Wednesday, 2016-02-03

*** smoriya_ has joined #openstack-cinder00:00
*** jamielennox|away is now known as jamielennox00:02
*** angela-s has quit IRC00:02
*** diablo_rojo has quit IRC00:03
*** Guest80711 is now known as med_00:15
*** med_ has quit IRC00:15
*** med_ has joined #openstack-cinder00:15
*** edtubill has joined #openstack-cinder00:21
*** alonma has joined #openstack-cinder00:22
*** alonma has quit IRC00:26
*** lprice has joined #openstack-cinder00:27
*** alonma has joined #openstack-cinder00:29
*** lprice1 has joined #openstack-cinder00:30
*** lprice has quit IRC00:32
*** alonma has quit IRC00:33
*** IlyaG has joined #openstack-cinder00:35
*** IlyaG has quit IRC00:37
*** jamielennox is now known as jamielennox|away00:38
*** jamielennox|away is now known as jamielennox00:39
*** mc_nair has quit IRC00:40
*** thurloat_isgone is now known as thurloat00:42
*** zhangjn has joined #openstack-cinder00:42
*** alonma has joined #openstack-cinder00:42
*** laughterwym has joined #openstack-cinder00:45
*** laughterwym has quit IRC00:46
*** alonma has quit IRC00:47
*** laughterwym has joined #openstack-cinder00:47
*** alonma has joined #openstack-cinder00:48
*** EinstCrazy has joined #openstack-cinder00:50
openstackgerritMitsuhiro Tanino proposed openstack/cinder: Permit volume type operations for policy authorized users  https://review.openstack.org/27420400:52
*** alonma has quit IRC00:53
*** edtubill has quit IRC00:54
*** qeelee has joined #openstack-cinder00:55
*** EinstCrazy has quit IRC00:55
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: 3PAR fix driver to work with image cache  https://review.openstack.org/23614300:55
*** cheneydc has joined #openstack-cinder00:57
*** mriedem_machell is now known as mriedem_afk00:58
*** JoseMello has quit IRC00:58
*** martyturner has quit IRC00:59
*** arch-nemesis has quit IRC01:02
*** rajinir has quit IRC01:05
*** garthb__ has quit IRC01:06
*** haomaiwang has joined #openstack-cinder01:12
*** IlyaG has joined #openstack-cinder01:18
*** gouthamr has joined #openstack-cinder01:21
*** gouthamr_ has joined #openstack-cinder01:23
*** davechen has joined #openstack-cinder01:24
*** gouthamr has quit IRC01:26
*** haomaiwang has quit IRC01:27
*** lixiaoy1 has joined #openstack-cinder01:30
*** jamielennox is now known as jamielennox|away01:30
*** Lee1092 has joined #openstack-cinder01:32
*** cknight has joined #openstack-cinder01:40
*** mtanino has quit IRC01:45
*** leeantho has quit IRC01:47
*** alonma has joined #openstack-cinder01:48
*** alonma has quit IRC01:53
*** alonma has joined #openstack-cinder01:55
*** alonma has quit IRC01:59
*** alonma has joined #openstack-cinder02:01
*** yhayashi has joined #openstack-cinder02:03
*** alonma has quit IRC02:05
*** alonma has joined #openstack-cinder02:07
*** salv-orl_ has joined #openstack-cinder02:09
*** salv-orlando has quit IRC02:12
*** alonma has quit IRC02:12
*** alonma has joined #openstack-cinder02:16
*** EinstCrazy has joined #openstack-cinder02:17
lixiaoy1wanghao:02:18
lixiaoy1wanghao: ping02:18
*** alonma has quit IRC02:21
*** alonma has joined #openstack-cinder02:22
*** haomaiwang has joined #openstack-cinder02:23
*** alonma has quit IRC02:27
*** jamielennox|away is now known as jamielennox02:31
*** xinyong has joined #openstack-cinder02:38
*** Thelo has quit IRC02:39
*** dims_ has quit IRC02:40
*** Apoorva has quit IRC02:41
openstackgerrityangweiwei proposed openstack/cinder: Update the method of backup-restore  https://review.openstack.org/27547802:43
*** daneyon_ has joined #openstack-cinder02:51
*** mc_nair has joined #openstack-cinder02:53
*** alonma has joined #openstack-cinder02:53
*** wilson-liu has quit IRC02:54
*** daneyon has quit IRC02:54
*** bill_az has quit IRC02:55
*** alonma has quit IRC02:57
*** mc_nair has quit IRC02:57
*** alonma has joined #openstack-cinder02:59
*** haomaiwang has quit IRC03:01
*** haomaiwang has joined #openstack-cinder03:01
*** alonma has quit IRC03:03
*** links has joined #openstack-cinder03:06
*** alonma has joined #openstack-cinder03:06
openstackgerritLisaLi proposed openstack/cinder: Clean uploading volume when c-vol restarts  https://review.openstack.org/27548003:07
*** alonma has quit IRC03:11
openstackgerritMerged openstack/cinder: Replication V2 for Pure Storage  https://review.openstack.org/21482503:32
openstackgerritMerged openstack/cinder: Remove deprecated options from NFS driver  https://review.openstack.org/26984103:33
*** alonma has joined #openstack-cinder03:34
*** alonma has quit IRC03:39
*** alonma has joined #openstack-cinder03:41
openstackgerritOpenStack Proposal Bot proposed openstack/python-cinderclient: Updated from global requirements  https://review.openstack.org/26616303:43
*** alonma has quit IRC03:45
*** esp_ has joined #openstack-cinder03:47
*** alonma has joined #openstack-cinder03:47
*** esp_ has quit IRC03:51
*** alonma has quit IRC03:52
*** alonma has joined #openstack-cinder03:53
*** qeelee has quit IRC03:53
*** gouthamr_ has quit IRC03:58
*** alonma has quit IRC03:58
*** alonma has joined #openstack-cinder04:00
*** haomaiwang has quit IRC04:01
*** haomaiwang has joined #openstack-cinder04:01
*** sheel has joined #openstack-cinder04:02
*** laughter_ has joined #openstack-cinder04:04
*** alonma has quit IRC04:05
*** alonma has joined #openstack-cinder04:06
*** laughterwym has quit IRC04:08
*** jamielennox is now known as jamielennox|away04:09
*** alonma has quit IRC04:10
*** mc_nair has joined #openstack-cinder04:12
*** alonma has joined #openstack-cinder04:12
*** alonma has quit IRC04:16
*** sheel has quit IRC04:18
*** deepakcs has joined #openstack-cinder04:20
*** mriedem_afk has quit IRC04:22
*** shyama has joined #openstack-cinder04:30
*** pradipm has joined #openstack-cinder04:33
*** sheel has joined #openstack-cinder04:33
*** qeelee has joined #openstack-cinder04:33
*** pradipm has quit IRC04:33
*** mc_nair has quit IRC04:41
*** qeelee has quit IRC04:50
*** EinstCrazy has quit IRC04:53
*** haomaiwang has quit IRC05:01
*** haomaiwang has joined #openstack-cinder05:01
*** esker has joined #openstack-cinder05:06
*** wanghao has quit IRC05:07
*** wanghao has joined #openstack-cinder05:07
*** bardia has quit IRC05:08
*** esker has quit IRC05:11
*** cknight has quit IRC05:12
openstackgerritAnkit Agrawal proposed openstack/cinder: NFS snapshots  https://review.openstack.org/14718605:13
*** cknight has joined #openstack-cinder05:13
*** cknight has quit IRC05:13
*** laughter_ has quit IRC05:21
*** laughterwym has joined #openstack-cinder05:21
*** mc_nair has joined #openstack-cinder05:28
*** mc_nair has quit IRC05:32
*** sgotliv has joined #openstack-cinder05:43
*** vgridnev has joined #openstack-cinder05:43
*** qeelee has joined #openstack-cinder05:49
*** nkrinner has joined #openstack-cinder05:50
openstackgerrityangweiwei proposed openstack/cinder: Update the method of backup-restore  https://review.openstack.org/27547805:52
*** mc_nair has joined #openstack-cinder05:56
*** mc_nair has quit IRC06:00
*** alonma has joined #openstack-cinder06:00
*** haomaiwang has quit IRC06:01
*** haomaiwang has joined #openstack-cinder06:01
openstackgerritMerged openstack/cinder: LeftHand: Implement v2 replication (unmanaged)  https://review.openstack.org/25554406:01
*** qeelee has quit IRC06:04
*** alonma has quit IRC06:04
*** ChubYann has quit IRC06:10
*** sheel has quit IRC06:10
*** cfriesen has quit IRC06:11
*** alonma has joined #openstack-cinder06:12
*** alonma has quit IRC06:16
openstackgerritvenkatamahesh proposed openstack/cinder: Fix the replication spelling in message  https://review.openstack.org/27550906:17
*** sgotliv has quit IRC06:19
*** alonma has joined #openstack-cinder06:20
*** alonma has quit IRC06:24
*** alonma has joined #openstack-cinder06:26
*** jaypipes has quit IRC06:26
*** mc_nair has joined #openstack-cinder06:29
*** alonma has quit IRC06:30
*** zul has joined #openstack-cinder06:33
*** qeelee has joined #openstack-cinder06:33
*** mc_nair has quit IRC06:36
*** alonma has joined #openstack-cinder06:38
*** anshul has joined #openstack-cinder06:40
*** alonma has quit IRC06:43
*** avishay has joined #openstack-cinder06:43
*** alonma has joined #openstack-cinder06:44
*** alonma has quit IRC06:49
*** zul has quit IRC06:51
*** chhavi has joined #openstack-cinder06:51
*** alonma has joined #openstack-cinder06:52
*** alonma has quit IRC06:56
*** alonma has joined #openstack-cinder06:58
*** abhishekk has joined #openstack-cinder06:58
*** haomaiwang has quit IRC07:01
abhishekkHi all, please review this cinderclient patch along with the dependent patches, https://review.openstack.org/#/c/257170/07:01
*** haomaiwang has joined #openstack-cinder07:01
abhishekkthank you07:01
*** alonma has quit IRC07:03
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Zanata  https://review.openstack.org/27552907:04
*** alonma has joined #openstack-cinder07:04
*** alonma has quit IRC07:08
*** alonma has joined #openstack-cinder07:10
*** esker has joined #openstack-cinder07:11
*** alonma has quit IRC07:14
*** IlyaG has quit IRC07:15
*** esker has quit IRC07:16
*** lpetrut has joined #openstack-cinder07:16
*** alonma has joined #openstack-cinder07:17
*** alonma has quit IRC07:22
*** alonma has joined #openstack-cinder07:24
*** laughterwym has quit IRC07:25
*** sandhya has joined #openstack-cinder07:25
*** laughterwym has joined #openstack-cinder07:25
sandhyaHi, I am trying to configure HP left hand in cinder. I find hplefthandclient for doing an install of the client. But I don't find hpelefthandclient. Any pointers on this?07:26
*** alonma has quit IRC07:28
*** laughterwym has quit IRC07:29
*** alonma has joined #openstack-cinder07:30
*** esker has joined #openstack-cinder07:32
*** alonma has quit IRC07:34
*** alonma has joined #openstack-cinder07:36
*** esker has quit IRC07:37
openstackgerrityangweiwei proposed openstack/cinder: Update the method of backup-restore  https://review.openstack.org/27554207:38
*** alonma has quit IRC07:40
*** alonma has joined #openstack-cinder07:42
*** alonma has quit IRC07:46
*** belmoreira has joined #openstack-cinder07:47
*** alonma has joined #openstack-cinder07:48
*** alonma has quit IRC07:53
openstackgerritLisaLi proposed openstack/cinder: Update quota when volume type renames.  https://review.openstack.org/26799907:53
*** esker has joined #openstack-cinder07:53
*** alonma has joined #openstack-cinder07:54
*** jwcroppe has quit IRC07:55
*** mvk_ has quit IRC07:56
*** esker has quit IRC07:58
*** jwcroppe has joined #openstack-cinder07:58
*** alonma has quit IRC07:59
openstackgerritLisaLi proposed openstack/cinder: Update quota when volume type renames.  https://review.openstack.org/26799907:59
