Tuesday, 2015-02-24

*** marcusvrn has quit IRC00:08
*** bswartz has joined #openstack-cinder00:10
*** Apoorva_ has quit IRC00:17
*** Apoorva has joined #openstack-cinder00:17
*** tbarron has joined #openstack-cinder00:21
*** tbarron has joined #openstack-cinder00:22
*** tsekiyam_ has joined #openstack-cinder00:24
*** Apoorva_ has joined #openstack-cinder00:25
*** asmith_brcd has quit IRC00:26
hemnainteresting read00:26
*** ebalduf has joined #openstack-cinder00:28
*** tsekiyama has quit IRC00:28
*** Apoorva has quit IRC00:28
*** marcusvrn has joined #openstack-cinder00:32
*** ebalduf has quit IRC00:32
*** ho has joined #openstack-cinder00:38
*** annashen has quit IRC00:39
*** Apoorva_ has quit IRC00:40
*** erlon has quit IRC00:41
*** Apoorva has joined #openstack-cinder00:41
*** mtanino has joined #openstack-cinder00:42
*** asselin_ has joined #openstack-cinder00:43
*** asselin_ has quit IRC00:48
*** dannywil_ has quit IRC00:57
*** IanGovett has quit IRC00:57
*** rmesta has quit IRC01:03
*** drjones has quit IRC01:06
*** adam_g is now known as adam_g_out01:09
*** sigmavirus24 is now known as sigmavirus24_awa01:10
*** devlaps has quit IRC01:11
*** hemna is now known as hemnafk01:17
*** david-lyle is now known as david-lyle_afk01:19
*** garthb_ has quit IRC01:21
*** garthb has quit IRC01:21
*** ebalduf has joined #openstack-cinder01:28
*** ebalduf has quit IRC01:33
*** bill_az has quit IRC01:36
*** Mandell_ has quit IRC01:36
*** Mandell has joined #openstack-cinder01:43
*** _cjones_ has joined #openstack-cinder01:43
*** krtaylor has quit IRC01:43
*** jamielennox is now known as jamielennox|away01:46
*** Mandell has quit IRC01:46
vilobhmm1thingee :ping01:46
*** asselin_ has joined #openstack-cinder01:48
*** tsekiyam_ has quit IRC01:48
*** asselin_ has quit IRC01:49
*** thingee has quit IRC01:51
*** krtaylor has joined #openstack-cinder01:55
anishsmcginnis: xyang was the final consensus the same as http://paste.openstack.org/show/180819/ ?01:58
*** xyang has quit IRC01:58
openstackgerritxing-yang proposed openstack/cinder: Kilo Consistency Group API update  https://review.openstack.org/14456101:59
*** Thelo has joined #openstack-cinder02:00
*** leeantho has quit IRC02:05
*** jcru has joined #openstack-cinder02:12
*** jamielennox|away is now known as jamielennox02:12
*** xyang has joined #openstack-cinder02:17
*** lcurtis has joined #openstack-cinder02:17
*** markvoelker has quit IRC02:25
*** markvoelker has joined #openstack-cinder02:26
*** ebalduf has joined #openstack-cinder02:29
*** markvoelker has quit IRC02:30
*** ebalduf has quit IRC02:34
*** vilobhmm1 has quit IRC02:36
*** casusbel_ has joined #openstack-cinder02:37
*** casusbelli has quit IRC02:40
*** kaisers1 has joined #openstack-cinder02:45
*** xyang has quit IRC02:45
openstackgerritxing-yang proposed openstack/cinder: Create Consistency Group from CG Snapshot API  https://review.openstack.org/14595202:46
*** kaisers has quit IRC02:46
*** diegows has quit IRC02:47
*** kaufer has joined #openstack-cinder02:50
openstackgerritJay Bryant proposed openstack/cinder: Sync scheduler module from oslo-incubator  https://review.openstack.org/15853402:53
openstackgerritJay Bryant proposed openstack/cinder: Sync scheduler.filters module from oslo-incubator  https://review.openstack.org/15853502:55
openstackgerritPatrick East proposed openstack/cinder: Add CHAP support to PureISCSIDriver  https://review.openstack.org/13717503:06
openstackgerritPatrick East proposed openstack/cinder: Add DB table for driver specific data  https://review.openstack.org/15183703:06
*** xyang has joined #openstack-cinder03:07
*** markvoelker has joined #openstack-cinder03:08
*** Apoorva_ has joined #openstack-cinder03:08
*** kaufer has quit IRC03:10
*** Apoorva has quit IRC03:11
*** MasterPiece has quit IRC03:11
openstackgerritPatrick East proposed openstack/cinder: Add support to PureISCSIDriver for updating consistency groups  https://review.openstack.org/15763703:12
*** lcurtis has quit IRC03:12
*** Apoorva_ has quit IRC03:13
*** Thelo has quit IRC03:14
*** _cjones_ has quit IRC03:14
*** marcusvrn has quit IRC03:15
*** lcurtis has joined #openstack-cinder03:25
*** patrickeast has quit IRC03:27
*** ebalduf has joined #openstack-cinder03:30
*** ebalduf has quit IRC03:34
*** bkopilov has quit IRC03:39
*** tbarron has quit IRC03:51
*** tbarron has joined #openstack-cinder03:52
*** kaisers1 has quit IRC03:52
*** dannywilson has joined #openstack-cinder03:53
*** dannywilson has quit IRC03:57
*** tbarron has quit IRC03:58
*** vilobhmm1 has joined #openstack-cinder04:04
*** pdeore_ has joined #openstack-cinder04:09
*** Apoorva has joined #openstack-cinder04:10
*** annashen has joined #openstack-cinder04:12
*** lcurtis has quit IRC04:13
*** harlowja is now known as harlowja_away04:14
*** harlowja_away is now known as harlowja04:15
*** devlaps has joined #openstack-cinder04:17
*** lcurtis has joined #openstack-cinder04:25
*** Tross has quit IRC04:27
*** annashen has quit IRC04:28
*** ebalduf has joined #openstack-cinder04:31
*** ebalduf has quit IRC04:35
*** markvoelker has quit IRC04:39
*** markvoelker has joined #openstack-cinder04:39
*** jcru has quit IRC04:43
*** markvoelker has quit IRC04:44
*** lpetrut has joined #openstack-cinder04:46
*** lcurtis has quit IRC04:49
*** krtaylor has quit IRC04:55
*** annashen has joined #openstack-cinder04:57
*** annashen has quit IRC04:58
*** annashen has joined #openstack-cinder05:00
*** bkopilov has joined #openstack-cinder05:01
*** krtaylor has joined #openstack-cinder05:07
*** annashen has quit IRC05:10
*** markvoelker has joined #openstack-cinder05:10
*** casusbel_ has quit IRC05:11
*** annashen has joined #openstack-cinder05:12
*** krtaylor has quit IRC05:14
*** devlaps has quit IRC05:17
*** markvoelker has quit IRC05:17
*** annashen has quit IRC05:22
*** Mandell has joined #openstack-cinder05:26
*** krtaylor has joined #openstack-cinder05:26
*** BharatK has joined #openstack-cinder05:30
*** ebalduf has joined #openstack-cinder05:31
*** jamielennox is now known as jamielennox|away05:33
*** jamielennox|away is now known as jamielennox05:34
*** sgotliv has quit IRC05:34
*** ebalduf has quit IRC05:36
*** lpetrut has quit IRC05:43
*** deepakcs has joined #openstack-cinder05:48
openstackgerritxing-yang proposed openstack/cinder: Modify Consistency Group API  https://review.openstack.org/14456105:56
*** xyang has quit IRC05:56
*** coolsvap_ is now known as coolsvap05:57
*** atmark1 is now known as atmark06:04
*** atmark has quit IRC06:04
*** atmark has joined #openstack-cinder06:04
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Transifex  https://review.openstack.org/15240106:06
*** ho__ has joined #openstack-cinder06:09
*** ho has quit IRC06:09
*** harlowja is now known as harlowja_away06:11
*** markvoelker has joined #openstack-cinder06:13
*** _cjones_ has joined #openstack-cinder06:14
*** markvoelker has quit IRC06:17
*** annashen has joined #openstack-cinder06:20
*** _cjones_ has quit IRC06:20
*** anshul has joined #openstack-cinder06:29
*** thingee has joined #openstack-cinder06:31
*** ebalduf has joined #openstack-cinder06:32
*** anshul has quit IRC06:36
*** anshul has joined #openstack-cinder06:36
thingeewanghao: please push an update for this https://review.openstack.org/#/c/144590/06:36
*** ebalduf has quit IRC06:37
*** annashen has quit IRC06:37
thingeewanghao: deadline is approach for passing code on march 1st http://lists.openstack.org/pipermail/openstack-dev/2015-February/056964.html06:37
*** thingee has quit IRC06:38
*** Mandell has quit IRC06:39
*** BharatK has quit IRC06:40
*** aniket has joined #openstack-cinder06:40
*** BharatK has joined #openstack-cinder06:41
*** pradipta has joined #openstack-cinder06:42
anikethello all :) , I am working on cinder component. I just want to know that how can we boot an instance from SAN storage?06:45
*** rushiagr_away is now known as rushiagr06:49
*** Apoorva has quit IRC07:00
openstackgerritEric Brown proposed openstack/cinder: Use oslo_config choices support  https://review.openstack.org/15858007:03
aniketanyone have any idea about this?07:09
*** nkrinner has joined #openstack-cinder07:15
openstackgerritEric Brown proposed openstack/cinder: Password config options should be marked secret  https://review.openstack.org/15859007:17
*** xyang1 has quit IRC07:19
*** sgotliv has joined #openstack-cinder07:20
*** sgotliv has quit IRC07:26
*** Miouge has joined #openstack-cinder07:27
*** ebalduf has joined #openstack-cinder07:33
*** chlong has quit IRC07:34
*** annashen has joined #openstack-cinder07:35
*** TobiasE has joined #openstack-cinder07:35
*** kaisers has joined #openstack-cinder07:36
*** coolsvap is now known as coolsvap_07:36
*** ebalduf has quit IRC07:37
*** vilobhmm1 has quit IRC07:38
*** coolsvap_ is now known as coolsvap07:39
*** annashen has quit IRC07:41
*** rushiagr is now known as rushiagr_away07:42
*** sgotliv has joined #openstack-cinder07:42
*** annashen has joined #openstack-cinder07:46
*** e0ne has joined #openstack-cinder07:50
*** nshaikh has joined #openstack-cinder07:50
*** vilobhmm has joined #openstack-cinder07:51
*** vilobhmm has quit IRC07:53
*** dulek has joined #openstack-cinder07:59
*** annashen has quit IRC08:00
*** aviramb has joined #openstack-cinder08:00
*** lpetrut has joined #openstack-cinder08:07
*** ndipanov has joined #openstack-cinder08:08
openstackgerritLi Yingjun proposed openstack/python-cinderclient: Add missing all-tenants option to transfer-list  https://review.openstack.org/11417608:12
openstackgerritEric Brown proposed openstack/cinder: Use oslo_config choices support  https://review.openstack.org/15858008:22
openstackgerritEric Brown proposed openstack/cinder: Use oslo_config choices support  https://review.openstack.org/15858008:28
*** ronis has joined #openstack-cinder08:29
*** emaisne has joined #openstack-cinder08:29
*** ho__ has quit IRC08:31
*** ho has joined #openstack-cinder08:31
*** ebalduf has joined #openstack-cinder08:34
*** teju has joined #openstack-cinder08:37
*** ebalduf has quit IRC08:38
*** enterprisedc has quit IRC08:39
openstackgerritTobias Engelbert proposed openstack/cinder: Recall mechanism for terminate/initialize  https://review.openstack.org/15437708:44
openstackgerritTobias Engelbert proposed openstack/python-cinderclient: Recall mechanism for terminate/initialize  https://review.openstack.org/15437608:45
