thanks for bringing that up.14:47
efriedjungleboyj: Okay. I'm talking in -requirements to figure out how to do that mechanically.14:47
jungleboyjefried:  Thanks.  I will come watch over your shoulder.14:48
*** yury-kulaz has quit IRC14:49
jungleboyjhemna_: ^^^14:50
*** yury-kulaz has joined #openstack-cinder14:54
openstackgerritsahid proposed openstack/cinderlib master: add python 3.7 unit test job
*** mszwed has joined #openstack-cinder15:01
*** psachin has quit IRC15:07
*** mriedem has quit IRC15:10
efriedjungleboyj: you have the power to +W right? Are you waiting for someone else to weigh in?15:11
efriedoh, I'm also looking with pretty narrow goggles; are there other patches in flight that will need to be part of the release?15:11
*** mriedem has joined #openstack-cinder15:11
jungleboyjefried:  Well, I was hoping to dot i's and cross t's in the Cinder meeting.15:12
jungleboyjhemna_:  Doesn't appear to be around though.15:12
efriedokay, nbd, just lmk if I can help with anything.15:13
jungleboyjThe original author has +1'd so I guess I can +W.15:13
efriedwhoami-rajat: ack15:13
*** _alastor_ has quit IRC15:13
jungleboyjwhoami-rajat:  Any concerns?15:13
*** _alastor_ has joined #openstack-cinder15:14
*** lpetrut has joined #openstack-cinder15:14
whoami-rajatjungleboyj: None from my side on this. my concerns are still with RSD one.15:14
*** Luzi has quit IRC15:14
jungleboyjwhoami-rajat:  Ok, so no reason to not get the ScaleIO ones merged and talk about RSD during the meeting?15:15
whoami-rajatjungleboyj: since Yury agreed on keeping the same protocol and connection properties as previous(ScaleIO), i think we don't need the VXFlexOs patch in os-brick anymore.15:19
whoami-rajatjungleboyj:  that seems like the right way.15:22
*** martinkennelly has joined #openstack-cinder15:22
*** dpawlik has quit IRC15:22
jungleboyjRight.  I am merging those now.  So, just trying to figure out what, if anything we have left to do with RSD.15:23
jungleboyjwhoami-rajat:  I just talked to Walt and he will be in the meeting.15:24
*** _alastor_ has quit IRC15:24
*** _alastor_ has joined #openstack-cinder15:24
whoami-rajatjungleboyj:  just not sure why are we keeping the nvmeof name in the connector list when a connector uses nvme.15:26
whoami-rajatjungleboyj: ok we can discuss it there then, Thanks :)15:26
*** erlon has joined #openstack-cinder15:29
jungleboyjYeah, lets talk about it there since hemna_  Will be around then.15:32
*** zhubx has joined #openstack-cinder15:32
*** irclogbot_0 has quit IRC15:36
*** irclogbot_0 has joined #openstack-cinder15:39
*** enriquetaso has quit IRC15:41
hemna_what'd I miss?15:47
*** eharney has quit IRC15:49
*** irclogbot_0 has quit IRC15:49
*** henriqueof has quit IRC15:49
*** irclogbot_0 has joined #openstack-cinder15:51
*** irclogbot_0 has quit IRC15:52
jungleboyjhemna_:  Howdy sir.  I have merged your reverts for the ScaleIO change.  Want to make sure we are all on the same page as to if there is anything else we need to do for os-brick.15:53
*** irclogbot_0 has joined #openstack-cinder15:56
hemna_ok sounds good15:58
hemna_I have a work meeting at 9am15:59
hemna_I'll try and multitask for the cinder-meeting15:59
*** erlon has quit IRC16:11
openstackgerritYury Kulazhenkov proposed openstack/os-brick master: Fix ScaleIO KeyError after upgrade
yury-kulazhi. This ^^ patch is replacement for
*** yury-kulaz has quit IRC16:21
*** thgcorrea has joined #openstack-cinder16:25
*** yury-kulaz has joined #openstack-cinder16:25
*** yury-kulaz has quit IRC16:25
*** FlorianFa has quit IRC16:41
*** udesale has quit IRC16:42
*** dave-mccowan has joined #openstack-cinder16:43
*** lpetrut has quit IRC16:44
*** eandersson_ is now known as eandersson16:54
*** FlorianFa has joined #openstack-cinder16:54
*** jgriffith has quit IRC16:59
*** e0ne has quit IRC17:01
hemna_how much time do we have to release cinderlib?17:04
hemna_bugs can be backported17:04
*** lpetrut has joined #openstack-cinder17:04
hemna_I think we need to get the pypi release job in place17:04
hemna_so we can push a release to pypi on time17:04
hemna_that's a good one to get in IMHO17:05
jungleboyjhemna_:  Yeah, that would be good to get in and get the pypi push set up.  Then we can deal with the bugs.17:10
openstackgerritMerged openstack/os-brick master: Revert "Fix VxFlexOs KeyError after upgrade"
mriedemnon-client lib release freeze was 2 weeks ago17:16
mriedemis enharney out this week?17:16
mriedemjbernard: is there anything you know about this?