*** alonma has joined #openstack-cinder08:00
*** haomaiwang has quit IRC08:01
*** haomaiwa_ has joined #openstack-cinder08:01
*** jwcroppe has quit IRC08:02
*** jwcroppe has joined #openstack-cinder08:04
*** alonma has quit IRC08:04
*** liverpooler has joined #openstack-cinder08:06
*** alonma has joined #openstack-cinder08:07
*** geguileo has joined #openstack-cinder08:07
*** zhangjn has quit IRC08:07
*** zhangjn has joined #openstack-cinder08:09
*** zhangjn has quit IRC08:10
*** salv-orlando has joined #openstack-cinder08:10
*** zhangjn has joined #openstack-cinder08:10
*** zhangjn has quit IRC08:11
*** laughterwym has joined #openstack-cinder08:11
*** zhangjn has joined #openstack-cinder08:11
*** zhangjn has quit IRC08:11
*** alonma has quit IRC08:12
*** zhangjn has joined #openstack-cinder08:12
*** zhangjn has quit IRC08:12
*** salv-orl_ has quit IRC08:12
openstackgerritKazumasa Nomura proposed openstack/cinder: Refactor Hitachi Block Storage Driver  https://review.openstack.org/27176308:13
*** lpetrut has quit IRC08:13
*** jistr has joined #openstack-cinder08:14
*** salv-orlando has quit IRC08:14
*** qeelee has quit IRC08:14
*** esker has joined #openstack-cinder08:14
*** salv-orlando has joined #openstack-cinder08:14
*** zhangjn has joined #openstack-cinder08:15
*** laughterwym has quit IRC08:15
*** zhangjn has quit IRC08:16
*** zhangjn has joined #openstack-cinder08:17
*** boris-42 has joined #openstack-cinder08:18
*** alonma has joined #openstack-cinder08:18
*** esker has quit IRC08:19
*** sgotliv has joined #openstack-cinder08:21
*** markus_z has joined #openstack-cinder08:21
*** alonma has quit IRC08:22
*** jistr is now known as jistr|mtg08:22
*** alonma has joined #openstack-cinder08:24
*** arecknag has joined #openstack-cinder08:25
*** sinese has joined #openstack-cinder08:25
*** zhangjn has quit IRC08:25
*** davechen has left #openstack-cinder08:27
*** mvk_ has joined #openstack-cinder08:27
*** alonma has quit IRC08:28
*** pots2 has quit IRC08:29
*** mc_nair has joined #openstack-cinder08:29
*** alonma has joined #openstack-cinder08:30
*** vgridnev has quit IRC08:32
*** mc_nair has quit IRC08:34
*** alonma has quit IRC08:35
*** esker has joined #openstack-cinder08:35
*** alonma has joined #openstack-cinder08:36
*** alonma has quit IRC08:37
*** alonma has joined #openstack-cinder08:38
*** zul has joined #openstack-cinder08:40
*** esker has quit IRC08:41
*** lpetrut has joined #openstack-cinder08:42
*** salv-orlando has quit IRC08:52
*** esker has joined #openstack-cinder08:56
*** smoriya_ has quit IRC08:58
*** haomaiwa_ has quit IRC09:01
*** EinstCrazy has joined #openstack-cinder09:01
*** esker has quit IRC09:01
*** haomaiwa_ has joined #openstack-cinder09:01
*** laurent\ has quit IRC09:09
*** chhavi has quit IRC09:16
*** esker has joined #openstack-cinder09:18
*** chlong has quit IRC09:20
*** ankit_ag has joined #openstack-cinder09:21
*** alonma has quit IRC09:22
*** markus_z has quit IRC09:22
*** zhangjn has joined #openstack-cinder09:22
*** alonma has joined #openstack-cinder09:22
*** esker has quit IRC09:23
*** zhangjn has quit IRC09:24
*** jordanP has joined #openstack-cinder09:24
*** yhayashi has quit IRC09:24
*** alonma has quit IRC09:26
*** EinstCrazy has quit IRC09:29
*** EinstCrazy has joined #openstack-cinder09:29
*** esp_ has joined #openstack-cinder09:32
*** alonma has joined #openstack-cinder09:34
*** alonma has quit IRC09:35
*** alonma has joined #openstack-cinder09:36
*** nkrinner has quit IRC09:38
*** esp_ has quit IRC09:39
*** itzdilip has joined #openstack-cinder09:42
*** e0ne has joined #openstack-cinder09:45
*** vgridnev has joined #openstack-cinder09:51
*** lpetrut has quit IRC09:54
*** tpsilva has joined #openstack-cinder09:55
*** zhangjn has joined #openstack-cinder09:56
*** zhangjn_ has joined #openstack-cinder09:57
*** zhangjn_ has quit IRC09:58
*** esker has joined #openstack-cinder09:59
*** zhangjn_ has joined #openstack-cinder10:00
*** zhangjn_ has quit IRC10:00
*** haomaiwa_ has quit IRC10:01
*** vivekd has joined #openstack-cinder10:01
*** haomaiwang has joined #openstack-cinder10:01
*** jordanP has quit IRC10:01
*** jordanP has joined #openstack-cinder10:01
*** zhangjn_ has joined #openstack-cinder10:02
*** zhangjn_ has quit IRC10:02
*** cheneydc has quit IRC10:02
*** zhangjn_ has joined #openstack-cinder10:02
*** zul has quit IRC10:03
*** sheel has joined #openstack-cinder10:04
*** esker has quit IRC10:04
*** e0ne has quit IRC10:05
*** nkrinner has joined #openstack-cinder10:05
*** zul has joined #openstack-cinder10:07
*** vgridnev_ has joined #openstack-cinder10:13
*** vgridnev has quit IRC10:14
*** zhangjn_ has quit IRC10:15
*** zhangjn_ has joined #openstack-cinder10:15
*** salv-orlando has joined #openstack-cinder10:16
*** laurent\ has joined #openstack-cinder10:17
*** ociuhandu has joined #openstack-cinder10:18
*** salv-orlando has quit IRC10:21
*** ildikov has quit IRC10:27
sheelDuncanT : Hi there10:27
*** alonma has quit IRC10:28
*** alonma has joined #openstack-cinder10:29
*** alonma has quit IRC10:29
*** alonma has joined #openstack-cinder10:30
*** lpetrut has joined #openstack-cinder10:32
*** avishay has quit IRC10:32
*** esker has joined #openstack-cinder10:41
*** esker has quit IRC10:45
*** e0ne has joined #openstack-cinder10:51
*** vgridnev_ has quit IRC10:52
*** vgridnev_ has joined #openstack-cinder10:54
*** deepakcs has quit IRC10:59
*** zul has quit IRC11:00
*** ndipanov has joined #openstack-cinder11:00
*** ndipanov has quit IRC11:00
*** haomaiwang has quit IRC11:01
*** ndipanov has joined #openstack-cinder11:01
*** zul has joined #openstack-cinder11:01
*** haomaiwang has joined #openstack-cinder11:01
*** esker has joined #openstack-cinder11:02
*** zhangjn has quit IRC11:02
*** zhangjn has joined #openstack-cinder11:03
*** zhangjn has quit IRC11:05
*** chhavi has joined #openstack-cinder11:07
*** esker has quit IRC11:07
*** salv-orlando has joined #openstack-cinder11:07
*** zhangjn_ has quit IRC11:13
*** deepakcs has joined #openstack-cinder11:13
*** zhangjn has joined #openstack-cinder11:13
*** esker has joined #openstack-cinder11:23
*** ndipanov has quit IRC11:27
*** jistr|mtg has quit IRC11:27
*** esker has quit IRC11:28
*** sandhya has quit IRC11:31
*** sheel has quit IRC11:38
*** ntt has joined #openstack-cinder11:39
*** deepakcs has quit IRC11:39
nttHi, it possible to run openstack-cinder-volume on a node and use lvm as backend from another server?11:39
*** jklare has joined #openstack-cinder11:42
jklarehi, anybody around?11:42
*** boris-42 has quit IRC11:43
*** baojg has joined #openstack-cinder11:44
*** vivekd_ has joined #openstack-cinder11:44
*** vivekd has quit IRC11:45
*** vivekd_ is now known as vivekd11:45
*** bkumar has joined #openstack-cinder11:45
*** sheel has joined #openstack-cinder11:46
*** merooney has joined #openstack-cinder11:52
e0nejklare: hi11:54
jklaree0ne hi :)11:54
e0nentt: no. you have to run cinder-volume on the same node where lvm is11:54
*** deepakcs has joined #openstack-cinder11:55
*** deepakcs has quit IRC11:55
ntte0ne: I'm searching for a HA solution. I already have a SAN, so I can expose a lun to the node where openstack-cinder-volume runs and use the lun as a disk for PV... but in this way I need to ensure HA for openstack-cinder-volume and the disk used as a backend. Is this the right way?11:56
jklaree0ne i am the ptl of the openstack-chef project and currently refactoring our cookbooks. For some reason i can not get the image service to appear on the dashboard and regarding this path (https://review.openstack.org/#/c/239091) it seems, that we would have to deploy cinderv1 and cinderv2, but that seems quite strange to me. Any comments on that? Could not find much helpful information while googling so far...11:57
nttfurthermore, using this method, datas will be multiple "hop" .... because if I write on one of the LV used by cinder volume, datas first will go to cinder-volume-node and then arrive on the san....11:57
e0nentt: we don't have HA for LVM now:(11:58
nttso.... if openstack-cinder-volume goes down, I need to switch services on another node and manually restart11:58
ntt?11:58
*** vivekd_ has joined #openstack-cinder12:00
*** haomaiwang has quit IRC12:01
*** haomaiwa_ has joined #openstack-cinder12:01
e0nejklare: AFAIR, it should works w/o cinder v112:02
e0nejklare: but you need to check for all services that they use cinderv2 endpoint12:02
jklaree0ne mmmh.. yeah i think so too, maybe i have to dive deeper into the logs12:02
e0nejklare: yea, logs will be very helpful12:02
*** julim_ has quit IRC12:03
jklaree0ne which service do you mean? shouldnt the appearance of the volumes catergory in horizon be idependent of any other service config?12:03
*** vivekd has quit IRC12:04
*** vivekd_ is now known as vivekd12:04
jklaree0ne the local_settings.py for horizon states ' "volume": 2 '12:04
*** esker has joined #openstack-cinder12:04
e0nejklare: I didn't try toget horizon working without api v112:05
e0nejklare: do you have any errors in logs?12:05
jklaree0ne sadly no :(12:05
e0ne:(12:05
jklaree0ne thats why i came here to maybe figure out if its v1 is always needed to get horizon to work :)12:06
*** yumapath has joined #openstack-cinder12:06
*** ildikov has joined #openstack-cinder12:07
david-lylejklare: I think v1 is required, shouldn't be, but is12:07
jklaredavid-lyle yeah, seems like that12:07
david-lylehopefully won't be by m-312:07
david-lylebut there should be a bug open for that12:08
jklaredavid-lyle have not found any, but maybe i used the wrong criteria to search12:09
*** esker has quit IRC12:09
*** avishay has joined #openstack-cinder12:10
david-lylejklare: https://bugs.launchpad.net/horizon/+bug/1415712 is what I was thinking of12:10
openstackLaunchpad bug 1415712 in OpenStack Dashboard (Horizon) ""Volumes" tab doesn't show up if cinder v1 isn't registered" [High,In progress] - Assigned to Itxaka Serrano (itxakaserrano)12:10
jklaredavid-lyle thank you so much12:11
jklaredavid-lyle no idea why i did not find this12:11
*** julim has joined #openstack-cinder12:11
*** links has quit IRC12:11
david-lylejklare: was tagged helpfully12:12
david-lyle*wasn't12:12
david-lylefixed that12:12
jklaredavid-lyle ahhh, its in the horizon launchpad bugs, i was looking through the cinder one... my bad12:12
jklaredavid-lyle but yeah, its more a bug in horizon than in cinder12:13