*** jistr has joined #openstack-cinder08:54
*** pradipta has quit IRC08:59
*** dalgaaf has joined #openstack-cinder08:59
*** jordanP has joined #openstack-cinder09:00
*** TobiasE has quit IRC09:03
*** vukcrni has quit IRC09:04
*** teju has quit IRC09:05
*** aviramb has quit IRC09:05
*** aniket has quit IRC09:05
*** btran has quit IRC09:05
*** nikesh_vedams has quit IRC09:05
*** nuritv has quit IRC09:05
*** ankit_ag has joined #openstack-cinder09:09
*** yuriy_n17 has joined #openstack-cinder09:13
*** markvoelker has joined #openstack-cinder09:17
*** karimb has joined #openstack-cinder09:21
*** markvoelker has quit IRC09:22
*** karimb has quit IRC09:23
*** TobiasE has joined #openstack-cinder09:23
*** aix has quit IRC09:25
*** karimb has joined #openstack-cinder09:27
*** ndipanov has quit IRC09:30
*** ndipanov has joined #openstack-cinder09:30
dulekwinston-d: Hi, I'm curious about your cleaner fix, can you share the idea?09:34
*** ebalduf has joined #openstack-cinder09:34
dulekwinston-d: I'm talking about https://launchpad.net/bugs/1409012 of course09:35
openstackLaunchpad bug 1409012 in Cinder "Volume becomes in 'error' state after scheduler starts" [High,In progress] - Assigned to Michal Dulko (michal-dulko-f)09:35
*** openstackgerrit has quit IRC09:36
*** openstackgerrit has joined #openstack-cinder09:36
*** ebalduf has quit IRC09:39
*** nellysmitt has joined #openstack-cinder09:44
*** aix has joined #openstack-cinder09:52
*** takedakn has joined #openstack-cinder09:59
*** annashen has joined #openstack-cinder10:00
openstackgerritRonen Mesonzhnik proposed openstack/cinder: Provide backup service that uses ProtecTIER as the backend  https://review.openstack.org/15396910:01
*** harlowja_away has quit IRC10:04
*** alecv has joined #openstack-cinder10:05
*** annashen has quit IRC10:05
*** TobiasE1 has joined #openstack-cinder10:15
*** TobiasE has quit IRC10:16
*** markvoelker has joined #openstack-cinder10:18
openstackgerritYuriy Nesenenko proposed openstack/cinder: Handle incorrect '--config-dir' param  https://review.openstack.org/15774910:20
*** markvoelker has quit IRC10:23
openstackgerritHuang Zhiteng proposed openstack/cinder: Allow scheduler to receive volume stats when starting service  https://review.openstack.org/15862310:23
*** marcusvrn has joined #openstack-cinder10:28
smcginnisanish: Sorry, was away until now and just saw your question.10:29
smcginnisanish: Yes, logging is correct in http://paste.openstack.org/show/180819/10:30
smcginnisanish: As long as it's passed in to the logger, the logger will take care of it.10:30
smcginnisanish: It's only an issue if you try to format it first and just pass in the string.10:30
*** aix has quit IRC10:30
smcginnisanish: BAD: LOG.info(_LI('Error: %s') % e), GOOD: LOG.info(_LI('Error: %s'), e)10:31
smcginnisanish: Subtle. :)10:31
smcginnisanish: More reason to follow the convention of using , instead of % and letting the logger handle things.10:31
*** ndipanov has quit IRC10:32
*** marcusvrn1 has joined #openstack-cinder10:32
*** tbarron has joined #openstack-cinder10:33
smcginnisanish: Oh, wait.10:33
smcginnisanish: Looking at that paste again, debug is wrong there. No difference if it is debug or info or whatever.10:33
smcginnisanish: Point is to pass the exception in to the logger. Then no need to explicitly call six.text_type().10:34
*** ebalduf has joined #openstack-cinder10:35
*** marcusvrn has quit IRC10:35
*** ebalduf has quit IRC10:40
*** jamielennox is now known as jamielennox|away10:42
*** aix has joined #openstack-cinder10:43
*** ndipanov has joined #openstack-cinder10:46
*** ho has quit IRC10:49
openstackgerritnikeshmahalka proposed openstack/cinder: Add CHAP persistence to SCST target helper  https://review.openstack.org/15735610:49
*** takedakn has quit IRC10:57
*** jamielennox|away is now known as jamielennox10:58
*** jamielennox is now known as jamielennox|away11:01
*** jamielennox|away is now known as jamielennox11:02
*** jamielennox is now known as jamielennox|away11:08
*** sgotliv has quit IRC11:08
*** jamielennox|away is now known as jamielennox11:10
*** markvoelker has joined #openstack-cinder11:19
*** sgotliv has joined #openstack-cinder11:20
*** markvoelker has quit IRC11:24
*** takedakn has joined #openstack-cinder11:31
*** takedakn has quit IRC11:32
*** takedakn has joined #openstack-cinder11:32
*** takedakn has quit IRC11:36
*** ebalduf has joined #openstack-cinder11:36
*** jamielennox is now known as jamielennox|away11:38
*** diegows has joined #openstack-cinder11:39
*** takedakn has joined #openstack-cinder11:40
*** takedakn has quit IRC11:40
*** ebalduf has quit IRC11:40
*** jamielennox|away is now known as jamielennox11:41
*** sgotliv has quit IRC11:51
*** sgotliv has joined #openstack-cinder11:51
*** tbarron has quit IRC11:52
*** enterprisedc has joined #openstack-cinder11:56
*** chlong has joined #openstack-cinder11:57
*** annashen has joined #openstack-cinder12:01
*** markvoelker has joined #openstack-cinder12:05
*** annashen has quit IRC12:06
*** markvoelker has quit IRC12:11
*** IanGovett has joined #openstack-cinder12:27
*** ebalduf has joined #openstack-cinder12:37
*** sgotliv has quit IRC12:37
*** ebalduf has quit IRC12:41
*** chlong has quit IRC12:44
*** akerr has joined #openstack-cinder12:45
*** Miouge has quit IRC12:53
*** Miouge has joined #openstack-cinder12:54
*** Mandell has joined #openstack-cinder12:58
*** Mandell has quit IRC13:03
*** bkopilov has quit IRC13:04
*** Tross has joined #openstack-cinder13:05
*** TobiasE1 has quit IRC13:06
*** TobiasE has joined #openstack-cinder13:06
*** markvoelker has joined #openstack-cinder13:06
*** Yogi1 has joined #openstack-cinder13:07
*** nikesh_vedams has joined #openstack-cinder13:08
*** markvoelker has quit IRC13:11
*** Longgeek has joined #openstack-cinder13:12
*** jamielennox is now known as jamielennox|away13:18
*** jamielennox|away is now known as jamielennox13:20
*** sgotliv has joined #openstack-cinder13:27
*** ankit_ag has quit IRC13:29
*** deepakcs has quit IRC13:30
*** sgotliv has quit IRC13:31
*** BharatK has quit IRC13:31
*** rcallawa has joined #openstack-cinder13:34
*** sgotliv has joined #openstack-cinder13:34
*** jaypipes has quit IRC13:36
*** ebalduf has joined #openstack-cinder13:38
*** xyang has joined #openstack-cinder13:40
*** jseiler has joined #openstack-cinder13:40
*** xyang1 has joined #openstack-cinder13:40
*** ebalduf has quit IRC13:42
*** sgotliv has quit IRC13:43
*** rushiagr_away is now known as rushiagr13:43
*** sgotliv has joined #openstack-cinder13:43
*** Yogi1 has quit IRC13:46
*** anshul has quit IRC13:51
*** ozamiatin has joined #openstack-cinder13:54
*** thangp has joined #openstack-cinder13:55
*** nlevinki has joined #openstack-cinder13:55
*** emagana has joined #openstack-cinder13:56
*** Miouge_ has joined #openstack-cinder13:57
*** Miouge has quit IRC13:58
*** Miouge_ is now known as Miouge13:58
*** annashen has joined #openstack-cinder14:02
dulekthangp: Hi!14:03
dulekthangp: I have some questions regarding objects patches. You've commented that we should pass objects instead of resource id's in RPC calls.14:04
thangpdulek: sure14:04
dulekthangp: I wonder if this doesn't introduce races.14:05
thangpdulek: why would it?14:05
thangpdulek: the purpose of the spec was to pass objects over rpc14:05
dulekthangp: As we're passing an object state it can change when Rabbit is doing it's processing.14:05
dulekAnd we're not refreshing the state.14:05
thangpdulek: you could implement the check in the save() function to check14:06
*** Miouge has quit IRC14:06
*** Miouge has joined #openstack-cinder14:06
thangpto make sure that the object is in the correct state14:06
*** kaufer has joined #openstack-cinder14:06
dulekthangp: I'm not opposed to the idea, just thinking of possible problems.14:06
*** Yogi1 has joined #openstack-cinder14:06
thangpbut the existing implementation might have locks in place to stop this14:06
*** markvoelker has joined #openstack-cinder14:07
*** annashen has quit IRC14:07
dulekthangp: By existing implementation you mean current master?14:07
thangpe.g. snapshots has a lock decorator to stop multiple deletes IIRC14:07
thangpdulek: existing implementation in the current cinder code base (not using objects)14:07
thangpdulek: some of them do14:08
thangpdulek: i havent looked at backups, but it may have them14:08
dulekSo locks would be the answer here.14:08
thangpdulek: in part14:08
thangpdulek: for scenarios where races are apparent, we need to check the state before we actually save14:08
thangpdulek: and fail is the "state" is old14:09
dulekthangp: Yeah, to not mess the DB.14:09
thangpdulek: or figure out how to handle it correctly14:09
thangpdulek: so...you could add an option (e.g. expected_state) in the save() function14:10
thangpdulek: and when you do save, you could check if the backup state is what you expected14:10
dulekthangp: Actually I think we may just check if state we have in the object is the same as in the DB.14:10
thangpit causes a lookup, but you do it when you know there can be a race14:10
dulekthangp: If I understand it correctly.14:10
thangpcorrect, that's what i mean14:11
thangpwe check the object state vs. the db state14:11
thangpand compare14:11
*** rmesta has joined #openstack-cinder14:11
dulekI cannot see locks in the backups but checks should do it.14:12
dulekOkay, great. I'm reworking the backups patch now.14:12
thangpdulek: yeah...you could do it either way14:13
thangplocks or checks in save()14:13
thangpwhichever one gives you better performance :)14:13
thangpor both14:13
*** changbl has quit IRC14:13
dulekthangp: Would you mind if I start the discussion on this topic in cinder-meeting tomorrow? Having core team opinion on this will be beneficial.14:14
thangpsure go for it14:14
dulekthangp: ok, another news - there will be no oslo.versionedobjects release this week14:15
thangpdulek: yeah...i didnt expect them to release anything anytime soon14:15
thangpdulek: and any release might be potentially buggy14:15
thangpuntil everything is ironed out14:15
*** rmstar has quit IRC14:16
dulekthangp: The release isn't probably much more than you did with forking the Nova code14:16
dulekthangp: So I don't expect it won't be so much difference14:17
thangpdulek: there were a few more things that got removed/moved/renamed14:17
*** navid__ has joined #openstack-cinder14:17
thangpdulek: but it should be ok14:17
dulekthangp: anyway - I wanted to start to rewrite the base patch for current state of oslo.versionedobjects14:17
dulekthangp: For now - only locally, no commits ;)14:18
thangpdulek: thanks for helping me on it!14:18