openstackLaunchpad bug 1819944 in OpenStack Compute (nova) "nova-grenade-live-migration job failing on ubuntu bionic" [Undecided,Confirmed]17:16
openstackgerritMerged openstack/os-brick master: Revert "rename ScaleIO connector to VxFlex OS"
mriedemi guess that's not really a red hat thing though...17:17
jungleboyjgeguileo:  Did you do a release of cinderlib17:18
geguileojungleboyj: I had done a couple of them on PyPi17:18
geguileojungleboyj: and when we added cinderlib to openstack I added admin privileges over the pypi account to the openstack ci17:19
geguileojungleboyj: and I haven't done any releases since then17:19
mriedemso it would go through openstack/releases and smcginnis17:19
mriedemand that ship sailed 2 weeks ago17:19
* jungleboyj was just looking through openstack/releases.17:20
jungleboyjYeah.  UGH.17:20
mriedemif only there were a person per project that was responsible for being on top of that stuff ... o)17:23
mriedemthere we go17:23
jungleboyjmriedem:  We might have one of those but people seem to have been spread thin lately.17:24
mriedemthin to win17:25
jungleboyjgeguileo:  So, it appears that we need to get cinderlib added to releases for Train.  What is the impact of not having it in Stein?17:25
*** dave-mccowan has quit IRC17:27
*** lpetrut has quit IRC17:30
*** sahid has quit IRC17:36
*** eharney has joined #openstack-cinder17:41
whoami-rajateharney:  I've a query regarding
eharneywhoami-rajat: ok17:44
whoami-rajateharney:  our gate is failing cinder-tempest-dsvm-lvm-lio-barbican job in bionic with the following error17:45
* smcginnis swings by17:45
smcginnisjungleboyj, geguileo: We could probably do an "independent" release of cinderlib for now, then get it as officially part of train.17:45
jungleboyjsmcginnis: Ok.  Is there an official process for that?17:46
geguileosmcginnis: sounds like a good idea17:46
eharneywhoami-rajat: hmm, interesting17:46
smcginnisjungleboyj: Look under the deliverables/_independent directory in openstack/releases17:47
whoami-rajateharney:  since the legacy jobs migration is soon, our gates at cinder and os-brick may start failing due to this :(17:47
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Enable publish-to-pypi Zuul job
openstackgerritGorka Eguileor proposed openstack/cinderlib master: Add DevStack plugin
geguileojungleboyj: smcginnis maybe like this?
geguileobut I don't know if that triggers the documentation, branch creation, etc17:49
eharneywhoami-rajat: pretty strange, ubuntu packages just failing to install correctly...17:49
whoami-rajateharney: yeah, looked like random failure so rechecked it.17:50
eharneywhoami-rajat: it has a pretty clear error message saying it can't use /etc/target17:51
jungleboyjsmcginnis: and geguileo  I am good with doing an independent release for cinderlib to start with and then get it into the official releases for Train.17:52
geguileosmcginnis: jungleboyj or I could try to add it to the releases project17:52
geguileosmcginnis: jungleboyj I'll write a patch now and see what they say  XD17:53
jungleboyjgeguileo:  I am guessing that that will be rejected by the release team given that we have passed the freeze.17:53
eharneywhoami-rajat: my bet is the packages aren't in great shape, and we can probably work around it by adding a mkdir to the devstack script before that install is performed, but i don't have a bionic environment handy to test this with...17:53
geguileojungleboyj: but it's only the feature freeze, right?17:53
geguileojungleboyj: shouldn't it be fine until RC1?17:54
jungleboyjThe problem is the final release for the non-client libraries.17:54
jungleboyjThat was Mar 01.17:54
geguileojungleboyj: yeah, but cinderlib is not one of those libraries17:54
*** ianychoi has quit IRC17:55
geguileojungleboyj: that freeze is for libraries that are used by other projects17:55
geguileojungleboyj: but cinderlib is a library that uses cinder as a library17:55
geguileojungleboyj: so it should follow cinder's cycle, imo17:55
geguileojungleboyj: I'll try to sell it to them that way  :-)17:55
jungleboyjOMG my head is going to explode today.17:55
jungleboyjOk.  Give it a shot and see what happens.17:56
openstackgerritJon Bernard proposed openstack/cinder master: RBD: save and restore multiattach features
jbernardmriedem: that's new to me18:00
*** martinkennelly has quit IRC18:02
mriedemjbernard: yeah nvm it's an issue in the nova scripts for our live migration job, which use the devstack-plugin-ceph functions but there is something else that's busted,18:03