*** gouthamr has joined #openstack-cinder12:15
*** raildo-afk is now known as raildo12:17
*** geguileo has quit IRC12:20
*** zul has quit IRC12:24
*** aagrawal has joined #openstack-cinder12:28
*** baojg has quit IRC12:29
*** ociuhandu_ has joined #openstack-cinder12:30
*** yumapath has quit IRC12:31
*** dims has joined #openstack-cinder12:32
*** haomaiw__ has joined #openstack-cinder12:32
*** qeelee has joined #openstack-cinder12:33
*** ankit_ag has quit IRC12:34
*** haomaiwa_ has quit IRC12:34
*** lixiaoy1 has quit IRC12:34
*** shyama has quit IRC12:34
*** ociuhandu has quit IRC12:34
*** ociuhandu_ is now known as ociuhandu12:34
*** shyama has joined #openstack-cinder12:34
*** abhishekk has quit IRC12:36
*** zhangjn has quit IRC12:39
*** gouthamr has quit IRC12:39
-openstackstatus- NOTICE: Infra running with lower capacity now, due to a temporary problem affecting one of our nodepool providers. Please expect some delays in your jobs. Apologies for any inconvenience caused.12:39
*** zhangjn has joined #openstack-cinder12:42
*** esker has joined #openstack-cinder12:46
*** zhangjn has quit IRC12:47
*** zhangjn has joined #openstack-cinder12:48
*** zhangjn has quit IRC12:48
*** zhangjn has joined #openstack-cinder12:49
*** esker has quit IRC12:51
*** julim has quit IRC12:54
*** zul has joined #openstack-cinder12:55
*** sinese has quit IRC12:56
*** julim has joined #openstack-cinder12:57
*** sinese has joined #openstack-cinder12:57
*** bill_az has joined #openstack-cinder13:00
*** haomaiw__ has quit IRC13:01
*** haomaiwa_ has joined #openstack-cinder13:01
*** sinese_ has joined #openstack-cinder13:01
*** openstackgerrit has quit IRC13:02
*** openstackgerrit has joined #openstack-cinder13:03
*** sinese has quit IRC13:04
*** geguileo has joined #openstack-cinder13:05
*** zhipeng has joined #openstack-cinder13:10
*** merooney has quit IRC13:10
*** merooney has joined #openstack-cinder13:10
*** nibalizer has joined #openstack-cinder13:12
*** jistr|mtg has joined #openstack-cinder13:12
*** haomaiwa_ has quit IRC13:16
*** alonma has quit IRC13:17
*** alonma has joined #openstack-cinder13:17
*** lprice1 has quit IRC13:20
*** gouthamr has joined #openstack-cinder13:21
*** alonma has quit IRC13:22
*** lprice has joined #openstack-cinder13:22
*** xyang1 has joined #openstack-cinder13:23
*** alonma has joined #openstack-cinder13:23
*** alonma has quit IRC13:28
*** timcl has joined #openstack-cinder13:29
*** alonma has joined #openstack-cinder13:29
*** laughterwym has joined #openstack-cinder13:30
*** gman-tx has joined #openstack-cinder13:31
*** gman-tx has quit IRC13:32
*** aagrawal has quit IRC13:33
openstackgerritMatan Sabag proposed openstack/cinder: Support for consistency groups in ScaleIO driver  https://review.openstack.org/26962413:33
*** itzdilip has quit IRC13:34
*** alonma has quit IRC13:35
*** alonma has joined #openstack-cinder13:35
*** sinese_ has quit IRC13:36
*** haomaiwang has joined #openstack-cinder13:36
*** itzdilip has joined #openstack-cinder13:38
*** edmondsw has joined #openstack-cinder13:38
*** amoturi has joined #openstack-cinder13:39
*** cdelatte has quit IRC13:39
*** sinese_ has joined #openstack-cinder13:39
*** fthiagogv has quit IRC13:40
*** qeelee has quit IRC13:42
*** zhipeng has quit IRC13:46
*** vivekd has quit IRC13:46
*** merooney has quit IRC13:47
*** akerr has joined #openstack-cinder13:48
*** baojg has joined #openstack-cinder13:51
*** merooney has joined #openstack-cinder13:52
*** cdelatte has joined #openstack-cinder13:55
*** abehl has joined #openstack-cinder13:56
*** erlon has joined #openstack-cinder13:57
*** isaacb has joined #openstack-cinder13:57
*** martyturner has joined #openstack-cinder13:58
*** cknight has joined #openstack-cinder13:59
*** martyturner has left #openstack-cinder14:00
*** baojg has quit IRC14:00
*** haomaiwang has quit IRC14:01
*** haomaiwa_ has joined #openstack-cinder14:01
*** itzdilip has quit IRC14:04
*** julim has quit IRC14:05
*** baojg has joined #openstack-cinder14:07
*** baojg has quit IRC14:08
*** skraynev has quit IRC14:08
*** baojg has joined #openstack-cinder14:09
*** isaacb has quit IRC14:09
*** mc_nair has joined #openstack-cinder14:11
*** baumann has joined #openstack-cinder14:12
*** zhangjn_ has joined #openstack-cinder14:15
*** rlrossit has joined #openstack-cinder14:16
*** jistr|mtg is now known as jistr14:20
*** zhangjn_ has quit IRC14:20
*** zhangjn_ has joined #openstack-cinder14:22
*** cdelatte has quit IRC14:22
*** krotscheck1 has joined #openstack-cinder14:23
*** baojg has quit IRC14:23
krotscheck1I've got a patch that's been idling in the queue for about a month. Is there some process thing I need to do to get cores to look at it? https://review.openstack.org/#/c/265409/14:23
*** merooney has quit IRC14:23
smcginniskrotscheck1: I'll take a look. No special process.14:24
smcginniskrotscheck1: Usually things get looked at sooner, but I think folks have been busy lately.14:24
krotscheck1smcginnis: No worries, I just want to make sure I don't get my hand slapped like I did when I asked nova to review things.14:25
*** jwcroppe has quit IRC14:26
scottdakrotscheck1: We don't slap here. We use the Love and Logic method14:26
*** vgridnev_ has quit IRC14:27
smcginnis:)14:27
*** vgridnev_ has joined #openstack-cinder14:27
smcginnisAnd then we get out the crowbar.14:27
* krotscheck1 throws a rainbow at scottda 14:27
* krotscheck1 throws a rainbow crowbar at smcginnis 14:29
*** merooney has joined #openstack-cinder14:30
*** bkumar has quit IRC14:30
*** baojg has joined #openstack-cinder14:32
*** baumann has quit IRC14:32
*** jgregor has joined #openstack-cinder14:34
*** skraynev has joined #openstack-cinder14:35
*** edtubill has joined #openstack-cinder14:36
*** cdelatte has joined #openstack-cinder14:40
*** delattec has joined #openstack-cinder14:42
*** merooney has quit IRC14:43
*** Julien-zte has joined #openstack-cinder14:44
*** Julien-zte has quit IRC14:44
*** cdelatte has quit IRC14:45
*** baojg has quit IRC14:46
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements  https://review.openstack.org/27572914:46
*** tongli has joined #openstack-cinder14:47
*** julim has joined #openstack-cinder14:48
*** JoseMello has joined #openstack-cinder14:48
*** baojg has joined #openstack-cinder14:49
*** anteaya has quit IRC14:49
*** cfriesen has joined #openstack-cinder14:50
*** diablo_rojo has joined #openstack-cinder14:50
*** pots1 has joined #openstack-cinder14:52
*** anteaya has joined #openstack-cinder14:54
*** pradipm has joined #openstack-cinder14:54
*** mriedem has joined #openstack-cinder14:54
*** cknight has quit IRC14:56
openstackgerritRyan McNair proposed openstack/cinder: WIP Fix enforcement of default child quota values  https://review.openstack.org/27573414:59
*** zhangjn has quit IRC15:00
*** haomaiwa_ has quit IRC15:01
*** pradipm has quit IRC15:01
*** cknight has joined #openstack-cinder15:01
openstackgerritMerged openstack/python-brick-cinderclient-ext: Update typos  https://review.openstack.org/26669615:01
*** haomaiwang has joined #openstack-cinder15:01
*** pradipm has joined #openstack-cinder15:02
*** delatte has joined #openstack-cinder15:03
*** delattec has quit IRC15:03
*** delatte has quit IRC15:04
*** zhangjn has joined #openstack-cinder15:04
*** timcl has quit IRC15:06
*** timcl has joined #openstack-cinder15:07
DuncanTkrotscheck1: Asking here is totally reasonable, though I wouldn't recommend more often than once every few days at most15:09
*** baumann has joined #openstack-cinder15:09
*** dustins has joined #openstack-cinder15:09
krotscheck1DuncanT: So what'd you say is the nag-threshold?15:09
*** eharney has joined #openstack-cinder15:10
DuncanTkrotscheck1: Every day feels like nagging to me to be sure.15:10
DuncanTkrotscheck1: A couple of times a week is fine15:10
DuncanTkrotscheck1: I actually like people who pop up to ask, it gives me a good feeling they'll respond to feedback rather than letting code rot with a -115:11
krotscheck1DuncanT: Yeah, I feel about the same.15:11
DuncanTkrotscheck1: Also, there's a header used by os-profiler I think, that I can't see mentioned in the patch15:12
*** cdelatte has joined #openstack-cinder15:12
DuncanTkrotscheck1: I don't remember what the header is, I vaguely remember it existing15:12
krotscheck1DuncanT: Lemme look15:12
*** jwcroppe has joined #openstack-cinder15:13
openstackgerritRyan McNair proposed openstack/cinder: WIP Fixup nested quotas to enforce child defaults  https://review.openstack.org/27482515:14
*** Yogi1 has joined #openstack-cinder15:14
krotscheck1DuncanT: X-Trace-Info, X-Trace-HMAC15:16
DuncanTkrotscheck1: Any chance of a followup patch, please? Thanks15:16
*** boris-42 has joined #openstack-cinder15:16
krotscheck1Of course, just a minute15:16
*** kurtmartin has joined #openstack-cinder15:17
*** mtanino has joined #openstack-cinder15:19
*** delattec has joined #openstack-cinder15:19
openstackgerritMichael Krotscheck proposed openstack/cinder: Added Keystone and RequestID headers to CORS middleware  https://review.openstack.org/26540915:19
openstackgerritMichael Krotscheck proposed openstack/cinder: Added osprofiler headers to cors middleware  https://review.openstack.org/27575115:19
krotscheck1ARGH15:19
krotscheck1WHAT15:19
krotscheck1NO15:19
* krotscheck1 did NOT want a rebase15:20
* krotscheck1 facepalms15:20
krotscheck1Oh well. Let's start this over again ;)15:20
*** Yogi1 has quit IRC15:21
*** cdelatte has quit IRC15:22
*** garthb has joined #openstack-cinder15:22
openstackgerritMitsuhiro Tanino proposed openstack/cinder: Permit volume type operations for policy authorized users  https://review.openstack.org/27420415:23
openstackgerritJin Nan Zhang proposed openstack/python-cinderclient: add tags to .gitignore  https://review.openstack.org/27575415:23
openstackgerritJose Porrua proposed openstack/python-cinderclient: Avoid logging sensitive info in http requests.  https://review.openstack.org/27575515:24
*** delatte has joined #openstack-cinder15:27
*** arecknag has quit IRC15:27
*** merooney has joined #openstack-cinder15:27
*** delattec has quit IRC15:28
*** delatte has quit IRC15:29
*** liverpooler has quit IRC15:31
*** esker has joined #openstack-cinder15:32
*** pradipm has quit IRC15:33
*** knikolla has joined #openstack-cinder15:33
*** alonma has quit IRC15:34
*** arch-nemesis has joined #openstack-cinder15:34
*** baojg has quit IRC15:35
*** baojg has joined #openstack-cinder15:35
*** cdelatte has joined #openstack-cinder15:37
*** delattec has joined #openstack-cinder15:40
*** alonma has joined #openstack-cinder15:40
*** cdelatte has quit IRC15:43
*** alonma has quit IRC15:44
*** porrua has joined #openstack-cinder15:45
*** laughterwym has quit IRC15:45
*** laughterwym has joined #openstack-cinder15:46
*** zul has quit IRC15:46
*** alonma has joined #openstack-cinder15:46
smcginniskrotscheck1: Since that's going through again, want to just add the osprofile headers to the one and abandon that other.15:48