*** rmstar has joined #openstack-cinder14:18
thangpdulek: refactoring the unit tests was a pain :P14:18
dulekthangp: I saw the changelog, it definitely was ;)14:18
thangpdulek: good luck, let me know if you have any questions14:19
dulekthangp: so if you have no objections I'll start rewriting the base patch and see if there will be any problems14:19
thangpbase patch?14:19
thangpthe nova objects base?14:20
dulekthangp: Cinder objects base I mean.14:20
thangpthe expected_state stuff?14:20
*** dustins has joined #openstack-cinder14:21
dulekI meant oslo.versionedobjects. I plan to use current state of their repo so transition would be painless once they release.14:21
thangpfor your backup objects or for everything?14:21
dulekI was thinking of everything.14:22
dulekOr do you prefer another approach - creating _new_ patch that depends on Backup's patch?14:22
thangperr...i really dont know when versionedobjects would be released...and the current patches we have up are ready14:22
thangpso rebasing everything to oslo.versionedobjects might delay us getting anything into kilo14:23
thangpif oslo.versionedobjects is releases really close to the deadline14:23
thangpnot opposed to using oslo.versionobjects...but we need some runway to test it out14:24
dulekthangp: true, that's why I don't want to commit it to gerrit - just to have patch ready14:24
thangpah ok14:24
thangpwe can still look at it14:24
dulekthangp: maybe it'll make into K, maybe in L14:24
dulekthangp: but to have it ready would be cool I think14:25
thangpmaybe have another patch after your backup patch reworking it to oslo.versionedobjects?14:25
thangpis that what you mean?14:25
dulekthangp: That's another approach. Now as you shared your concerns about getting objects into Kilo I think this will be better approach.14:25
thangpdulek: ok cool, so just to confirm...you will create a child patch after your backup objects one that would rebase things to oslo.versionedobjects?14:27
dulekthangp: yes, this is how I want to proceed14:27
thangpdulek: i good with that, thx14:27
dulekthangp: don't worry, I wasn't planning to touch your patches without permission ;)14:28
thangpdulek: thx14:28
*** nellysmitt has quit IRC14:28
dulekthangp: okay, so one last thing - I have one more developer on the team who can work on Cinder objectification14:28
*** nshaikh has left #openstack-cinder14:29
thangpdulek: sure14:29
*** bkopilov has joined #openstack-cinder14:29
*** nellysmitt has joined #openstack-cinder14:29
*** navid__ has left #openstack-cinder14:29
*** nellysmitt has quit IRC14:29
*** bill_az has joined #openstack-cinder14:29
*** timcl has joined #openstack-cinder14:29
*** nellysmitt has joined #openstack-cinder14:29
dulekthangp: I wonder if it would be better if he starts to create a new object (Service perhaps) or try to replace rest of Snapshots DB accesses with objects.14:29
dulekthangp: (in a dependent patch of course)14:30
thangpdulek: anywhere...i guess14:30
thangpdulek: service is probably the easiest14:30
dulekSo right now you're not planning to do the rest of Snapshots?14:31
thangpdulek: i'm trying to figure out the snapshots get_all thing14:31
thangpdulek: then planning on the rest of volume14:31
*** pdeore_ has quit IRC14:32
thangpdulek: there are CG groups, service, ...14:32
*** markvoelker has quit IRC14:32
thangpthat need to switch to objects14:32
*** markvoelker has joined #openstack-cinder14:33
dulekthangp: Okay, I understand that you advise to go on with a new object.14:33
*** mtanino has quit IRC14:34
thangpdulek: it depends on where he/she wants to work on - service, volume, volume type, snapshot, quota, backup, consistency group and consistency group snapshot objects14:34
thangpaside from the ones we already have...service, volume type, quota, and consistency groups are left14:35
dulekthangp: I need to step out for a moment, will message you a bit later14:36
thangpok...and then we have to switch the rest of the code to use objects, i.e. drivers and such14:36
thangpto use the dot notation14:36
*** markvoelker has quit IRC14:38
*** ebalduf has joined #openstack-cinder14:38
*** ebalduf has quit IRC14:42
*** mriedem has joined #openstack-cinder14:47
openstackgerritHuang Zhiteng proposed openstack/cinder: Allow scheduler to receive volume stats when starting service  https://review.openstack.org/15862314:49
openstackgerritPetrut Lucian proposed openstack/cinder: SMBFS: Fix retrieving total allocated size  https://review.openstack.org/15871314:49
openstackgerritHuang Zhiteng proposed openstack/cinder: Allow scheduler to receive volume stats when starting service  https://review.openstack.org/15862314:51
openstackgerritPetrut Lucian proposed openstack/cinder: SMBFS: Fix retrieving total allocated size  https://review.openstack.org/15871314:51
*** nellysmitt has quit IRC14:55
*** Mandell has joined #openstack-cinder14:55
*** emagana has quit IRC15:04
*** emagana has joined #openstack-cinder15:05
*** alexpilotti has quit IRC15:07
*** Mandell has quit IRC15:09
dulekthangp: Sorry for that, I've had ad-hoc meeting.15:09
*** jungleboyj has joined #openstack-cinder15:10
thangpdulek: np15:10
*** emagana has quit IRC15:10
dulekthangp: I would prefer to complete work on Snapshots first but if you're already working on that there's no need to overlap.15:10
thangpdulek: ok15:10
dulekthangp: So it's your call I guess if you need any help there?15:11
thangpdulek: we can start on the service object15:11
thangpthat should be simple to implement15:11
*** amoturi has quit IRC15:11
dulekthangp: OK, understood. :)15:11
*** amoturi has joined #openstack-cinder15:11
thangpdulek: ok cool, thx15:11
*** Manshoon has joined #openstack-cinder15:12
dulekthangp: Thanks for syncing up!15:12
thangpsure np15:12
*** nlevinki has quit IRC15:13
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements  https://review.openstack.org/15764215:14
*** sigmavirus24_awa is now known as sigmavirus2415:16
bswartzthingee: awake yet?15:16
*** Yogi1 has quit IRC15:17
bswartzI have a question for anyone who was in austin last month15:17
*** nellysmitt has joined #openstack-cinder15:18
jungleboyjbswartz: Yes sirt?15:18
bswartzwhen we were talking about API versions and extensions, someone mentioned a way to query for the list of enabled extensions15:18
bswartzdoes anyone remember that and remember how to do it?15:18
jungleboyjbswartz: Are you talking about querying Keystone?15:19
smcginnisbswartz: I remember but do not recall how. I thingee brought that up.15:19
smcginnis*think thingee15:19
bswartzno I'm referring to all the different APIs implemented in contrib as extensions15:19
*** Yogi1 has joined #openstack-cinder15:19
bswartzeveryone agreed it was a bad idea, but I didn't hear a solution proposed15:20
jungleboyjbswartz: Oh, yeah, I vaguely remember that but don't remember how.15:20
jungleboyjbswartz: Think that was when I was trying to deal with the migration mess.15:20
smcginnisIf only someone would have taken detailed notes.15:20
jungleboyjsmcginnis: :-p15:21
bswartzjay takes excellent notes15:21
smcginnisbswartz: Totally agree! Way better than I ever do.15:21
jungleboyjI ahve to do that or I don't pay attention in long meetings like that.15:22
jungleboyjOr IRC meetings.15:22
smcginnisjungleboyj: Good strategy. I wander off too often and have to remind myself to pay attention. :)15:23
*** mtanino has joined #openstack-cinder15:23
openstackgerritJay Bryant proposed openstack/cinder: Sync scheduler module from oslo-incubator  https://review.openstack.org/15853415:23
*** devlaps has joined #openstack-cinder15:24
*** devlaps has quit IRC15:24
*** deepakcs has joined #openstack-cinder15:25
bswartzin IRC meetings the bot takes notes for you15:26
akerrbswartz: sounds like a bad "in communist russia..." joke15:26
smcginnisIn Soviet Russia the notes take you? :D15:26
bswartzI only have problems in teleconferences and face to face meetings remembering what was said15:27
smcginnisakerr: :)15:27
akerrsmcginnis: +115:27
*** primechuck has joined #openstack-cinder15:31
*** nellysmitt has quit IRC15:32
*** emaisne_ has joined #openstack-cinder15:32
*** nellysmitt has joined #openstack-cinder15:32
openstackgerritSean McGinnis proposed openstack/cinder: Remove use of contextlib.nested  https://review.openstack.org/15873215:34
*** emaisne has quit IRC15:36
openstackgerritEric Brown proposed openstack/cinder: Password config options should be marked secret  https://review.openstack.org/15859015:36
*** emaisne_ has quit IRC15:37
*** ebalduf has joined #openstack-cinder15:39
*** ebalduf has quit IRC15:44
*** aviramb has joined #openstack-cinder15:45
*** david-lyle_afk is now known as david-lyle15:47
*** markvoelker has joined #openstack-cinder15:49
*** rushiagr is now known as rushiagr_away15:50
*** lpetrut has quit IRC15:54
*** BharatK has joined #openstack-cinder15:56
*** vilobhmm has joined #openstack-cinder15:57
*** annegent_ has joined #openstack-cinder15:58
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: Add volume multi attach support  https://review.openstack.org/8584715:59
*** amoturi1 has joined #openstack-cinder16:01
*** coolsvap is now known as coolsvap_16:02
openstackgerritMikhail Khodos proposed openstack/cinder: Nexenta Edge iSCSI backend driver.  https://review.openstack.org/15874516:03
*** annashen has joined #openstack-cinder16:03
*** amoturi1 has quit IRC16:04
*** amoturi1 has joined #openstack-cinder16:04
*** Mandell has joined #openstack-cinder16:06
*** vilobhmm has quit IRC16:06
*** lpetrut has joined #openstack-cinder16:07
openstackgerritHuang Zhiteng proposed openstack/cinder: Allow scheduler to receive volume stats when starting service  https://review.openstack.org/15862316:08
*** annashen has quit IRC16:08
*** hodos has joined #openstack-cinder16:08
*** btran has joined #openstack-cinder16:10
*** lcurtis has joined #openstack-cinder16:10
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Allow archiving deleted rows to shadow tables, for performance  https://review.openstack.org/13118216:12
*** emagana has joined #openstack-cinder16:12
*** mdenny has joined #openstack-cinder16:13
*** jamielennox is now known as jamielennox|away16:15
*** jamielennox|away is now known as jamielennox16:16
*** changbl has joined #openstack-cinder16:16
*** ronis has quit IRC16:18
*** rhe00 has quit IRC16:18
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Allow archiving deleted rows to shadow tables, for performance  https://review.openstack.org/13118216:21