mriedemso we're just going to disable ceph in these jobs temporarily18:03
jbernardmriedem: ok, if you do find the plugin has a flaw let me know18:04
openstackgerritKeith Berger proposed openstack/os-brick stable/pike: Handle multiple errors in multipath -l parsing
*** bal has quit IRC18:07
geguileojungleboyj: I checked the docs, I have to add it to that repo like smcginnis said, as an independent deliverable, and then on the next cycle add it as a normal cycle18:08
geguileoso I'll propose it as independent18:08
hemna_since cinderlib is new and nobody in openstack is really using it, what's the harm in releasing it?18:17
whoami-rajateharney: wow, worked! Thanks a lot... proposed devstack fix for it
eharneywhoami-rajat: awesome, that saves me a bunch of time :)18:23
*** pcaruana has quit IRC18:29
*** enriquetaso has joined #openstack-cinder18:34
jungleboyjgeguileo:  Ok.  Sorry it can't go in to the main release path yet.18:41
geguileojungleboyj: It shouldn't be that big of a deal  :-)18:42
geguileojungleboyj: it'll be good enough if we can release it  XD18:42
jungleboyjOk.  How will the difference impact releasing it?18:43
geguileonot sure yet18:44
geguileoI don't even know how to do a normal release...18:44
geguileoI'll have a look at the docs if the accept the patch18:44
openstackgerritGhanshyam Mann proposed openstack/cinder master: DNM: Testing cinder legacy jobs on bionic
openstackgerritGhanshyam Mann proposed openstack/os-brick master: DNM: Testing os-brick legacy jobs on bionic
*** gmann is now known as gmann_afk18:48
jungleboyjhemna_: whoami-rajat
jungleboyjCan we get some more eyes on that so we can get it merged if appropriate.18:52
whoami-rajatjungleboyj: was waiting for CI, +2ed18:53
jungleboyjhemna_:  Thanks.  When that merges I will push up the patch for the release.18:56
*** e0ne has joined #openstack-cinder19:01
*** mchlumsky_ has joined #openstack-cinder19:03
*** mchlumsky has quit IRC19:04
*** zzzeek has quit IRC19:04
*** zzzeek has joined #openstack-cinder19:07
*** enriquetaso has quit IRC19:13
*** enriquetaso has joined #openstack-cinder19:29
*** luizbag_ has quit IRC19:30
*** luizbag has joined #openstack-cinder19:30
*** bal has joined #openstack-cinder19:30
*** thgcorrea has quit IRC19:31
*** bal_ has quit IRC19:31
*** bal has quit IRC20:02
*** dklyle has quit IRC20:09
*** luizbag has quit IRC20:12
efriedjungleboyj: ready to release os-brick yet?20:14
*** dklyle has joined #openstack-cinder20:16
enriquetasojungleboyj, o/ How are u? would you mind check this backport ? it has one +2! :)20:16
jungleboyjefried:  waiting for one more patch to land.20:19
jungleboyjShould happen yet tonight.20:19
efriedjungleboyj: ack. lmk if you want me to propose and do the ffe request20:19
jungleboyjenriquetaso: I will take a look.20:19
efriedjungleboyj: lmk either way, cause I'll need to update that blacklist patch20:19
jungleboyjefried: I got it. Thanks.20:19
jungleboyjOk. Will do.20:20
*** s10 has joined #openstack-cinder20:24
*** lpetrut has joined #openstack-cinder20:27
*** lpetrut has quit IRC20:31
*** e0ne has quit IRC20:38
*** gmann_afk is now known as gmann20:41
*** e0ne has joined #openstack-cinder20:41
*** e0ne has quit IRC20:45
openstackgerritMerged openstack/cinder stable/ocata: ScaleIO Driver - adding cache and refactoring tests
enriquetasothanks jungleboyj !20:52
*** enriquetaso has quit IRC20:52
openstackgerritMerged openstack/cinder-tempest-plugin master: Add metadata to snapshot for revert test
*** abishop has quit IRC21:07
*** whoami-rajat has quit IRC21:11
*** rcernin has joined #openstack-cinder21:57
*** s10 has quit IRC21:59
*** carlos_silva has quit IRC22:07
*** mriedem is now known as mriedem_afk22:18
*** lifeless has quit IRC22:27
*** lifeless has joined #openstack-cinder22:27
*** adriant has quit IRC22:29
gmannjungleboyj: for cinder gate failure due to bionic migration, this fix is working fine. once this is merged cinder gate will be up -
*** ianychoi has joined #openstack-cinder23:06
*** sapd1 has joined #openstack-cinder23:09
*** dave-mccowan has joined #openstack-cinder23:28
*** sapd1 has quit IRC23:30
*** lseki has quit IRC23:31
*** dave-mccowan has quit IRC23:38
openstackgerritKeith Berger proposed openstack/os-brick stable/pike: Handle multiple errors in multipath -l parsing
*** erlon has joined #openstack-cinder23:48

Generated by 2.15.3 by Marius Gedminas - find it at!