smcginniskrotscheck1: I'm actually fine either way, just thinking that might be simpler.15:48
*** baojg has quit IRC15:49
*** zhangjn has quit IRC15:50
*** laughterwym has quit IRC15:50
mtaninoxyang1: Hi,  Do you know Peter still have retype problem? https://bugs.launchpad.net/cinder/+bug/153687315:50
openstackLaunchpad bug 1536873 in Cinder "'module' object has no attribute 'is_admin' when retyping" [High,New] - Assigned to Mitsuhiro Tanino (mitsuhiro-tanino)15:50
*** alonma has quit IRC15:51
*** zhangjn has joined #openstack-cinder15:51
*** zhangjn has quit IRC15:51
*** alonma has joined #openstack-cinder15:52
*** salv-orl_ has joined #openstack-cinder15:54
xyang1mtanino: I need to check with him15:56
*** baojg has joined #openstack-cinder15:56
mtaninoxyang1: If you have a chance, please check it. I tried to reproduced a bug, but it seems the bug was already fixed.15:56
*** salv-orlando has quit IRC15:57
*** alonma has quit IRC15:57
*** sheel has quit IRC15:57
*** amoturi has quit IRC15:57
*** nkrinner has quit IRC15:57
*** amoturi has joined #openstack-cinder15:58
*** alonma has joined #openstack-cinder15:59
xyang1mtanino: I'll get back to you tomorrow15:59
mtaninoxyang1: Thank you. I'll left a comment on bug from my test result.16:00
xyang1mtanino: thanks16:00
*** jgriffith_away is now known as jgriffith16:00
*** haomaiwang has quit IRC16:01
*** haomaiwang has joined #openstack-cinder16:01
*** jdurgin1 has joined #openstack-cinder16:02
*** vgridnev_ has quit IRC16:02
*** lennyb has quit IRC16:03
*** alonma has quit IRC16:03
*** thurloat is now known as thurloat_isgone16:03
*** timcl1 has joined #openstack-cinder16:04
*** esker has quit IRC16:05
*** alonma has joined #openstack-cinder16:05
*** esker has joined #openstack-cinder16:05
*** timcl has quit IRC16:07
*** alonma has quit IRC16:10
*** alonma has joined #openstack-cinder16:11
openstackgerritMerged openstack/python-cinderclient: CLI for backup snapshots  https://review.openstack.org/24563216:12
*** rajinir has joined #openstack-cinder16:12
*** shyama has quit IRC16:15
*** alonma has quit IRC16:15
*** shyama has joined #openstack-cinder16:15
*** lennyb has joined #openstack-cinder16:16
*** daneyon has joined #openstack-cinder16:24
jgriffithsmcginnis: replication v2.1 start https://review.openstack.org/#/c/275797/16:25
smcginnisjgriffith: Thank you!!16:25
jgriffithsmcginnis: sure... I don't think the object versioning fix I made for the merge conflict is *right* but I'll look at it later and check with dulek if I can't figure it out16:26
*** daneyon_ has quit IRC16:27
smcginnisjgriffith: Good enough for now. At least we can see some code.16:27
*** belmoreira has quit IRC16:28
openstackgerritKedar Vidvans proposed openstack/cinder: Volume manage/unmanage support to ZFSSA drivers  https://review.openstack.org/27146216:29
*** vincent_hou has joined #openstack-cinder16:29
*** shausy has joined #openstack-cinder16:34
*** anshul has quit IRC16:39
*** abehl has quit IRC16:39
*** mgarza_ has joined #openstack-cinder16:39
*** yumapath_ has joined #openstack-cinder16:40
*** mriedem is now known as mriedem_afk16:40
*** huanan_L has joined #openstack-cinder16:41
*** merooney has quit IRC16:42
*** julim has quit IRC16:42
*** huanan has quit IRC16:44
*** merooney has joined #openstack-cinder16:44
*** julim has joined #openstack-cinder16:44
*** ndipanov has joined #openstack-cinder16:48
*** aarefiev has quit IRC16:48
*** aarefiev has joined #openstack-cinder16:49
flip214smcginnis: yeah, it moved forward .... right into a -2. "Discussed at the nova/cinder midcycles and agreed that new rootwrap filters for nova/cinder/brick have to go through privsep in os-brick."16:49
smcginnisflip214: Heh, at least it got an answer I guess. :]16:49
smcginnisWe have a few things depending on that now.16:50
openstackgerritVincent Hou proposed openstack/cinder: Storwize: Implement v2 replication  https://review.openstack.org/23738716:50
*** merooney has quit IRC16:51
*** delattec has quit IRC16:51
*** jordanP has quit IRC16:52
*** jordanP has joined #openstack-cinder16:53
flip214jgriffith: I'll read the cheesecake spec, and change/redo the replication API for DRBD ... had a patch lined up, after some other (more important) changes, but I guess I can reuse quite some of that anyway.16:54
*** sheel_ has joined #openstack-cinder16:54
mc_nairsmcginnis: here are the bugs where you can still currently get yourself in trouble with Nested Quotas in Liberty today - https://bugs.launchpad.net/cinder/+bug/1531502, https://bugs.launchpad.net/cinder/+bug/153718916:54
openstackLaunchpad bug 1531502 in Cinder "Child project's default quota not enforced" [High,New] - Assigned to Ryan McNair (rdmcnair)16:54
*** lpetrut has quit IRC16:54
openstackLaunchpad bug 1537189 in Cinder "Cinder nested quotas does not handle unlimited quotas (-1) well" [High,New] - Assigned to Ryan McNair (rdmcnair)16:54
smcginnismc_nair: Thanks for working on that.16:56
mc_nairhttps://review.openstack.org/#/c/275734/ (still needs a few minor fixups) should fix the first issue and avoid the second bug going forward (because disables -1 on child)16:56
DuncanTsheel_: Ok, so I think patch set 5 can go in now, and it would be great if you could work with scottda to propose patch set 6 as a microversion against his new microversion code (which isn't up in the state it needs to be yet, I don't think, though he'd be able to tell you more about it)16:56
*** EinstCrazy has quit IRC16:56
sheel_DuncanT:ok thnx16:57
scottdasheel_: Yes, let me know and I can help16:57
scottdasheel_: What is the patch set in question?16:57
sheel_Scottda: thanks, its https://review.openstack.org/27223216:57
*** jistr has quit IRC16:57
*** belmoreira has joined #openstack-cinder16:57
scottdacool, I'll have a look16:58
sheel_scottda: would u be able to give some idea abt it today?16:58
DuncanTsheel_: That way you get the feature in, then we get a simple and easy-to-test change for a our first microversion, with a slightly cleaner API16:58
sheel_DuncanT: yes its right way16:59
*** e0ne has quit IRC16:59
scottdasheel_: possibly, but I'm prioritizing work on the microversion patch itself.16:59
sheel_scottda:please take ur time16:59
*** IlyaG has joined #openstack-cinder17:00
sheel_just let me know whenevee u find time to discuss17:00
scottdaok17:00
sheel_DuncanT:so i ll be releasing patchset5in a while17:00
mc_nairsmcginnis: sure thing - was happy to have something to dig into.  I'll give you some time before continuing my rant, but just lemme know when you want discuss more on the current state / where to go next17:00
*** haomaiwang has quit IRC17:01
*** Apoorva has joined #openstack-cinder17:01
sheel_Duncan:please review as per time availibility17:01
DuncanTsheel_: I'm only on intermittently since it is getting late, but ping me a link and I'll get to it17:01
*** 7F1AA87I3 has joined #openstack-cinder17:01
*** sheel_ has quit IRC17:01
*** sheel has joined #openstack-cinder17:02
*** itlinux has quit IRC17:02
sheelscottda DuncanT: Thnx17:02
*** alonma has joined #openstack-cinder17:04
mtaninoDuncanT: Hi, Could you revisit this metadata support patch? https://review.openstack.org/#/c/258649/ I have fixed your comment. Thanks.17:04
mtaninoDuncanT: I'd like to push glance side too.17:04
*** dims has quit IRC17:04
*** shyama has quit IRC17:05
*** shyama has joined #openstack-cinder17:05
*** dims has joined #openstack-cinder17:06
*** garthb has quit IRC17:06
*** fthiagogv has joined #openstack-cinder17:07
*** alonma has quit IRC17:08
*** alonma has joined #openstack-cinder17:10
*** dims has quit IRC17:12
*** sinese_ has quit IRC17:13
*** sgotliv has quit IRC17:14
*** alonma has quit IRC17:15
*** geguileo has quit IRC17:15
*** baojg has quit IRC17:16
*** baumann has quit IRC17:16
*** mvk_ has quit IRC17:16
*** leeantho has joined #openstack-cinder17:17
*** baumann has joined #openstack-cinder17:17
*** dims has joined #openstack-cinder17:18
*** jgriffith is now known as jgriffith_away17:25
*** avishay has quit IRC17:26
*** chhavi has quit IRC17:30
*** julim has quit IRC17:33
*** IlyaG has quit IRC17:33
openstackgerritMerged openstack/cinder: Fix the replication spelling in message  https://review.openstack.org/27550917:34
*** anshul has joined #openstack-cinder17:37
*** abehl has joined #openstack-cinder17:37
*** jgriffith_away is now known as jgriffith17:39
*** merooney has joined #openstack-cinder17:40
*** julim has joined #openstack-cinder17:42
*** garthb has joined #openstack-cinder17:43
jgriffithpatrickeast: Swanson smcginnis hemna let me know if that looks like what you expected so far17:44
*** belmoreira has quit IRC17:45
*** shyama has quit IRC17:45
patrickeastjgriffith: have only taken a quick look but so far so good17:45
smcginnisjgriffith: Yeah, quick pass so far. Nothing big stood out.17:45
patrickeastjgriffith: there was one comment on the spec after it merged asking about the list replication targets api, would be easy enough to leave that one in and take out the vref part17:46
patrickeastjgriffith: dunno if we want to do it now though or leave it for a later iteration17:46
patrickeasti'm not sure how helpful it really is17:47
*** vincent_hou has quit IRC17:47
jgriffithpatrickeast: I didn't see the question, but I kinda figured there wasn't much value in having it...17:48
jgriffithpatrickeast: the info is in the config file, and the capabilities should reflect that it's succesfuly configured17:48
jgriffithAt least I *think* :)17:49
patrickeastjgriffith: yep, totally agree17:49
jgriffithpatrickeast: I'll work up the reporting for that in the driver next maybe17:49
smcginnisjgriffith: Only reason for having a list I can think of would be to be able to pull out that information via API for some kind of external system.17:50
smcginnisOtherwise, yeah, just look at the config.17:50
jgriffithsmcginnis: maybe, but IMHO that would just morph into a list of backend_id's in this model17:50
patrickeastyea it could be much more simple17:51
jgriffithsmcginnis: Probably useful... I can look at adding that in17:51
smcginnisjgriffith: Yeah, not saying it's needed. I'm happy wiht the way it is. ;)17:51