*** dannywilson has joined #openstack-cinder16:23
*** tbarron has joined #openstack-cinder16:23
*** tsekiyama has joined #openstack-cinder16:25
*** tsekiyama has quit IRC16:26
*** tsekiyama has joined #openstack-cinder16:26
*** Mandell has quit IRC16:27
*** dencaval has joined #openstack-cinder16:28
*** jdurgin1 has joined #openstack-cinder16:28
*** TobiasE has quit IRC16:29
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: FCZM fix reading of cinder config entries  https://review.openstack.org/15877616:34
*** yatin has joined #openstack-cinder16:39
*** deepakcs has quit IRC16:40
*** ebalduf has joined #openstack-cinder16:40
*** _cjones_ has joined #openstack-cinder16:40
*** _cjones_ has quit IRC16:41
yatinhow to change owner in git review16:41
*** _cjones_ has joined #openstack-cinder16:42
*** alecv has quit IRC16:43
dulekyatin: It's impossible as far as I know16:43
yatindulek: then how can i submit a patch to a already opened review16:44
*** ebalduf has quit IRC16:44
dulekyatin: You just amend your changes and do git review.16:45
yatindulek: I did this but the owner remained the same as before16:46
*** _cjones_ has quit IRC16:46
*** jcru has joined #openstack-cinder16:46
*** _cjones_ has joined #openstack-cinder16:46
*** jcru has quit IRC16:47
dulekyatin: So I don't know how to help you. Downloading the patch, amending my changes and doing git review always worked for me.16:47
*** lpetrut has quit IRC16:48
yatindulek: i checked out master, did changes, added changeID(of already submitted patch) in commit message and submitted for review16:51
smcginnisyatin: I don't believe that's right.16:52
smcginnisYou would have had to check out their review, not master.16:52
smcginnisThen just update and ammend it.16:53
smcginnisyatin: I remember seeing a guide somewhere. I'll see if I can track that down.16:53
yatinsmcginnis: it will be good if you can share16:54
smcginnisyatin: It's basically this process: https://wiki.openstack.org/wiki/Documentation/HowTo#How_to_amend_a_review-in-progress16:56
smcginnisThough I thought I read something somewhere specifically about updating someone else's commit.16:57
*** yatin has quit IRC16:57
thangpsmcginnis: https://wiki.openstack.org/wiki/Gerrit_Workflow16:57
thangpthere's a section on "Maintaining Dependency Chains"16:58
*** emagana has quit IRC16:58
*** emagana has joined #openstack-cinder16:59
smcginnisthangp: +1 that's good background too yatin if you have not read that.17:00
*** BaconIsKing has joined #openstack-cinder17:02
*** esker has joined #openstack-cinder17:03
*** emagana has quit IRC17:03
*** emagana has joined #openstack-cinder17:05
e0nenew issues with running unit tests with concurrency>1. :(17:06
*** amoturi1 has quit IRC17:06
*** annegent_ has quit IRC17:06
*** Apoorva has joined #openstack-cinder17:09
*** annegent_ has joined #openstack-cinder17:10
*** amoturi1 has joined #openstack-cinder17:11
*** nkrinner has quit IRC17:15
*** nellysmitt has quit IRC17:17
*** markstur has quit IRC17:18
*** jistr has quit IRC17:18
*** nellysmitt has joined #openstack-cinder17:18
*** rwsu-afk is now known as rwsu17:18
*** erlon has joined #openstack-cinder17:19
*** emagana has quit IRC17:19
*** coolsvap_ is now known as coolsvap17:20
openstackgerritShay Halsband proposed openstack/cinder: XtreamIO version 4.0 support  https://review.openstack.org/15297317:21
*** rhe00_ has joined #openstack-cinder17:21
*** jungleboyj has quit IRC17:23
*** bill_az has quit IRC17:25
*** kallebe has joined #openstack-cinder17:25
*** afazekas has quit IRC17:26
*** vilobhmm has joined #openstack-cinder17:27
*** Ilja has joined #openstack-cinder17:27
*** emagana has joined #openstack-cinder17:28
*** annashen has joined #openstack-cinder17:28
*** afazekas has joined #openstack-cinder17:28
kallebeHello. Does CI have a SAN storage which I can to test for SAN generic driver?17:28
IljaHello, is there anyone here from HP/3par driver team? There's a pretty critical bug on cinder for a while that doesn't seem to be addressed.17:30
*** bill_az has joined #openstack-cinder17:30
*** leeantho has joined #openstack-cinder17:30
*** ronis has joined #openstack-cinder17:30
*** hemnafk is now known as hemna17:31
hemnaIlja, ?17:31
*** markstur has joined #openstack-cinder17:31
*** _cjones_ has quit IRC17:33
hemnamust not be that urgent17:35
*** _cjones_ has joined #openstack-cinder17:36
*** annegent_ has quit IRC17:37
*** timcl has quit IRC17:39
*** rhe00_ is now known as rhe0017:40
openstackgerritTomoki Sekiyama proposed openstack/cinder: Limit volume copy bandwidth per backend  https://review.openstack.org/14944617:40
*** ebalduf has joined #openstack-cinder17:41
*** lpetrut has joined #openstack-cinder17:42
*** ebalduf has quit IRC17:45
*** jungleboyj has joined #openstack-cinder17:49
*** fischerw has joined #openstack-cinder17:55
*** karimb has quit IRC17:56
*** rhagarty has quit IRC17:56
Ilja@hemna, sorry, urgent, but irc client is not so stable17:57
IljaI was referring to this one: https://bugs.launchpad.net/cinder/+bug/141683617:57
openstackLaunchpad bug 1416836 in Cinder "Cinder 3PAR driver does not convert volume to base when restoring from snapshot" [Undecided,New]17:57
hemnaIlja, yes, that's how it's suppose to work17:58
hemnasnapshots are not base volumes17:59
hemnaif you want base volumes, you clone it.17:59
hemnaotherwise it's not a snapshot17:59
*** coolsvap is now known as coolsvap_18:00
*** annegent_ has joined #openstack-cinder18:01
Iljawell, the bug is not about that18:03
Iljait's possible to create a volume from snapshot in openstack terms18:03
Iljabut volume/snapshot have a bit different meaning in HP 3PAR18:03
hemnayes, which is a r/w snapshot18:03
hemnait's not a base volume.18:03
hemnaif you want a base volume, you have to clone it.18:03
hemnathat's how 3PAR works.18:03
hemnathis is not a driver bug18:04
hemnanor a 3PAR bug18:04
Iljanot really — it's possible to switch a snapshot volume to a base volume18:04
hemnait's a feature. :)18:04
*** annashen_ has joined #openstack-cinder18:04
*** dulek has quit IRC18:04
hemnaIlja, that's a clone18:04
hemnayou have to clone the volume to get a base volume.18:04
hemnawhich is what I mentioned above.18:04
Iljawell, I do feel it's a bug — conversion _does_ happen if the size of a cloned volume is larger. And _doesn't_ if the size is the same18:04
Iljalet'me look up code pointers18:05
*** Mandell has joined #openstack-cinder18:06
*** kmartin has quit IRC18:06
Iljanow what if https://github.com/openstack/cinder/blob/master/cinder/volume/drivers/san/hp/hp_3par_common.py#L1350 is False?18:07
*** ndipanov has quit IRC18:07
*** annashen_ has quit IRC18:09
hemnaso it looks like we try and comply with the new size, but creating a base volume18:09
hemnaif the size of the volume from snap is bigger than the original snap18:09
Iljaactually, this flow _must_ be creating a base volume in 3PAR terms18:10
hemnabut if they are the same, then we don't.  because it's only a r/w snapshot for a new volume.18:10
hemnaonly to support bigger sizes18:10
Iljar/w is only part of the lifecycle18:10
hemnabut if it's the same size, you don't want a new base volume18:11
hemnabecause it actually occupies more storage on the array18:11
hemnasnaps are just diffs18:11
Iljait introduces dependency on 3par level, which is not visible to openstack18:11
*** devlaps has joined #openstack-cinder18:11
hemnavolumes from snaps always have deps18:11
Iljaso imagine you are creating a volume with 'snapshot_to_volume' function18:11
hemnaotherwise it's not a snapshot18:11
hemnadon't use snapshots if you don't want deps.18:11
Iljaeee, what? In my world 'volume' is always a standalone entity I can delete. It can _have_ snapshots, that stop it from deletion, but if it doesn't have snapshots, I should be able to delete it18:12
hemnaIlja, no18:13
hemnayou are wrong18:13
hemnavolumes from snapshots on 3PAR are not base volumes.18:13
hemnaif you want a base volume, create a clone.18:13
hemnaI can only say that so many times.18:13
hemnainvalid bug.18:13
IljaBut look, do you agree that then the following openstack (not 3par!) behaviour is incorrect:18:14
Ilja- I see a volume. It doesn't have any snapshots. I issues a DELETE request. Systems says 202.18:14
Ilja- Nothing happens. In logs on debug level I see a message from 3par saying that the volume is in hierarchy (or smth similar, don't remember the exact wording)18:14
*** lpetrut has quit IRC18:15
*** Longgeek has quit IRC18:15
Iljaand I have no means with OpenStack (no metadata, no links) to detect how the deletion of this volume should happen18:15
*** marcusvrn1 has quit IRC18:16
*** ronis has quit IRC18:16
*** changbl has quit IRC18:17
*** marcusvrn1 has joined #openstack-cinder18:18
*** cdelatte has joined #openstack-cinder18:19
*** harlowja has joined #openstack-cinder18:20
*** patrickeast has joined #openstack-cinder18:20
*** annegent_ has quit IRC18:22
vilobhmmthingee : ping18:23
*** harlowja has quit IRC18:24
*** harlowja has joined #openstack-cinder18:24
*** changbl has joined #openstack-cinder18:27
*** asmith_brcd has joined #openstack-cinder18:27
vilobhmmdulek : ping yt ?18:27
*** marcusvrn has joined #openstack-cinder18:28
*** jordanP has quit IRC18:29
*** marcusvrn1 has quit IRC18:29
*** aix has quit IRC18:29
hemnahodos, https://wiki.openstack.org/wiki/Cinder/how-to-contribute-a-driver18:30
*** diegows has quit IRC18:30
hodoshemna: thanks18:31
*** jungleboyj has quit IRC18:31
hemnahodos, but please don't stop working on it.  Feel free to solicit feedback and reviews, just be aware that drivers are not a priority right now18:31
*** diemt has quit IRC18:33
openstackgerritAnthony Lee proposed openstack/cinder: HP 3par driver filter and evaluator function  https://review.openstack.org/15135318:33
*** diemt has joined #openstack-cinder18:34
*** ronis has joined #openstack-cinder18:34
openstackgerritHuang Zhiteng proposed openstack/cinder: Allow scheduler to receive volume stats when starting service  https://review.openstack.org/15862318:38
Ilja@hemna - I know I'm being annoying, but could you please comment on the this comment: https://bugs.launchpad.net/cinder/+bug/1416836/comments/5 ? We are running a somewhat big deployment with HP 3par and this issue is hitting us badly. Could you perhaps comment on what could be a correct way of treating of such immortal volumes?18:41
openstackLaunchpad bug 1416836 in Cinder "Cinder 3PAR driver does not convert volume to base when restoring from snapshot" [Undecided,Invalid]18:41