jgriffithpatrickeast: yeah, honestly that whole thing was kind of a mess17:51
patrickeastjgriffith: hehe yea, the trick will be to avoid falling into the same trap this time too17:52
* smcginnis crosses fingers17:52
* patrickeast says that after suggesting we add another api >.>17:52
jgriffithpatrickeast: yeah... but I'm just going to be a jerk this time around (or... more of a jerk) :)17:52
smcginnisHah17:52
patrickeastlol17:52
jgriffithNO MORE SNOW FLAKES17:52
smcginnisBut they're so purdy.17:52
patrickeastwhat if its a really beautiful snowflake?!17:52
jgriffithI have a blow torch17:53
jgriffithMelt all the snow-flakes17:53
smcginnisHeh, I just shovelled out from 11" of snow this morning. Kill the snowflakes!17:53
patrickeasthaha17:53
jgriffithsmcginnis: don't get me started!!!  2 Feet at my place yesterday17:53
jgriffithBurried the Jeep/Plow in a 4 foot drift a mile from home17:53
jgriffiththen ran out of gas17:53
smcginnisjgriffith: Is it all melted by today?17:53
jgriffithhaha.. not this time17:54
patrickeastjgriffith: for the active_backend_id should I assume that will be a new parameter for the driver __init__ or do_setup?17:54
jgriffithThis winter has been weird/cold17:54
smcginnisOK, you win this round! :)17:54
*** jordanP has quit IRC17:54
jgriffithpatrickeast: actually, I'm putting that in the stats update... so the driver sets it when receiving the failover call.  Then it gets persisted in the DB so that yes, on init the driver can figure out where it's supposed to be pointing to17:54
jgriffithsmcginnis: LOL... one of the few times we beat out Minnesota17:55
jgriffithThankfully!!17:55
smcginnisWe still win on being frickin' cold.17:55
jgriffithno doubt... cold in my book is low twenties17:55
*** IlyaG has joined #openstack-cinder17:56
jgriffithbut the sun is still shining bright at least17:56
smcginnisHah!17:56
*** thurloat_isgone is now known as thurloat17:56
patrickeastjgriffith: yea so for the on init part, it would be a new param? or should i plan to do some db querying?17:57
* patrickeast might be missing something in the code17:57
patrickeastjust tryin to scope out the changes for our driver17:57
jgriffithpatrickeast: yeah, just realized what you're asking.  I had anticipated passing it in as an optional kwarg17:57
patrickeastjgriffith: ok cool17:57
smcginnisSwanson: Might want to pay attention if you're around. ^^17:57
jgriffithpatrickeast: keep the db stuff out of the driver17:57
SwansonWhat?17:58
patrickeastjgriffith: yea def, why i asked ;)17:58
smcginnisSwanson: Now's the time to hit up jgriffith with 20 questions. :D17:58
jgriffithpatrickeast: cool, yeah good point17:58
*** bkumar has joined #openstack-cinder17:58
smcginnisSwanson: https://review.openstack.org/#/c/275797/17:58
*** 7F1AA87I3 has quit IRC18:01
*** ndipanov has quit IRC18:01
*** haomaiwang has joined #openstack-cinder18:01
*** mudassirlatif has joined #openstack-cinder18:02
*** sinese has joined #openstack-cinder18:04
smcginnisjgriffith: I like this: +246, -55718:05
jgriffithsmcginnis: haha... we'll see how that ratio works out18:06
smcginnisjgriffith: It's a good start at least.18:06
jgriffithsmcginnis: yeah, certainly the right direction18:06
jgriffithsmcginnis: I'm planning to let others revert/remove all their old code in their drivers18:07
jgriffithsmcginnis: starting querying on rep patches to propose reverts, but it was a bit tangled and realized it might be best to let driver maintainers do that18:07
smcginnisjgriffith: That's probably safer. I know HPE wanted to refactor what they had rather than revert.18:07
smcginnisI'm guessing most will just modify what they have.18:07
jgriffithsmcginnis: yeah, I think that will be true for most18:07
smcginnisAnd hopefully be able to rip out a lot of code.18:07
jgriffithyeah, should be able to18:08
*** merooney has quit IRC18:08
*** bkumar has quit IRC18:09
*** alonma has joined #openstack-cinder18:11
*** liverpooler has joined #openstack-cinder18:14
*** alonma has quit IRC18:15
*** liverpooler has quit IRC18:15
*** merooney has joined #openstack-cinder18:15
*** liverpooler has joined #openstack-cinder18:15
*** baojg has joined #openstack-cinder18:17
*** alonma has joined #openstack-cinder18:17
ApoorvaHello, could you please review https://review.openstack.org/#/c/261126/ and https://review.openstack.org/#/c/268370/18:18
*** Apoorva is now known as apoorvad18:18
*** shausy has quit IRC18:20
*** alonma has quit IRC18:21
*** alonma has joined #openstack-cinder18:23
*** e0ne has joined #openstack-cinder18:24
patrickeastjgriffith: ok had a more thorough look through the code and how things would work, i had one concern about the active backend id thing and sort of pushing the drivers to maintain state for it18:28
patrickeastafter failover we would need to track that state in our driver which means any other c-vols (most problematic being active-active setups) may not know about it18:28
jgriffithpatrickeast: I don't think so18:28
*** IlyaG has quit IRC18:28
patrickeastjgriffith: so like, in our driver (and i think some others) for the v2 replication we modify the volume entry in the db so every driver method we can see who to talk to18:28
*** alonma has quit IRC18:28
patrickeastjgriffith: but we don't have it with this, right?18:28
jgriffithpatrickeast: so my take was that the service column would tell the driver which backend to access18:29
jgriffithpatrickeast: so the "some volumes here, some there" problem just goes away18:29
jgriffithpatrickeast: does that make sense?18:29
*** IlyaG has joined #openstack-cinder18:29
jgriffithpatrickeast: because that case is kinda silly IMHO... either the backend is toast or it's not18:30
patrickeastjgriffith: yea that makes sense, lemmie take another stab at explaining18:30
*** alonma has joined #openstack-cinder18:30
patrickeastjgriffith: so if you have two c-vols looking for the same backend in HA kind of setup18:30
jgriffithslowwwly for thick headed people like me :)18:30
jgriffithyes... I'm with ya18:30
patrickeastjgriffith: one of them handles the failover and updates its reference for which backend is active now18:30
patrickeastjgriffith: the other one wouldn't know the failover happened, and would, until restart or something, not get the right active id18:30
jgriffithpatrickeast: right, I think I see where you're going18:31
*** vgridnev has joined #openstack-cinder18:31
jgriffithpatrickeast: so you're saying the problem is there's no guarantee that the other backend ever gets the failover command?18:31
jgriffithpatrickeast: and as a result never updates it's internal memory version of the info18:31
openstackgerritSheel Rana proposed openstack/cinder: Bootable filter for listening volumes from CLI  https://review.openstack.org/27222918:32
patrickeastjgriffith: right, the older one we were basically stashing that info in the db and every driver method that would need to know could look at the volume/snapshot/whatever and do the right thing18:32
patrickeastjgriffith: dunno which way we want to go with this18:32
patrickeastjgriffith: could lean in and put some mechanism in to handle it... or maybe just update all the volumes/snaps after a failover18:33
*** baojg has quit IRC18:33
jgriffithpatrickeast: thinking18:33
*** alonma has quit IRC18:34
openstackgerritJose Porrua proposed openstack/python-cinderclient: Avoid logging sensitive info in http requests.  https://review.openstack.org/27575518:35
*** bkumar has joined #openstack-cinder18:36
*** alonma has joined #openstack-cinder18:36
*** alonma has quit IRC18:40
*** mvk_ has joined #openstack-cinder18:41
*** bkumar has quit IRC18:42
jgriffithpatrickeast: wonder if we need a reverse stats update18:43
*** alonma has joined #openstack-cinder18:46
*** liverpooler has quit IRC18:47
eharneydoes someone know what's supposed to happen currently when you SIGHUP cinder-volume?18:49
smcginniseharney: I remember some patches for that, but can't recall.18:49
patrickeastjgriffith: yea something like that would work18:50
smcginniseharney: Was SIGHUP for guru meditation reports?18:50
eharneysmcginnis: no, that was SIGUSR1/218:50
*** alonma has quit IRC18:50
*** lcurtis has joined #openstack-cinder18:50
smcginniseharney: Ah, that's it.18:50
patrickeastjgriffith: or even like calling do_setup again... although i guess that probably breaks tons of stuff18:50
eharneytl;dr funny stuff happens when you SIGHUP it18:50
smcginnisI remember it had to change from the original signal because it wasn't supported on every platform or something.18:51
patrickeastjgriffith: sort of like a step towards doing live config changes on a c-vol service :D18:51
smcginniseharney: "Funny" stuff.18:51
eharneywhich is probably a bug18:51
eharneysmcginnis: well, on one platform it said "Alarm clock" and died.  The less funny one was where it re-inited the LVM driver and started running a second dd wipe on the same LV device18:51
*** esp_ has joined #openstack-cinder18:51
*** crose has quit IRC18:51
*** ChubYann has joined #openstack-cinder18:52
smcginnisAlarm clock? That is funny.18:52
dimssmcginnis : we switched from SIGUSR1 to SIGUSR2 for guru meditation reports18:52
smcginnisdd'ing your drive, not so funny.18:52
dims(in oslo.reports)18:52
smcginnisdims: OK, thanks!18:52
*** alonma has joined #openstack-cinder18:52
smcginnisCouldnt' recall18:52
*** jgriffith is now known as jgriffith_away18:53
smcginnisGotta leave for a work event. I'll be offline the rest of the afternoon.18:53
smcginnisLeave a ping if you need anything from me. Might be back on this evening.18:53
smcginnisUnless the work event turns into a non-work event. Hah!18:54
*** mc_nair_ has joined #openstack-cinder18:55
openstackgerritSheel Rana proposed openstack/cinder: Use of metadata id instead of metadata name  https://review.openstack.org/27458918:56
*** yumapath_ has quit IRC18:56
*** mc_nair has quit IRC18:56
*** mc_nair_ is now known as mc_nair18:56
*** alonma has quit IRC18:56
*** jgriffith_away is now known as jgriffith18:57
*** esp_ has quit IRC18:58
*** alonma has joined #openstack-cinder18:58
*** liverpooler has joined #openstack-cinder18:58
*** bardia has joined #openstack-cinder18:59
*** rlrossit has quit IRC18:59
*** crose has joined #openstack-cinder19:00
*** haomaiwang has quit IRC19:01
*** haomaiwang has joined #openstack-cinder19:01
diablo_rojoxyang1: Hey Xing, I was wondering if, when you had a chance, you could look at the CI on this patch: https://review.openstack.org/#/c/267085/ ? I can't see anything related to the patch in the logs as to why the patch would be failing EMX-VNX CI when it passes everyone elses. No big rush, just thought I'd ask :)19:01
*** sheel has quit IRC19:02
*** alonma has quit IRC19:03
*** julim has quit IRC19:04
*** alonma has joined #openstack-cinder19:05
xyang1diablo_rojo: sure19:06
*** mylu has joined #openstack-cinder19:06
*** julim has joined #openstack-cinder19:06
*** geguileo has joined #openstack-cinder19:09
*** alonma has quit IRC19:09
diablo_rojoxyang1: Thank you!19:12
xyang1diablo_rojo: the failure was not caused by the patch. It failed before running any tests.  let's see how the rerun goes19:13
*** alonma has joined #openstack-cinder19:13
diablo_rojoxyang1: Okay cool. Thanks for looking at it so quickly.19:15
xyang1diablo_rojo: np19:15
*** esp_ has joined #openstack-cinder19:17
*** alonma has quit IRC19:18
*** jgriffith is now known as jgriffith_away19:18
*** esp_ has quit IRC19:21
*** jgriffith_away is now known as jgriffith19:29
*** merooney has quit IRC19:30
*** merooney has joined #openstack-cinder19:32
*** baojg has joined #openstack-cinder19:33
*** geguileo has quit IRC19:33
*** e0ne has quit IRC19:35
*** IlyaG has quit IRC19:36
*** liverpooler has quit IRC19:37
*** liverpooler has joined #openstack-cinder19:38
*** baojg has quit IRC19:38
*** JoseMello has quit IRC19:41
*** JoseMello has joined #openstack-cinder19:41
*** krotscheck1 is now known as krotscheck_dcm19:42
*** vgridnev has quit IRC19:43
*** boris-42 has quit IRC19:43
mc_nairxyang1: hey - I was looking at your comment on https://review.openstack.org/#/c/275408/1.  Looks like the change from ">=" to ">" from comments in https://review.openstack.org/#/c/185764/7/cinder/scheduler/filters/capacity_filter.py.  The suggestion was to make that change in order to have thin-volumes behave like thick volumes when using max_oversubscription = 1, but it seems to have the opposite affect because it fails back to looking at19:47
*** ociuhandu has quit IRC19:47
*** abehl has quit IRC19:48
*** anshul has quit IRC19:48
mc_nairbut then again I may just be totally off-base, was some tough mental gymnastics trying to figure out expected behavior :)19:49
*** mgarza__ has joined #openstack-cinder19:50
*** vincent_hou has joined #openstack-cinder19:51
*** liverpooler has quit IRC19:51
*** mgarza_ has quit IRC19:53
*** rajinir has quit IRC19:55
openstackgerritJose Porrua proposed openstack/python-cinderclient: Avoid logging sensitive info in http requests.  https://review.openstack.org/27575519:59
*** haomaiwang has quit IRC20:01
*** 7JTAAES9V has joined #openstack-cinder20:01
*** rlrossit has joined #openstack-cinder20:01
*** vgridnev has joined #openstack-cinder20:02
*** merooney has quit IRC20:06
*** mudassirlatif_ has joined #openstack-cinder20:06
*** mudassirlatif has quit IRC20:09
*** mudassirlatif_ is now known as mudassirlatif20:09
*** dustins has quit IRC20:09
openstackgerritEric Harney proposed openstack/cinder: Improve logging to debug invalid "extra_specs" entries  https://review.openstack.org/27590920:11
*** rlrossit has quit IRC20:12
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor5 iSCSI driver unit tests  https://review.openstack.org/27153720:13
*** geguileo has joined #openstack-cinder20:13
openstackgerritVincent Hou proposed openstack/cinder: Migration: take the local_path for the source volume  https://review.openstack.org/25022020:14
*** alonma has joined #openstack-cinder20:14
openstackgerritMitsuhiro Tanino proposed openstack/python-cinderclient: Support name option for volume restore  https://review.openstack.org/27591020:15
*** merooney has joined #openstack-cinder20:18
*** alonma has quit IRC20:18
*** mgarza__ has quit IRC20:19
xyang1mc_nair: that change does have problem.  I just want to make sure we don't keep changing them back and force20:19
*** rlrossit_ has joined #openstack-cinder20:20
mc_nairxyang1: yea, for sure.  I'm not overly familiar with that code path, but gave my take on it above.  You want to discuss here or is it easiest if I just add my comment to the patch and we'll hash out there?20:23
*** mylu has quit IRC20:23
*** sgotliv has joined #openstack-cinder20:25
*** mylu has joined #openstack-cinder20:26
*** lprice has quit IRC20:30
xyang1mc_nair: we can discuss here if you want20:30
*** lprice has joined #openstack-cinder20:30
*** sgotliv has quit IRC20:36
*** gouthamr has quit IRC20:36
*** gouthamr has joined #openstack-cinder20:36
*** baojg has joined #openstack-cinder20:37
mc_nairxyang1: cool.  So from reading the config help, we say "A ratio lower than 1.0 will be ignored and the default value will be used instead." I haven't seen any validation that checks if we're using < 1 ratio and defaults it to 20.  If we had that, I think the >1 part of the check could go away, as I think a ratio of "1" should enter the if check, because we really want to be comparing the provisioned capacity to 1 and not free capacity for20:39
mc_nairdoes that make sense to you?20:39
*** lprice1 has joined #openstack-cinder20:41
*** lprice has quit IRC20:41
xyang1mc_nair: I think that help text is no longer accurate20:43
*** thurloat is now known as thurloat_isgone20:43
*** baojg has quit IRC20:45
mc_nairxyang1: I'd agree it's not accurate, do you know if that was intentional or not?  I.e. should we be fixing the help tip or the code?  My thought was we should be changing the code to be inline with the help tip20:45
xyang1mc_nair: if someone sets it to negative and we use 20, that is too dramatic20:46
*** mylu has quit IRC20:49
*** angela-s has joined #openstack-cinder20:49
*** mylu has joined #openstack-cinder20:49
*** crose has quit IRC20:50
*** apoorvad has quit IRC20:53
openstackgerritPatrick East proposed openstack/cinder: Allow for eradicating Pure volumes on Cinder delete  https://review.openstack.org/27591920:53
*** crose has joined #openstack-cinder20:55
*** dims has quit IRC20:56
*** 7JTAAES9V has quit IRC21:01
*** haomaiwa_ has joined #openstack-cinder21:01
*** merooney has quit IRC21:03
*** vincent_hou has quit IRC21:05
*** edtubill has quit IRC21:05
*** edtubill has joined #openstack-cinder21:05
*** dims has joined #openstack-cinder21:05
*** kurtmartin has quit IRC21:06
*** mylu has quit IRC21:07
*** mylu has joined #openstack-cinder21:07
openstackgerritJay Bryant proposed openstack/python-cinderclient: Remove debug statement  https://review.openstack.org/27592721:09
*** raildo is now known as raildo-afk21:11
*** apoorvad has joined #openstack-cinder21:13
*** alonma has joined #openstack-cinder21:14
*** boris-42 has joined #openstack-cinder21:14
mc_nairxyang1: ok, so then why don't we just not allow them to set < 1 ratio if we're using some other value anyway21:14
xyang1mc_nair: so the check needs to be in the driver itself, not in the scheduler21:16
*** esp_ has joined #openstack-cinder21:16
xyang1mc_nair: we can add a check in lvm driver21:17
xyang1mc_nair: other driver can add the check too21:18
*** alonma has quit IRC21:18
*** vgridnev has quit IRC21:18
*** akerr has quit IRC21:19
mc_nairxyang1: not following.... how does that help us in the case that the user put max_over_subscription_ratio = 0.5?21:19
xyang1mc_nair: check_for_setup_error21:20
*** diablo_rojo has quit IRC21:20
xyang1mc_nair: handle it the same way as other invalid config21:20
*** esp_ has quit IRC21:21
*** Lee1092 has quit IRC21:22
*** alonma has joined #openstack-cinder21:23
*** geguileo has quit IRC21:24
mc_nairxyang1: are you just saying we would need to slide the validation of < 1 ratio somewhere that would fail early as oppose to in that part of the scheduler?21:24
*** huanan_L has quit IRC21:25
openstackgerritTom Barron proposed openstack/cinder: WIP -  ABC cleanup work - WIP  https://review.openstack.org/27231821:26
*** huanan_L has joined #openstack-cinder21:26
*** julim has quit IRC21:27
xyang1mc_nair: the check in scheduler is still needed21:27
xyang1mc_nair: but we can add an extra check in the driver21:27
*** alonma has quit IRC21:28
*** alonma has joined #openstack-cinder21:29
*** zul has joined #openstack-cinder21:29
*** geguileo has joined #openstack-cinder21:29
*** sgotliv has joined #openstack-cinder21:32
*** alonma has quit IRC21:33
*** pratap has joined #openstack-cinder21:35
*** alonma has joined #openstack-cinder21:35
mc_nairxyang1: the extra check that you speak of is enforcing that max_over_subscription_ratio >= 1 for a driver config, correct?21:38
*** jgregor has quit IRC21:39
*** alonma has quit IRC21:39
*** edtubill has quit IRC21:39
*** zul has quit IRC21:40
*** thingee has quit IRC21:41
*** alonma has joined #openstack-cinder21:41
xyang1mc_nair: yes21:41
*** timcl1 has quit IRC21:41
*** alonma has quit IRC21:46
*** jistr has joined #openstack-cinder21:46
*** jistr has quit IRC21:47
*** thingee has joined #openstack-cinder21:49
*** edtubill has joined #openstack-cinder21:50
*** edtubill has quit IRC21:50
*** mylu has quit IRC21:52
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: WIP os-brick refactor get_connector_properties  https://review.openstack.org/27594321:52
*** mylu has joined #openstack-cinder21:53
*** salv-orlando has joined #openstack-cinder21:54
*** jgregor has joined #openstack-cinder21:54
hemnasmcginnis, ^^21:54
hemnahacking around in os-brick to see if doing that makes a bit more sense.21:54
hemnathere is a windows connector patch in gerrit now that's ugly as can be.   I'm hoping that my WIP can make it easier and cleaner to manage.21:55
*** mylu has quit IRC21:56
*** mylu has joined #openstack-cinder21:56
*** mylu has quit IRC21:56
*** salv-orl_ has quit IRC21:56
*** edtubill has joined #openstack-cinder21:57
*** crose has quit IRC21:59
*** haomaiwa_ has quit IRC22:01
*** porrua has quit IRC22:01
*** haomaiwang has joined #openstack-cinder22:01
mc_nairxyang1: ok.  And so if we had the same (min, max) behavior we have for IntOpt in FloatOpt (which we don't currently) we could just specify that in the base driver.py class22:02
*** alyson_ has quit IRC22:02
hemnaman check is backed up bad22:02
hemna30022:02
hemna:(22:02
mc_nairxyang1: but regardless, I think all that stuff of what to do with config < 1 is sort of a side discussion.  The point of what to do with = 1 is still a discussion to have.  From what I can tell the check should be >= 122:03
xyang1mc_nair: I am not saying no, but I want to make sure no one will submit another patch to change it back again:)22:04
xyang1mc_nair: need to ask lots of people to review it22:05
*** PsionTheory has joined #openstack-cinder22:05
mc_nairxyang1: haha ok.  That seems fair.  I agree about having lots of eyes on it before it merges.22:06