*** ebalduf has joined #openstack-cinder18:41
*** jdurgin1 has quit IRC18:42
*** timcl has joined #openstack-cinder18:44
*** ebalduf has quit IRC18:46
*** krtaylor has quit IRC18:48
*** amoturi1 has quit IRC18:50
*** marcusvrn1 has joined #openstack-cinder18:52
*** marcusvrn has quit IRC18:52
*** jcru has joined #openstack-cinder18:52
kallebeHello. Does CI(Jenkins) have a SAN storage which I can to test for SAN generic driver?18:54
*** timcl has quit IRC18:54
*** kmartin has joined #openstack-cinder18:55
*** timcl has joined #openstack-cinder18:55
smcginniskallebe: Which CI do you mean? The gate, or a local CI you have set up?18:56
kallebesmcginnis the gate18:56
smcginniskallebe: I believe it uses LVM.18:57
kallebesmcginnis ok, thanks.18:57
smcginniskallebe: np18:57
*** changbl has quit IRC18:58
*** jungleboyj has joined #openstack-cinder18:58
*** amoturi1 has joined #openstack-cinder18:58
*** krtaylor has joined #openstack-cinder19:00
*** lpetrut has joined #openstack-cinder19:01
*** timcl has quit IRC19:02
*** emagana has quit IRC19:03
*** annegent_ has joined #openstack-cinder19:03
*** emagana has joined #openstack-cinder19:04
*** bill_az has quit IRC19:04
*** hodos has quit IRC19:06
*** amoturi1 has quit IRC19:06
*** timcl has joined #openstack-cinder19:12
*** Longgeek has joined #openstack-cinder19:15
* e0ne wants do delete test_huawei_t_dorado :(19:16
*** _cjones_ has quit IRC19:16
*** jcru has quit IRC19:18
*** jcru has joined #openstack-cinder19:18
*** _cjones_ has joined #openstack-cinder19:21
*** thingee has joined #openstack-cinder19:22
*** Mandell has quit IRC19:23
*** annegent_ has quit IRC19:23
*** eharney has joined #openstack-cinder19:24
thingeeeharney: what's up with the blocking bugs on nova side for this https://review.openstack.org/#/c/147186/19:24
*** rushil has quit IRC19:25
*** avishay has joined #openstack-cinder19:25
eharneythingee: currently i'm the only one working to fix them and haven't gotten too far.  Nobody liked my initial "do this as root" patch, looks like some work is needed to have Nova query some volume info in a different way19:26
thingeejgriffith: you've been summoned on https://review.openstack.org/#/c/135002/19:26
eharneythingee: (it picks through libvirt info to figure out some things which it should probably calculate via other means)19:26
thingeeeharney: are we still planning to have this ready for march 1st to land in K?19:26
smcginnisthingee: Could you take a look at the response to your concerns in https://review.openstack.org/#/c/155894/ when you get a moment?19:26
*** annegent_ has joined #openstack-cinder19:26
thingeetbarron: https://review.openstack.org/#/c/149726/19:27
eharneythingee: Hard to say... possible but I don't know that it's too likely.  (I've been out of the office for a few days with a move, just diving back in a bit now.)19:27
thingeeeharney: traveling and a move. fun!19:28
tbarronthingee: I need to recheck19:28
thingeethat's how I felt with atlanta summit =/19:28
eharneythingee: Yes, the fun never ends... I'm hanging out for hours in the airport now with snow delaying my flight. :)19:28
thingeesmcginnis: I interpret that example as don't do this for LOG.exception... in the case I raised it for it's LOG.error19:30
openstackgerritMitsuhiro Tanino proposed openstack/cinder: Implement IET target driver  https://review.openstack.org/15882919:30
smcginnisthingee: My understanding is anything passed in to an exception should not have _Lx(), only _().19:31
smcginnisthingee: Who can confirm (or deny) that?19:31
thingeethen don't pass the translation into the exception. It's still a log. we should be translating it19:31
*** marcusvrn has joined #openstack-cinder19:32
smcginnisIt will be translated by the _(). Just not put in the LE catalog.19:32
thingeesmcginnis: http://paste.openstack.org/show/181387/19:33
smcginnisthingee: Aha, here's the relevant bit: "The log marker functions (_LI(), _LW(), _LE(), and _LC()) must only be used when the message is only sent directly to the log. Anytime that the message will be passed outside of the current context (for example as part of an exception) the _() marker function must be used."19:33
*** devlaps has quit IRC19:33
smcginnisthingee: Make sense to you?19:34
*** BharatK has quit IRC19:34
thingeesmcginnis: not exactly. See my example.19:34
thingeeI understand msg is being passed around, however, it's directly being translated within the LOG.error19:35
thingeethe translation never leaves that context19:35
*** marcusvrn1 has quit IRC19:35
smcginnisthingee: I see what you're saying. But that statement from the i18n guidelines seems to support the way I have it.19:35
thingeesmcginnis: honestly it's not clear to me with what I just raised.19:35
smcginnisthingee: In your example the exception would not be translated.19:35
thingeethere seems to be a bit of a gap with it's msg = _LE('LOLOLO') or msg = 'LOLOLO'19:35
thingeeso the translation is NOT being passed19:36
smcginnisthingee: But it should be.19:36
thingeeonly the message.19:36
thingeepassed around*19:36
smcginnisthingee: The only thing we don't want translated are debug messages, no?19:36
thingeeas I understood, yes19:36
smcginnisSo it would need to be translated somehow before being passed to exception.Laugh().19:37
smcginnisthingee: And the guidelines say then to use _().19:37
smcginnisthingee: We could do exception.Laugh(_(msg)), but it says not to do two different translations.19:38
thingeesmcginnis: this is honestly confusing, and I think we need someone to confirm instead of us attempting to interpret it. The example you linked to shows doing LOG.exception(_LE('There was an error.'))19:38
thingeeraise LocalExceptionClass(_('An error occured.'))19:39
smcginnisthingee: But the statement above says if it's going to a log and an exception then use _().19:39
smcginnisthingee: That part at least seems pretty explicit to me.19:39
thingeeso do that. I'm now saying making sure the LOG.error has LE()19:39
thingeethe current patch does not do that19:40
smcginnisthingee: That would be against that statement though.19:40
openstackgerritMitsuhiro Tanino proposed openstack/cinder: Clear migration_status from a destination volume if migration fails  https://review.openstack.org/15473919:40
smcginnisthingee: _Lx() used if _only_ going to the log, _() used if used multiple places.19:40
*** emaisne has joined #openstack-cinder19:40
hemnaxyang, https://review.openstack.org/#/c/15877619:40
thingeesmcginnis: sorry, but the doc still says:19:41
thingee# RIGHT19:41
thingeeLOG.exception(_LE('There was an error.'))19:41
thingeeraise LocalExceptionClass(_('An error occured.'))19:42
thingeeso we need clarification19:42
*** ebalduf has joined #openstack-cinder19:42
openstackgerritSteven Kaufer proposed openstack/cinder: GET volumes API sorting REST/volume/DB updates  https://review.openstack.org/14191519:42
openstackgerritSteven Kaufer proposed openstack/cinder: GET volumes API sorting enhancements common utilities  https://review.openstack.org/14191419:42
smcginnisthingee: In the example. But this is pretty explicit: "Anytime that the message will be passed outside of the current context (for example as part of an exception) the _() marker function must be used"19:42
thingeesmcginnis: get clarification.19:43
smcginnisthingee: Any idea who can clear that up?19:43
smcginnisthingee: Not sure where to go for that.19:43
thingeedhellmann probably, but he's a busy guy19:43
*** e0ne has quit IRC19:43
smcginnisthingee: OK. Seems trivial, but we should clear this up. It's used many places.19:44
*** Miouge has quit IRC19:44
smcginnisdhellmann: Not sure if you're around and available, but if you could comment on above that would be great.19:45
patrickeastjgriffith: hey, you around?19:46
*** eharney has quit IRC19:46
*** ebalduf has quit IRC19:46
openstackgerritDanny Wilson proposed openstack/cinder: Over subscription for Pure Storage iSCSI driver.  https://review.openstack.org/15768119:48
kauferthingee: If you have a minute, can you lok at this review? It hasn't been updated since the mid-cycle meetup:  https://review.openstack.org/#/c/148373/ ... anything I should be doing for it?19:48
*** avishay has quit IRC19:48
thingeesmcginnis, dhellmann: I've responded to this review. https://review.openstack.org/#/c/155894/19:50
openstackgerritDanny Wilson proposed openstack/cinder: Over subscription for Pure Storage iSCSI driver.  https://review.openstack.org/15768119:51
thingeesmcginnis: if you need someone for clarification, look at git. Whoever worked on the library and or docs19:51
openstackgerritKallebe Monteiro proposed openstack/cinder: pep8 E251 (keyword arguments and spaces) cleanup  https://review.openstack.org/15884019:52
*** casusbelli has joined #openstack-cinder19:52
smcginnisthingee: Sorry if I came of too hard. Just trying to understand and make sure we are using this right.19:53
*** annegent_ has quit IRC19:54
xyanghemna: you fixed it? cool!  I'll take a look soon19:55
*** avishay has joined #openstack-cinder19:56
hemnaxyang, yah19:58
*** jcru has quit IRC19:58
casusbellijgriffith: Hi! Again lurking for you regarding sos-ci, have a minute? :-)19:59
*** kallebe has quit IRC20:01
*** avishay has quit IRC20:01
*** dencaval has quit IRC20:04
*** sandywalsh_ is now known as sandywalsh20:04
*** annegent_ has joined #openstack-cinder20:04
*** amoturi1 has joined #openstack-cinder20:05
*** amoturi1 is now known as amoturi_20:05
*** annashen_ has joined #openstack-cinder20:05
*** BaconIsKing has quit IRC20:05
xyangthingee: hi, xtremio driver is update: https://review.openstack.org/#/c/152973/820:09
xyangthingee: I have one question for you20:09
hemnathingee, are you in the TC meeting ?  os-brick is an agenda item, fwiw.20:09
*** annashen_ has quit IRC20:10
xyanghemna: isn't os-brick already under cinder?20:10
hemnait's a governance change20:10
xyanghemna: ok20:10
hemnanot sure what the conversation will be about20:10
hemnalurking :P20:10
xyanghemna: :)20:11
*** sigmavirus24 is now known as sigmavirus24_awa20:11
*** sigmavirus24_awa is now known as sigmavirus2420:13
*** ronis has quit IRC20:14
*** nellysmitt has quit IRC20:26
*** lcurtis has quit IRC20:26
*** dalgaaf has quit IRC20:27
jungleboyjsmcginnis: thingee Sean is right.  If you have a message that is going to an exception, _() should be used for translation.20:29
jungleboyjI.E. msg = _('My message')20:30
jungleboyjLOG.error("Encountered exception: %s", msg)20:30
jungleboyjOr, cinder.exception(_('My error message'))20:31
jungleboyjIf not going to an exception do LOG.exception(_LE('My message'))20:32
smcginnisjungleboyj: Would you mind taking a look at the specific example in question and see if that meets your understanding?20:33