xyang1mc_nair: at least add everyone from the previous patch22:06
mc_nairxyang1: gotcha.  I can do that.  Will add those people to the review.  Anything else we should do on this?  You care if the config validation we were talking about is this patch vs. follow up?22:08
xyang1mc_nair: that can be a follow up22:08
xyang1mc_nair: are you the author of the patch? i thought it is someone else22:09
mc_nairxyang1: not me. Just helping them out as it's their first patch and it's getting more complex22:11
xyang1mc_nair: ok22:12
mc_nairxyang1: but I'll make sure the stuff we discussed gets done22:12
xyang1mc_nair: thanks22:12
mc_nairxyang1: thanks for the chat - lemme know if you think of anything else on this22:12
xyang1mc_nair: sure22:12
*** gouthamr has quit IRC22:16
*** krotscheck_dcm has quit IRC22:17
guitarzanthis is a patch to give you the ability to not overprovision lvm thin volumes?22:17
*** krotscheck1 has joined #openstack-cinder22:17
*** dims_ has joined #openstack-cinder22:17
*** mylu has joined #openstack-cinder22:18
*** sinese has quit IRC22:18
*** sgotliv has quit IRC22:21
*** dims has quit IRC22:21
dulekscottda: You were asking about quota bug. I don't think you've linked the correct one. Or maybe you're observing the bug on stable/liberty?22:24
*** baumann has quit IRC22:24
mc_nairguitarzan: this is related to https://bugs.launchpad.net/cinder/+bug/153716222:24
openstackLaunchpad bug 1537162 in Cinder "accounting bug for lvm with thin provisioning and max_over_subscription_ratio=1" [Undecided,In progress] - Assigned to Michael Rowden (mrrowden)22:24
scottdadulek: I've observed it on stable/liberty on our distro. I haven't attempted to repro on Devstack.22:24
guitarzanmc_nair: yeah, I'm kind of following22:25
guitarzanmc_nair: I just wanted to voice my support for this working correctly :)22:25
xyang1mc_nair: eharney has a patch to fix the lvm thin provisioning22:25
mc_nairguitarzan: ah gotcha22:25
xyang1mc_nair: I forgot the link22:25
guitarzanas in there are other reasons to want lvm thin other than over provisioning22:25
wNsnapshots dont suck?22:26
wNheh22:26
guitarzanmostly "don't have to scrub"22:26
openstackgerritTom Barron proposed openstack/cinder: mock time.sleep in Broadcom unit test  https://review.openstack.org/27595222:26
scottdadulek: It's fixed with the latest Keystone policy.json22:26
guitarzanbeing the one I care about22:26
mc_nairscottda: quotas? bugs? no way! which one was this?22:27
mc_nairI'm just trying to keep them on my radar...22:27
*** dims_ has quit IRC22:27
dulekscottda: Okay! Sorry for the late response. I probably should go and catch my planeā€¦22:27
scottdamc_nair: https://bugs.launchpad.net/cinder/+bug/151851322:27
openstackLaunchpad bug 1518513 in Cinder "Quota update call fails in Liberty" [Medium,In progress] - Assigned to Dan Nguyen (daniel-a-nguyen)22:27
xyang1mc_nair: https://review.openstack.org/#/c/26698622:27
*** jgregor has quit IRC22:27
scottdadulek: np. Thanks22:27
xyang1mc_nair: so lvm will use a different parameter22:28
xyang1mc_nair: this patch you are looking at is still valid22:28
xyang1mc_nair: lvm will have 1 as default instead of 2022:28
xyang1mc_nair: your followup patch should check eharney's new parameter22:30
mc_nairxyang1: Thanks for the link.  To me it seems like that patch makes it *way* more important we land the fix for the max_over_subscription_ratio then and land https://review.openstack.org/#/c/275408/22:31
mc_nairxyang1: because from what I'm getting from eharney's commit message, the intention for this is to make overprovisioning opt-in, but with the current bug in the capacity filter by setting this default to 1 we're still making it opt-in (because the bug causes this to allow over-provisioning)22:33
mc_nairbasically right now if we wanted over-provisioning to be opt-in in LVM with that other bug we'd need to set it to default to 1.000000122:34
xyang1mc_nair: ?22:35
xyang1mc_nair: what other bug22:35
xyang1mc_nair: what we discussed earlier should not be affected22:35
mc_nairxyang: https://bugs.launchpad.net/cinder/+bug/153716222:35
openstackLaunchpad bug 1537162 in Cinder "accounting bug for lvm with thin provisioning and max_over_subscription_ratio=1" [Undecided,In progress] - Assigned to Michael Rowden (mrrowden)22:35
mc_nairxyang1: what we said before won't be affected, but it will affect what eharney's change does22:36
*** alonma has joined #openstack-cinder22:36
mc_nairlemme step back for a minute, it's possibly between quotas and this my brain has just melted22:36
xyang1mc_nair: I don't see a conflict22:37
*** amoturi has left #openstack-cinder22:38
mc_nairxyang1: here I'm taking 5, then I'll try to articulate my concern :)22:38
xyang1mc_nair: if thin and >=1, use it22:39
xyang1mc_nair: if thin but invalid ratio, set it to 122:39
mc_nairxyang1: with current bug it's "if thin and > 1, use it"22:40
mc_nairhttps://github.com/openstack/cinder/blob/master/cinder/scheduler/filters/capacity_filter.py#L8922:40
xyang1i know, I am proposing a change22:41
*** alonma has quit IRC22:41
mc_nairxyang1: to what?  https://review.openstack.org/#/c/275408/ already does the change we need I think22:42
xyang1invalid ratio is not handled22:42
xyang1still not affected by eharney's change22:42
mc_nairxyang1: ok - it will be good to have the invalid ratio change22:43
*** alonma has joined #openstack-cinder22:43
mc_nairxyang1: but my point is that without https://review.openstack.org/#/c/275408/, I think eharney's change doesn't do what we expect22:44
xyang1mc_nair: he just changed default from 20 to 1 fir lvm22:45
mc_nairxyang1: because without that fix ^, max_subscription_ratio == 1 incorrectly allows for overprovisioning22:45
xyang1that was agreed at the summit22:45
mc_nairxyang1: what was agreed at the summit, changing from 20 to 1?22:45
xyang1mc_nair: for lvm22:46
*** laughterwym has joined #openstack-cinder22:46
eharneywell the whole intent of changing it to 1 is to not overprovision22:46
mc_nairxyang1: perfect, I'm not against that change at all.  The reason for that change is to disable overprovisioning by default with LVM, correct?22:46
mc_naireharney: cool22:46
*** alonma has quit IRC22:47
mc_nairand that will work well once https://review.openstack.org/#/c/275408/ is in, but before that we would need to set to 1.00001 (which I'm not advocating) to default to not overprovision22:47
eharneyand while i'm a little fuzzy on some of the details here, i suspect the folks saying that we need to fix the calculation still are correct22:48
mc_nairat least that's my understanding based on the code and https://bugs.launchpad.net/cinder/+bug/153716222:48
openstackLaunchpad bug 1537162 in Cinder "accounting bug for lvm with thin provisioning and max_over_subscription_ratio=1" [Undecided,In progress] - Assigned to Michael Rowden (mrrowden)22:48
*** krotscheck1 has quit IRC22:49
eharneyif the driver is doing thin provisioning, even when limited to a ratio of 1, the calculations need to understand to be correct, AFAICT22:49
eharneyneed to understand that*22:49
*** alonma has joined #openstack-cinder22:49
mc_naireharney: yea - I don't think it does anything to your patch.  Was just pointing out that until the fix for ^ lands, I don't think if turns off overprovisioning by default.  Think it may allow for no max_ratio on overprovisioning22:50
mc_nair*it turns off22:50
*** laughterwym has quit IRC22:50
eharneywhich is not desirable behavior for any driver22:51
*** mylu has quit IRC22:53
*** alonma has quit IRC22:54
*** mylu has joined #openstack-cinder22:54
mc_nairagreed.  So basically in short - we want the fix for https://review.openstack.org/#/c/275408/ to land because it helps all drivers.  We probably want that fix to land before https://review.openstack.org/#/c/266986/4, because otherwise that by default would expose the max_ratio == 1 bug22:55
mc_naireharney, xyang1: ^ does that make sense to you?22:55
eharneymc_nair: yeah, except for the part where we already merged the LVM change22:55
mc_naireharney: haha ok then, new plan :)22:56
mc_nairI can't read22:56
xyang1mc_nair: yes, we only need to worry about 275408 now22:56
mc_nairhaha ok, cool.  I missed that part of it, sorry about that22:57
mc_nairthat was why I was talking in circles :)22:57
mc_nairwell at least one of the reasons22:57
*** rlrossit_ has quit IRC22:58
*** jwcroppe has quit IRC22:59
eharneyi have to run for today but i'll look at this again tomorrow22:59
*** haomaiwang has quit IRC23:01
*** haomaiwa_ has joined #openstack-cinder23:01
openstackgerritMerged openstack/cinder: Imported Translations from Zanata  https://review.openstack.org/27552923:02
*** baojg has joined #openstack-cinder23:02
openstackgerritMerged openstack/cinder: EMC VMAX - Fix for last volume in VMAX3 storage group  https://review.openstack.org/24433123:05
*** erlon has quit IRC23:06
*** tpsilva has quit IRC23:06
*** baojg has quit IRC23:07
*** xyang1 has quit IRC23:08
*** eharney has quit IRC23:09
openstackgerritAnthony Lee proposed openstack/python-brick-cinderclient-ext: Query volume paths implementation  https://review.openstack.org/26840523:09
jgriffithpatrickeast: sorry... lost my network screwing around with new VPN23:18
*** arch-nemesis has quit IRC23:18
jgriffithpatrickeast: then have been busy mourning the loss of my beloved google for business :(23:18
* jgriffith wants to jump out a window23:18
jgriffithpatrickeast: that's a tricky problem... because I'm not sure how we "guarantee" the message is sent to all "faked" backends23:19
*** mylu has quit IRC23:21
jgriffithpatrickeast: TBF though, that's not unique to our proposal here23:21
*** gouthamr has joined #openstack-cinder23:21
jgriffithpatrickeast: only thing I came up with is a periodic when things are "failed over" that continuously sends an update to the driver... but that's race prone; still may not work and is a whole can of worms23:22
*** gouthamr_ has joined #openstack-cinder23:23
*** e0ne has joined #openstack-cinder23:24
*** mylu has joined #openstack-cinder23:25
patrickeastjgriffith: yea i haven't figure out a better solution either23:26
*** gouthamr has quit IRC23:26
hemnatrying to wrap my head around cheesecake23:26
*** mylu has quit IRC23:26
patrickeastjgriffith: i'm almost leaning towards a 'if it breaks just to restart your other c-vol services' documentation punt23:26
patrickeasts/just to/just go/23:26
jgriffithpatrickeast: yeah, it's an interesting case23:27
patrickeasthemna: sounds messy23:27
hemnayah it is23:27