smcginnisjungleboyj: https://review.openstack.org/#/c/155894/2/cinder/volume/api.py20:33
mtaninojungleboyj: hi, thank you for the review.20:34
jungleboyjmtanino: You are welcome.  Sorry for having so many comments.20:35
mtaninojungleboyj: I saw a discussion about six.text_type about this https://review.openstack.org/#/c/154713/11/cinder/volume/targets/tgt.py yesterday on IRC.20:35
jungleboyjmtanino: Ok, hold on.20:35
mtaninojungleboyj: sure20:35
*** kvidvans has joined #openstack-cinder20:36
smcginnismtanino, jungleboyj: We tried that out. No, you should not need six.text_type here.20:36
smcginnismtanino: It is being passed in to the logger which does handle it correctly.20:36
smcginnismtanino: Only an issue if you format the string first using % then pass just the string in to the logger.20:36
jungleboyjsmcginnis: Oh, that is right.20:37
*** asmith_brcd has quit IRC20:37
jungleboyjmtanino: smcginnis is right.20:37
smcginnisxyang and I had to do some experimenting to confirm that.20:37
mtaninojungleboyj: So, in my case, I don't need use six.text_type. right?20:37
smcginnismtanino: Correct20:37
jungleboyjmtanino: Correct.  You can disregard those comments.20:38
mtaninosmcginnis: jungleboyj thanks, I understand.20:38
jungleboyjmtanino: Sorry for the confusion.20:39
mtaninojungleboyj: logger is confusiong :(20:39
jungleboyjmtanino: Tell me about it.20:39
smcginnisThese dang computer thingamajigs.20:40
*** dustins has quit IRC20:40
jungleboyjTell me about it.20:41
*** asmith_brcd has joined #openstack-cinder20:42
jungleboyjsmcginnis: I have Jim looking at that case to weigh in.  He is the expert.  I think you are correct but I want another opinion.20:42
*** ebalduf has joined #openstack-cinder20:43
*** kaisers1 has joined #openstack-cinder20:43
smcginnisjungleboyj: Perfect, that's all I want. I'd like to make sure we are using this properly as it is all over the place.20:43
jungleboyjsmcginnis: Agreed.  Jim helped write the code so he can give an authoritative answer.20:43
smcginnisjungleboyj: Awesome!20:44
jungleboyjNow, I need to figure out how to get the new oslo-config-generator to work for us.20:45
smcginnisjungleboyj: Bet you didn't know the full extent of what you were signing up for when you became the captian. :)20:46
jungleboyjsmcginnis: You are right.20:46
jungleboyjDon't want to admit how long it took for me to get the scheduler patch put together.20:47
*** rushil has joined #openstack-cinder20:47
*** ebalduf has quit IRC20:48
jungleboyjsmcginnis: Oh hey, it worked.  I just accidentally put it in the wrong file name.20:50
jungleboyjNow to figure out how to get all the right options in there.20:50
openstackgerritAbel Lopez proposed openstack/cinder: cinder-manage man update  https://review.openstack.org/15886120:52
smcginnisjungleboyj: :)20:52
*** Manshoon_ has joined #openstack-cinder20:53
*** BaconIsKing has joined #openstack-cinder20:54
*** Manshoon has quit IRC20:56
*** rushil_ has joined #openstack-cinder20:56
*** rushil has quit IRC20:56
*** akerr has quit IRC20:56
*** akerr has joined #openstack-cinder20:57
*** PaulCuzner has joined #openstack-cinder20:58
*** Mandell has joined #openstack-cinder20:59
thingeexyang: thanks20:59
*** BaconIsKing has quit IRC20:59
*** Lee1092 has quit IRC21:00
thingeehemna: I was not in the TC meeting. Was at lunch21:01
openstackgerritBob Callaway proposed openstack/cinder: Add Manage/Unmanage support to NetApp NFS drivers  https://review.openstack.org/15886421:03
xyangthingee: hi, so I tried the override option CONF.set_override('some_config_option', myvalue), it will always set some_config_option to myvalue whether it is already set in cinder.conf or not21:03
*** annegent_ has quit IRC21:04
openstackgerritAngela Smith proposed openstack/cinder: Fix issue where custom zone name prefix not being used by brcd driver  https://review.openstack.org/15886621:04
thingeejungleboyj: did you see my comment in https://review.openstack.org/#/c/155894/2/cinder/volume/api.py ?21:05
*** Manshoon has joined #openstack-cinder21:05
thingeexyang: can you check if it's set explicitly in the cinder.conf, and if not, override?21:07
xyangthingee: sure, that should work, if that is acceptable21:07
*** Manshoon_ has quit IRC21:07
xyangthingee: so the current xtremio patch still uses its own provisioning_factor, I'll ask him to check cinder.conf and override it if not there21:08
thingeexyang: I guess my only worry is you're overriding the default. The default is what's communicated to the user as what it'll be if you don't set it21:08
jungleboyjthingee: Yes and just got someone who works on the library to respond.21:08
*** emaisne has quit IRC21:08
xyangthingee: I start to wonder if we should change the default to a higher number21:08
thingeejungleboyj: ok, thank you!21:09
xyangthingee: pure also has a much higher default21:09
thingeexyang: changing default is fine, just can't backport it to stable.21:09
thingeexyang: I think changing the default is fine21:09
xyangthingee: it is not released yet21:09
thingeexyang: all the better :)21:09
*** maike has joined #openstack-cinder21:09
xyangthingee: this is the max_over_subscription_ratio we are talking about21:09
jungleboyjthingee: Welcome.  :-)  Glad to help.21:10
xyangthingee: jgriffith initially said my default of 1 is too low. it should be at least 2.  so I picked 2.  maybe I should set it to 2021:10
thingeejungleboyj: I'm still really confused with this documentation though. =/21:10
xyangthingee: agree documentation should have more examples to cover all cases21:11
thingeejungleboyj: ah nice james carey commented. :)21:12
jungleboyjthingee: Yes, and so he is going to patch the documentation and hopefully remove the confusion.21:13
thingeeexcellent, yes I'll definitely review this21:13
thingeeyay progress!21:13
smcginnisJust saw that. Great to have things clarified. I'd also like to take a look at the documentation update.21:13
smcginnisthingee: +1 for progress. Thanks for your time on this.21:14
thingeesmcginnis: I left a comment for james to add you to the review21:14
thingeesmcginnis: thank you21:14
xyangjungleboyj: one more thing to clarify.  I see LOG.debug(ex) in some places.  I tried it and it won't cause exception.  Is this correct?  will ex be translated in this format21:14
*** changbl has joined #openstack-cinder21:14
smcginnisthingee: Perfect, thanks!21:14
xyangjungleboyj: ignore the part about translation21:15
jungleboyjxyang: I think that is a bad idea.  Let me check.21:15
xyangjungleboyj: but if it is LOG.error(ex), I would worry about translation. thanks21:16
openstackgerritJon Bernard proposed openstack/cinder: RBD: remove non-layering support for antiquated versions  https://review.openstack.org/15887321:19
openstackgerritJon Bernard proposed openstack/cinder: RBD: Query volume features from ceph.conf  https://review.openstack.org/15887421:19
*** Manshoon has quit IRC21:19
*** Manshoon has joined #openstack-cinder21:20
jbernardthe ibm ds8000 ci is either the fastest CI ever, or terribly broken21:20
jungleboyjxyang: Do you have a specific example you are asking about?21:20
xyangjungleboyj: one sec21:20
jungleboyjjbernard: Yeah, I need to ping Eddie about that.  I saw that it is doing something weird.21:20
jungleboyjjbernard: Which review you looking at?21:21
jbernardjungleboyj: this one: https://review.openstack.org/#/c/158873/ and this one: https://review.openstack.org/#/c/158874/21:21
*** sandywalsh has quit IRC21:22
xyangjungleboyj: https://review.openstack.org/#/c/155989/19/manila/share/drivers/service_instance.py line 56521:22
jungleboyjjbernard: Thanks.  Let me ping Eddie.21:23
jbernardjungleboyj: thanks man21:23
jungleboyjjbernard: No problem.  Thank you.21:23
*** sandywalsh has joined #openstack-cinder21:23
*** diegows has joined #openstack-cinder21:23
*** Longgeek has quit IRC21:24
*** Manshoon has quit IRC21:24
openstackgerritEric Brown proposed openstack/cinder: Use oslo_config choices support  https://review.openstack.org/15858021:25
dannywilsonxyang: For max_over_subscription_ratio we decided to use 20 in the case of empty array/nothing provisioned21:26
xyangdannywilson: perfect. you just confirmed that 20 is a reasonable default21:27
openstackgerritAngela Smith proposed openstack/cinder: Fix issue where custom zone name prefix not being used by brcd driver  https://review.openstack.org/15886621:28
dannywilsonxyang: I think we just confirmed it together. :)21:28
xyangdannywilson: good:)21:28
jungleboyjxyang: You shouldn't do LOG.error(ex).  Your safest bet is LOG.error(six.text_type(ex))21:28
jungleboyjOtherwise you may blow up with the logger trying to do an str() on the exception.21:29
jungleboyjRegardless, sending the exception like that right to the logger is yucky.21:29
openstackgerritAngela Smith proposed openstack/cinder: Custom zone name prefix not being used by brcd driver  https://review.openstack.org/15886621:29
xyangjungleboyj: I tested it.  it doesn't blow up.  LOG.error("%s" % ex) will throw unicodeerror exception, but LOG.error(ex) doesn't21:30
xyangLOG.error("%s", ex) also doesn't throw exception21:30
xyangthe logger is doing six.text_type itself21:31
smcginnisjungleboyj: xyang and I had fun trying different permutations of logging unicode.21:31
smcginnisjungleboyj: She's the expert now. :)21:32
xyangjungleboyj, smcginnis: yes, Sean and I tested it yesterday:)21:32
jungleboyjsmcginnis: xyang masochists!21:32
smcginnisjungleboyj: Just been having fun trying to straighten out different advice. :)21:34
xyangjungleboyj: see here:  /opt/stack/cinder/cinder/openstack/common/log.py(318)process()21:34
xyangmsg = six.text_type(msg)21:34
*** lpetrut has quit IRC21:34
xyangjungleboyj: so as long as the msg reaches here, it will be wrapped with six21:35
jungleboyjxyang: Ok, that sounds familiar.21:36
xyangif we use LOG.error("%s" % ex) thought, it will throw unicode error at an earlier point21:36
xyangbecause it treats ex as str21:36
xyangjungleboyj: exception will be thrown here:  /usr/local/lib/python2.7/dist-packages/oslo_i18n/_message.py(160)__str__()21:37
xyangfrom oslo_i18n._i18n import _21:38
xyangmsg = _('Message objects do not support str() because they may '21:38
jungleboyjxyang: Ok, so, the answer is LOG.error(ex) is ok, but it is still yucky.21:39
xyangraise UnicodeError(msg)21:39
xyangjungleboyj: I wouldn't want to use that.  just wonder how I should comment on it if someone uses it:)21:39
*** _cjones_ has quit IRC21:39
jungleboyjI would say it is bad form.21:40
xyangask jungleboyj if you agree :)21:40
xyangif you don't agree21:40
jungleboyjxyang: :-)21:40
*** dustins has joined #openstack-cinder21:42
*** dustins has quit IRC21:42
jungleboyjxyang: If for no other reason, you are spitting an exception into the logs with no context around it.21:43
*** ebalduf has joined #openstack-cinder21:44
*** scottda has quit IRC21:44
xyangjungleboyj: so you suggest add some text explaining where the exception is coming from, what operation we are trying, etc?21:44
*** cdelatte has quit IRC21:44
mtreinishjgriffith: if you get a sec: https://review.openstack.org/158020 should be a quick one21:45
jungleboyjRight, so you have LOG.error(_LE("Exception encountered: %s", ex))21:45
xyangjungleboyj: ok, thanks21:45
jungleboyjThat is what is normally done.21:45
*** diegows has quit IRC21:46
jgriffithmtreinish: lgtm... do I need to test it or am I safe in assuming you've fully vetted it :)21:47
*** ebalduf has quit IRC21:48
anishmtanino: just in time, I was about to upload the same IET code today :)21:48
mtreinishjgriffith: heh, well I never bothered to run run_tests, but the tox output is here: http://logs.openstack.org/20/158020/2/check/gate-cinder-python27/189bd6d/console.html21:48
jgriffithmtreinish: hmm...21:48
jgriffithmtreinish: may be busted, just s sec21:48
*** akerr has quit IRC21:49
mtreinishit's also probably worth noting you guys also leak a lot of stdout and stderr during the unit tests21:49
jgriffithmtreinish: that all needs squashed, I agree21:49
*** akerr has joined #openstack-cinder21:49
jgriffithmtreinish: I thought we made some pretty good progress ripping that out, but maybe not21:49
mtreinishmy favorite one was: http://logs.openstack.org/20/158020/2/check/gate-cinder-python27/189bd6d/console.html#_2015-02-23_16_02_02_44221:49
mtreinishthat's a lot of wrapping21:49
*** _cjones_ has joined #openstack-cinder21:50
mtaninoanish: oh really....21:50
jgriffithmtreinish: oh dear  :(21:50
jgriffithmtreinish: ok.. I'm saving that off, maybe on the plane I'll nuke at least the zfssaapi stuff21:50
jgriffithmtreinish: ok... your test change looks good21:51
jgriffithmtreinish: just needed some updates on my env21:51
jgriffithmtreinish: but I will truly miss my colors :)21:51
mtaninoanish: Did you create test code? I don't have test code yet :(21:51
mtreinishjgriffith: heh, push a tempest-lib patch to add a color mode to subunit-trace :)21:51
jgriffithmtreinish: hehe.. don't know if I'll miss is that much :)21:52
*** karimb has joined #openstack-cinder21:52
jgriffithmtreinish: but that would be kinda fun to work on as a distractor at some point21:52
jgriffithmtreinish: oi!!  The stdout crap is very annoying isn't it21:52
jgriffithmtreinish: ok... thanks sir!21:52
jgriffithI'll look at what I can do to eradicate all the leaking output later tonight21:53
patrickeastjgriffith: hey, thanks for the reviews on my db table change, is it safe to assume a +2 on the code changes means you would be willing to do the same for the spec too?21:53
mtreinishjgriffith: sure, np21:53
mtreinishjgriffith: yeah it does get annoying. It's better than nova's problem though. They're leaking stdout outside the fixture so it doesn't get multiplexed properly. Leads to some interesting output in places21:54
jgriffithpatrickeast: specs!  specs!  You want specs!  You can't handle specs!21:54
thingeepatrickeast, jgriffith: I'll +2 the spec today21:55
jgriffiththingee: cool.. I won't have to try and find it :)21:55
patrickeastthingee: sweet, thanks21:55
thingeejgriffith: :)21:55
thingeepatrickeast: sorry about that21:55
jgriffithOk... gotta run, next leg of th ehourney is on!21:55
patrickeast^ https://review.openstack.org/#/c/153333/21:55
jgriffithcya all later21:55
thingeejgriffith: good luck!21:55
*** Yogi1 has quit IRC21:56
jgriffiththingee: left the +A for you :)21:56
thingeepatrickeast: was hemna ever going to comment on this?21:56
hemnawhich ?21:56
thingeethink I mentioned I would be comfortable with another vendor who was interested in this21:56
thingeehemna: https://review.openstack.org/#/c/153333/21:56
thingeehemna: in the midcycle meetup, you mentioned you had to work around this problem21:56
patrickeastah yea, sry i wasn’t too aggresive chasing people down on it21:56
hemnathingee, yah we did21:57
hemnaI had reviewed this one a while back21:57
anishmtanino: did not :) I also did the base code first21:57
hemnabut hadn't kept up on it21:57
anishSimilarly to yours, except you use cid better21:57
thingeehemna: would appreciate a +1 on this so I can +2 to it21:57
hemnaok reviewing now21:57
mtaninoanish: hahaha, same situation..21:57
patrickeastthingee: hemna: thanks for the help pushing it through!21:57
mtaninoanish: I got that cid idea from guitarzan21:58
anishmtanino: whoops, 1 second late21:58
hemnathingee, patrickeast question21:58
hemnasince this DB table is going to have credentials in it21:58
*** kvidvans_ has joined #openstack-cinder21:58
hemnashouldn't rows be hard deleted?21:58
hemnainstead of just marked as 'deleted' like all of our other tables do ?21:59
patrickeasthemna: i had wondered about that21:59
hemnanot sure how to enforce that at delete time.21:59
patrickeasti opted not to since it isn’t only for credentials21:59
hemnaI presume there is a way though21:59
patrickeastand even then i hope to eventually only store a barbican uuid21:59
patrickeastand the actual secret elsewhere21:59
hemnaso what happens to a row that does have credentials that's soft deleted ?21:59
hemnathe credentials are left in the db21:59
patrickeastit sticks around21:59
patrickeastnot ideal21:59
patrickeastits an easy change to hard delete them22:00
*** kvidvans has quit IRC22:00
patrickeastim cool with it either way22:00
*** tsekiyama has quit IRC22:00
*** tsekiyama has joined #openstack-cinder22:00
hemnawell, the reason I bring this is up, is that we also have another review in flight22:00
hemnathat archives the 'deleted' rows into shadow tables22:00
mtaninoanish: np.22:00
hemnaand so this data will be left around forever22:00
hemnaand migrated to another table eventually22:01
*** dustins has joined #openstack-cinder22:01
hemnaI'm not sure we want that for private credentials22:01
*** thangp has quit IRC22:01
hemnathingee, ?22:01
hemnajungleboyj, ?22:01
patrickeastyea i agree with that22:01
hemnapatrickeast, sorry about this.22:01
hemnaI just thought of it as I was looking this over again22:02
hemnabecause I'm also reviewing the shadow tables review as well22:02
patrickeasthemna: no worries, i’m glad to find this out now instead of later on when its a problem22:02
anishmtanino: I did all the cxt tests in mock, can I take a quick shot at getting you started while you concentrate on the main code22:03
patrickeastso it sounds to me like the question is whether the table should hard delete all the time, or only some of the time like a delete_now=True kind of deal22:03
hemnado we do this anywhere else?22:03
hemnacontrolled conditional deletes ?22:04
jungleboyjhemna: ?22:04
hemnajungleboyj, ^^ on the https://review.openstack.org/#/c/153333 review22:04
*** vilobhmm has quit IRC22:04
hemnajungleboyj, issue with the new db table I'm raising.22:04
hemnajungleboyj, the new table will be used to store CHAP creds22:05
hemnaand deleted rows are soft deletes22:05
*** vilobhmm1 has joined #openstack-cinder22:05
hemnawhich is just marking the row as deleted, but all the data is still there.22:05
*** annashen_ has joined #openstack-cinder22:06
patrickeasti kind of lean towards always hard delete22:06
mtaninoanish: That's much appreciate22:07
hemnapatrickeast, yah I think so as well22:07
*** cdelatte has joined #openstack-cinder22:08
hemnais there any reason why we would need those rows to stay around ?22:08
*** delattec has joined #openstack-cinder22:08
jungleboyjhemna: Yeah, seems like we would want to hard delete that.22:08
patrickeastmm other than convienent debugging tool, no22:08
patrickeasteven then22:08
hemnayah ok22:08
patrickeastok, one sec, i’ll update the spec22:08
hemnaso I vote for hard delete on that table22:08
*** tsekiyama has quit IRC22:08
hemnalet me add a comment in the review for history22:09
*** tsekiyama has joined #openstack-cinder22:09
*** kvidvans_ has quit IRC22:09
*** amoturi_ has quit IRC22:09
*** annashen_ has quit IRC22:10
hemnaok added22:10
*** maike has quit IRC22:13
*** maike has joined #openstack-cinder22:13
patrickeasthemna: thanks! i’ll have an update posted in just a min22:14
hemnaok cool22:14
*** changbl has quit IRC22:17
*** timcl has quit IRC22:17
*** jungleboyj has quit IRC22:19
openstackgerritPatrick East proposed openstack/cinder-specs: Add DB table for driver private data  https://review.openstack.org/15333322:19
patrickeastmade one other little change too, just noticed it was silly to pass a value for removing an entry, we only need the key22:19
patrickeaston a somewhat related note, does anyone know if its possible to get an admin auth token for the cinder user in a driver? i was playing around with using the key manager w/ barbican to store secrets but couldn’t figure out a good way to get the right context object22:24
patrickeastnot sure if thats something we do elsewhere (or at all) anywhere else22:25
mriedemis this a known gate bug? http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/logs/screen-c-api.txt.gz?level=TRACE#_2015-02-24_19_59_10_82022:25
mriedemlooks like that started happening around 2/2022:26
mriedemwhen it shows up though, the job is ultimately successful 93% of the time22:27
hemnapatrickeast, dunno about that one22:28
*** diemt has quit IRC22:28
mriedemmtreinish: look at this test which failed with a timeout: http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/console.html#_2015-02-24_20_12_55_70622:29
mriedemthen look at the slowest test in the run above it22:29
mriedemmtreinish: wonder if we should make those mutually exclusive?22:29
mtreinishmriedem: printing slowest and failures?22:30
*** erlon has quit IRC22:31
mriedemmtreinish: no, lock fixture on those tests22:32
mriedemtest_list_get_volume_attachments and test_rescued_vm_detach_volume22:33
mriedemmaybe that's just papering over an issue22:33
mtreinishmriedem: yeah, the operations aren't mutually exclusive so I don't think locking is correct there22:33
mriedemk, some of the traceback is confusing here http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/console.html22:34
mtreinishmriedem: fwiw, jgriffith has been looking at test_rescued_vm_detach_volume there are some interesting interactions going on there which is causing lvm to sit in a wait loop22:34