jgriffithLOL23:27
hemnaI'm just trying to figure how this works for us at all23:27
hemna:(23:27
jgriffithhemna: what are you having trouble with?23:27
jgriffithhemna: it's actually super easy compared to volume-level rep23:27
hemnaheh23:27
hemnayah, it's supposed to be23:28
*** geguileo has quit IRC23:28
hemnaand I see the API23:28
hemnait looks easier23:28
jgriffithhemna: LOL.. well, can I help?23:28
hemnaI'm trying to figure out what to do on our backend for enable_replication23:28
jgriffithhemna: Ahh... ok, so23:28
hemnaI'm not sure how to find which volumes to turn on at that point on our array.23:28
jgriffithhemna: there's only one real case for that IMHO23:28
hemnaespecially, since our array can have multiple openstack's pointing to it. :(23:28
jgriffithhemna: your backend should setup replication based on cinder.conf at init time23:29
jgriffithhemna: BUT, disable-rep gives an opportunity for maintenance or whatever if needed23:29
patrickeastjgriffith: oh speaking of the api, the one right now in the WIP is still on the volume manager, does that mean its still like a /v2/volumes/<uuid>/action kind of thing?23:29
jgriffithhemna: enable_replication just says... "turn that shit back on"23:29
jgriffithpatrickeast: yeah, I'm hoping to just keep the old methods/api's where possible and modify their parameters23:30
jgriffithpatrickeast: I don't see any value in yet anoter set of things23:30
jgriffithOhhhh23:30
jgriffithpatrickeast: No!23:30
patrickeastjgriffith: mm so the api still needs to be called on some volume?23:30
jgriffithsorry23:30
jgriffithpatrickeast: so... that's all "host/c-vol" based now.23:30
jgriffithpatrickeast: it all gets sucked in under services api23:30
jgriffithhemna: does that help at all?23:30
hemnachewing....23:30
hemnait may..23:31
*** dustins has joined #openstack-cinder23:31
patrickeastjgriffith: yep yep, ok cool23:31
hemnaI'm also trying to unravel what we currently have23:31
jgriffithpatrickeast: pheww... sorry23:31
patrickeasthemna: can't you just do what you had before on the per-volume basis, but for alll the volumes?23:31
jgriffithfrik.. I have to go23:31
hemnakinda23:31
patrickeastjgriffith: oh i see part of the confusion, i meant the volume controller, not the manager23:32
patrickeastthe manager ones all are good the way they are in the WIP23:32
jgriffithpatrickeast: yeah, I figured that out :)23:32
jgriffithpatrickeast: so yeah... I would move the API calls to contrib/service23:32
patrickeastjgriffith: gotcha23:33
*** e0ne has quit IRC23:33
*** edtubill has quit IRC23:33
hemnajgriffith, ok I think we have a way to make this work23:35
jgriffithhemna: honestly I was just looking at your code, it should be more code-removal than anything else23:36
hemnawe do so much work/effort to setup each volume replication23:36
jgriffithhemna: you already have all the mechanisms to do the *things*, we just remove some of the crazy options we had23:36
hemnaI think we can just force the admin to create the replication group name before hand, stuff it into the volume type and we are done.23:37
jgriffithanyway.. don't hesitate to ping me on questions.  We can all figure this out together.  like patrickeast 's case23:37
hemnainstead of us dynamically creating the group on every volume.23:37
jgriffithhemna: hmmm23:37
jgriffithhemna: not sure I follow that23:37
jgriffithhemna: so why not just have it in the cinder.conf... then the driver sets up the group in init23:37
jgriffithhemna: and if a type=replicated comes in it goes to said rep-group?23:37
hemnayah either way23:38
patrickeastthats what ours will do23:38
jgriffithhemna: well... don't take this the wrong way... but one way is "wrong" and the other "is not"23:38
jgriffith:)23:38
hemna:P23:38
jgriffithhemna: better to do it the "right" way :)23:38
jgriffithhemna: otherwise we might as well just stick with what we have23:38
hemnaok23:39
hemnafor some reason we were spinning on the need for getting the list of volumes from the manager at enable time23:39
hemnameh23:39
*** geguileo has joined #openstack-cinder23:41
patrickeastdo we have a way to query the db for all the replicated volumes for that backend? may be pretty easy to pass em in to the failover/enable/disable commands23:41
jgriffithhemna: nope, not needed23:41
patrickeastif you guys end up needed them23:41
jgriffithpatrickeast: yeah, they'll have the type associated with them23:41
hemnaif we know the group name, we can find them and enable/disable23:42
jgriffithpatrickeast: get-all-volume of type xxxx23:42
hemnaI think we're good23:42
jgriffithhemna: hmm... you're scaring me23:42
patrickeastlol23:42
hemnahehe23:42
patrickeastjgriffith: yea that should work, although there could be multiple replicated types, right?23:42
*** pratap has quit IRC23:43
jgriffithpatrickeast: yeah, but you could filter on backend-name etc23:43
patrickeastjgriffith: need like a join for type has replication and get all the volumes of that type on the backend23:43
jgriffithbut....23:43
jgriffithhonestly the disable/enable thing is kinda... well... weird23:43
jgriffithIn my case, I'm not planning to actually disable anything23:43
jgriffithI don't see any value in it23:43
patrickeastyea, i dont need it for my backend :D23:44
jgriffithif you want to stop replcating volumes, then they should be retyped23:44
jgriffithand then to start again... retype again23:44
jgriffithbut if the admin wants/needs some magic tool... well, I guess that's "ok"23:44
jgriffithpatrickeast: hemna maybe enable/disable doesn't even make sense?23:44
*** mylu has joined #openstack-cinder23:44
patrickeasti guess the only thing there is a user may not know why their volume got retyped, right?23:44
patrickeastjgriffith: dunno, i mean... for our simple use case it isn't really needed23:45
hemnaif you lose a switch between the 2 arrays and want to pause it in the mean time ?23:45
hemnadunno, thinking outloud23:45
patrickeastjgriffith: but i can see where maybe they want to turn it off while the dr site is under maintenance or something23:45
jgriffithhemna: sure, but what's it going to matter?23:45
patrickeasteven then i'm not sure it matters23:45
jgriffithhemna: I mean, it's not working anyway23:45
hemnaI'm not sure what our array will do if the target takes a dump in the process23:45
hemnatbh23:45
jgriffithso it's just semantics/hoops to jump through23:46
jgriffithhemna: yeah, but my point is it's not going to matter23:46
hemnaI'm not a replication expert hehe23:46
jgriffithhemna: I mean... say your switch or the remote target dumps23:46
jgriffithhemna: until the admin says "disable rep" it doesn't seem like it would make any difference... or offer anything that your driver couldn't figure out on it's own23:46
jgriffithI'm kinda leaning towards those calls being useless honestly23:47
*** yhayashi has joined #openstack-cinder23:47
patrickeastso... i could reasonably see something like they wanted to disable replication to one of the remote sites or something temporarily (in the multi-target scenario)23:47
*** jamielennox|away is now known as jamielennox23:47
patrickeastwhich could then make the driver not try and pick that one in failover23:47
hemnasure23:47
patrickeasteven if it could reach it23:47
patrickeastbuuut its still kind of a stretch imo23:47
jgriffithpatrickeast: yeah, that's feasible.  Although the impl I have doesn't allow that anyway :)  It's all or nothing23:48
*** mylu has quit IRC23:48
*** mylu has joined #openstack-cinder23:48
patrickeastjgriffith: not for that it isn't, right? it only fails-over to one of the targets and we return which one23:48
patrickeastjgriffith: so it would be like a short-cut to temporarily disable a target instead of changing the cinder.conf/restarting c-vol23:48
jgriffithpatrickeast: well, what I mean is that the enable/disable doesn't take any parameters23:48
patrickeastoooh yea23:49
jgriffithpatrickeast: it just tells the "main" backend... "stop/start replicating"23:49
jgriffiththat's all23:49
jgriffithall or nothing23:49
hemnaso it's not multi target then ?23:49
patrickeastyea, got it23:49
jgriffithhemna: sure it is23:49
*** edmondsw has quit IRC23:49
jgriffithor can be23:49
hemnaor it's all on for all targets23:49
hemnaor all off for all targets23:49
jgriffithhemna: exactly23:49
hemnaah k23:49
hemnaI suppose that's fine for now23:49
jgriffithhemna: replication is either enabled/possible or it's not23:49
patrickeasti guess in the future it could be expanded to be more specific if folks want that23:49
hemnasimplify this shit23:49
jgriffithhemna: for now?23:50
hemna:)23:50
*** alonma has joined #openstack-cinder23:50
hemnapatrickeast, yah that's what I'm thinking23:50
jgriffithhemna: honestly I think it's a wayy better answer than what we were trying to do23:50
hemnaall or nothing23:50
hemnawell it is simpler23:50
hemnaand I think that's a good thing23:50
patrickeasthemna: +123:50
jgriffithand it makes sense in terms of use-cases IMHO23:50
hemnaor it's impossible to make 80 drivers happy23:50
patrickeastjgriffith: so maybe back to your point earlier, would it make more sense to just not have enable/disable?23:51
jgriffithpatrickeast: honestly I think it might23:51
patrickeastsince we've just kind of said we dont have a good use case for it23:51
jgriffithpatrickeast: It's sort of weird now that I think about it23:52
jgriffithpatrickeast: I had thought some sort of maintenance mode... but an admin is going to have to be doing something on the backend for that anyway23:52
jgriffithso maybe it doesn't make sense right now23:52
patrickeastjgriffith: im on board to just pull it, not sure it really adds value with the way things would work now23:52
jgriffithpatrickeast: I think I'll remove it... maybe ponder it a bit more on the drive home23:53
jgriffithbut I don't see much value in it23:53
patrickeastlater on if we do have more specific stuff and we wanted to make a more flexible maintenance mode sort of deal it would make more sense to add it back in23:53
hemnawell, it doesn't harm anything23:53
hemnasince it's a simple passthrough to the driver23:53
jgriffithhemna: yeah... I suppose but then we get into that problem of "what is expected"23:53
hemnakinda23:53
jgriffithhemna: if you put it in there then you would expect that the backend is "disabling replication"23:54
hemnait's up to the driver....:P23:54
jgriffithso then you kinda have to do that23:54
jgriffithhemna: oh hush :)23:54
patrickeastlol23:54
hemna:)23:54
*** alonma has quit IRC23:54
*** cknight has quit IRC23:57

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