mriedemwith the order in teardown, so i'm trying to figure out what actually timed out22:34
mtreinishso that'll always be slow22:34
mriedemalright, there is a gate race there, just not sure if it's a known bug or not22:36
openstackgerritPatrick East proposed openstack/cinder-specs: Add DB table for driver private data  https://review.openstack.org/15333322:36
mriedemand my currenty logstash query is too generic probably22:36
openstackgerritEric Brown proposed openstack/cinder: Use oslo_config choices support  https://review.openstack.org/15858022:36
*** openstackgerrit has quit IRC22:37
*** openstackgerrit has joined #openstack-cinder22:37
mtreinishmriedem: looking through the logs it looks like it never detached22:40
mtreinishwhich was the initial failure, then it also raised an exc when it went to delete the server because there was still a volume attached to it22:40
jgriffithmriedem: mtreinish hey....22:42
jgriffithThat test sucks :)22:42
mtreinisherr probably volume delete is what failed actually22:42
jgriffithSo Nova/KVM do bad things22:42
jgriffithmakes LVM hang22:42
jgriffithbut in the past it hasn't caused "failures"22:42
jgriffithlooking now at the logstash you pasted22:43
*** dustins has quit IRC22:43
mtreinishmriedem: although the detach succeeded: http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/logs/screen-c-vol.txt.gz#_2015-02-24_20_00_48_05622:43
mriedemjgriffith: probably more useful: http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/logs/screen-c-vol.txt.gz#_2015-02-24_19_59_10_60322:43
*** Ilja has quit IRC22:43
*** dustins has joined #openstack-cinder22:43
jgriffithoh.. boy... that's new22:43
jgriffithmriedem: 20'th you say?22:44
*** ebalduf has joined #openstack-cinder22:44
mriedemjgriffith: yeah22:44
jgriffithmriedem: ok... two candidates, gimmie a minute22:45
mtreinishmriedem: oh ignore me, my link was the forced delete working22:45
mtreinish(dunno how I missed all that red)22:45
mriedemfor some reason i can't get logstash to find the stderr here http://logs.openstack.org/97/157497/2/check/check-tempest-dsvm-nova-v21-full/3cd8eef/logs/screen-c-vol.txt.gz#_2015-02-24_19_59_10_60322:45
jgriffithaccess control rule failure22:46
mtreinishmriedem: is it getting filtered out?22:46
jgriffithit's there...22:46
mriedemmriedem: idk22:46
jgriffithStderr: u'tgtadm: this access control rule does not exist\n'22:46
mriedemmtreinish: got myself that time22:46
mriedemmtreinish: i don't know why that wouldn't be indexed22:47
mtreinishmriedem: http://goo.gl/aB0OK822:47
jgriffithmriedem: mtreinish http://goo.gl/8xaf5522:47
mtreinishseems to work22:48
*** ebalduf has quit IRC22:49
mtreinishjgriffith: that broke things? looks safe to me :)22:50
mtreinishmriedem: heh, I just noticed you mixed up our nicks this time :)22:50
mriedemtime to leave soon22:51
jgriffithmriedem: not nice to mess with me like that :)22:51
mtreinishmriedem: do you want me to push a query for that error mesage?22:51
*** kaufer has quit IRC22:51
mriedemit's not 100% failure22:52
mriedemof course22:52
mriedembut it is newish as of 2/2022:52
mriedem59% fail when it shows up22:52
mtreinishheh, of course it isn't...22:52
mriedemmtreinish: jgriffith: bug reported https://bugs.launchpad.net/cinder/+bug/142531022:53
openstackLaunchpad bug 1425310 in Cinder "test_list_get_volume_attachments race fail - timeout waiting for volume to attach (stuck in detach)" [Undecided,New]22:53
mriedemmtreinish: go ahead with the e-r query, i have to head out22:53
mriedemnot sure if it's a dupe of 1421895 or not22:53
*** Ilja has joined #openstack-cinder22:54
*** dustins has quit IRC22:55
mtreinishmriedem: it might be a dup, or at least a partial dup22:56
mtreinishbug 1421895 looks like it's just looking at the tempest side22:56
openstackbug 1421895 in Cinder "test_encrypted_cinder_volumes_luks fails with "Volume is still attached, detach volume first."" [High,Confirmed] https://launchpad.net/bugs/142189522:56
mriedemi'm punting22:57
mriedemand leaving22:57
mtreinishtsk, tsk where's the work ethic :)22:57
*** mriedem has quit IRC22:57
jgriffithmtreinish: Not clear on where that ACL is getting setup or pulled in22:59
jgriffithmtreinish: unless my removal of retry_deactivate is doing something "funny"23:00
jgriffithmtreinish: but the lvm.conf has that set so it should still behave the same way23:01
jgriffithmtreinish: I can surely revert that change and see if our problem goes away23:01
jgriffithmtreinish: http://goo.gl/bGa3kT23:01
*** harlowja is now known as harlowja_away23:02
*** harlowja_away is now known as harlowja23:05
*** vilobhmm1 has quit IRC23:06
*** vilobhmm has joined #openstack-cinder23:06
*** emagana has quit IRC23:08
*** Longgeek has joined #openstack-cinder23:09
*** Yogi1 has joined #openstack-cinder23:09
mtreinishjgriffith: yeah if the lvm.conf has the same options, I'm not sure how that would make a diff23:12
*** xyang has quit IRC23:12
openstackgerritxing-yang proposed openstack/cinder: Modify Consistency Group API  https://review.openstack.org/14456123:12
*** _cjones_ has quit IRC23:12
mtreinishjgriffith: I'm assuming the --config flag doesn't drop all the options from the lvm.conf file (that would be weird) because that's the only thing I could think would be different23:12
jgriffithmtreinish: indeed23:13
jgriffithmtreinish: which is why I changed it23:13
jgriffithmtreinish: I can test that though... if it does drop everything else then that's LAME23:13
jgriffithmtreinish: BUT... the thing is we don't set up bind anywhere23:13
jgriffithmtreinish: so I don't know where that's coming from23:14
jgriffithmtreinish: re the ACL stuff... we use CHAP, no ACL23:14
*** _cjones_ has joined #openstack-cinder23:14
*** tbarron has quit IRC23:17
*** tbarron has joined #openstack-cinder23:17
*** emagana has joined #openstack-cinder23:17
mtreinishjgriffith: I've never used tgt-admin before (just tgtadm occasionally) but how are you telling it that? Because it loooks like the tgt-admin --force --delete is issuing 'tgtadm -C 0 --op unbind --mode target --tid 1 -I ALL'23:17
jgriffithmtreinish: so that's the thing :)23:18
jgriffithmtreinish: it is, but we don't use ACL so I don't know why the unbind is being called23:18
*** amoturi_ has joined #openstack-cinder23:18
jgriffithunless something "else" on the system is setting it in the tgtadm.conf file23:18
jgriffithbut given the timing, it sure seems like something with my change23:19
*** maike has quit IRC23:19
*** chlong has joined #openstack-cinder23:19
*** amoturi_ has quit IRC23:20
jgriffithmtreinish: I'm doing a fresh deploy on my end to see if I can repro and tag it23:20
mtreinishok, cool. fwiw, I'm still not making the jump where tgt-admin gets the config to know which tgtadm calls to make. (but that's just my general lvm ignorance)23:22
mtreinishso if you've got a grasp on it that's good enough for me :)23:22
jgriffithwell... in Cinder we have never paid any attention to the conf23:23
jgriffiththe only thing we do is persistent tgt files23:23
jgriffithbut just like most things, there is an /etc/tgt.conf file that can "do stuff"23:23
*** karimb has quit IRC23:25
*** jungleboyj has joined #openstack-cinder23:27
mtreinishjgriffith: ah, ok23:27
jgriffithmtreinish: anyway... loading a full stack and I'll see if I can find anything out before resorting to just a blind revert :)23:27
mtreinishI think I get it now. Do you think there would be value in also pulling the tgt.conf in the log output?23:29
jgriffithmtreinish: depends on if I can find antying locally or not :)23:29
jgriffithmtreinish: but honestly... for future anyway it might be a good idea23:30
*** _cjones_ has quit IRC23:30
jgriffithmtreinish: building wheel is sooooo slooowwwwww23:30
mtreinishjgriffith: heh, yeah it's that numpy install doing the blas lib compliation23:31
mtreinishblame nova23:31
jgriffithmtreinish: yeah... sighhh23:31
jgriffithmtreinish: I did/do23:31
*** primechuck has quit IRC23:32
*** thingee has quit IRC23:33
*** primechuck has joined #openstack-cinder23:34
*** jungleboyj has quit IRC23:34
mtreinishI should probably look into a way to speed up that compliation (assuming we can't remove it)23:34
*** rmesta has quit IRC23:34
jgriffithmtreinish: stack.d ?23:36
*** primechuck has quit IRC23:38
*** Longgeek has quit IRC23:42
*** _cjones_ has joined #openstack-cinder23:43
openstackgerritxing-yang proposed openstack/cinder: Create Consistency Group from CG Snapshot API  https://review.openstack.org/14595223:43
*** ebalduf has joined #openstack-cinder23:45
casusbellijgriffith: Hoi! Have a minute?23:46
*** jungleboyj has joined #openstack-cinder23:47
*** emagana has quit IRC23:47
openstackgerritxing-yang proposed openstack/python-cinderclient: Kilo Consistency Group CLI update  https://review.openstack.org/14456723:47
patrickeastxyang1: hey, just saw your comment on https://review.openstack.org/#/c/157681/5/cinder/volume/drivers/pure.py for the change dannywilson is working on, just to make sure i understand correctly, in the situation you described the scheduler would see 90GB available and still send volumes our way, right?23:48
*** amoturi_ has joined #openstack-cinder23:48
*** emagana has joined #openstack-cinder23:48
patrickeastxyang1: it wont stop scheduling just because the ratio hits 1.023:48
patrickeastxyang1: is my current understanding*23:48
*** ebalduf has quit IRC23:49
xyang1patrickeast: Can i ping you back in 10 min?23:50
*** chlong_ has joined #openstack-cinder23:50
patrickeastxyang1: yep23:50
*** emagana has quit IRC23:52
*** mriedem has joined #openstack-cinder23:53
jgriffithmriedem: so again comes back to the server rescue stuff...23:54
jgriffithmriedem: so this has been troubling to me23:54
jgriffithmriedem: because rescue does some things with LVM, and adding devs23:54
jgriffithhehe... mtreinish ^^23:55
jgriffithforgot who left and who stayed :)23:55
jgriffithcasusbelli: hey... what's up?23:55
*** Apoorva_ has joined #openstack-cinder23:56
mtreinishjgriffith: oh, fun. I feel like rescue will constantly be a problem23:57
jgriffithmtreinish: I'm HAPPY to propose a skip for those tests :)23:58
*** asmith_brcd has quit IRC23:58
jgriffithmtreinish: it seems to do "weird" things23:58
*** Apoorva has quit IRC23:59
mtreinishjgriffith: is there a bug for nova?23:59
jgriffithmtreinish: Yeah, I logged one...  just a sec23:59
mtreinishjgriffith: pushing a skip would basically be saying we can't test (or use) rescue reliably w/ cinder lvm23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!