Monday, 2015-08-31

*** haomaiwang has quit IRC00:35
*** takedakn has quit IRC00:42
*** bswartz has joined #openstack-cinder00:52
openstackgerritEli Qiao proposed openstack/cinder: Handler KeyManager exception when deleting a volume
*** chlong has joined #openstack-cinder00:55
*** sparr has quit IRC01:07
*** tbarron has quit IRC01:07
*** salv-orlando has joined #openstack-cinder01:07
openstackgerritTeruaki Ishizaki proposed openstack/cinder: Sheepdog: Improve image operations
*** salv-orlando has quit IRC01:13
*** salv-orlando has joined #openstack-cinder01:13
*** simondodsley has quit IRC01:16
*** EinstCrazy has joined #openstack-cinder01:19
*** zhangjn has joined #openstack-cinder01:20
*** zhenguo has joined #openstack-cinder01:24
*** salv-orlando has quit IRC01:26
*** tobasco_ has joined #openstack-cinder01:36
*** tobasco has quit IRC01:37
*** kambiz has quit IRC01:37
*** takedakn has joined #openstack-cinder01:39
*** zhangjn has quit IRC01:40
*** haomaiwang has joined #openstack-cinder01:40
*** zhangjn has joined #openstack-cinder01:41
*** aix_ has joined #openstack-cinder01:41
*** vilobhmm1 has joined #openstack-cinder01:42
*** aix has quit IRC01:42
*** vilobhmm11 has joined #openstack-cinder01:43
*** kambiz has joined #openstack-cinder01:43
*** vilobhmm1 has quit IRC01:47
*** chlong has quit IRC01:49
*** Lee1092 has joined #openstack-cinder01:50
openstackgerritPeter Wang proposed openstack/cinder: Update name_id and provider_location for migration
openstackgerritPeter Wang proposed openstack/cinder: Update name_id and provider_location for migration
*** takedakn has quit IRC02:12
*** zhangjn has quit IRC02:19
*** tbarron has joined #openstack-cinder02:20
*** sparr has joined #openstack-cinder02:21
*** hw__chenying has joined #openstack-cinder02:23
*** salv-orlando has joined #openstack-cinder02:28
openstackgerritTina Tang proposed openstack/cinder: Clone cg support in VNX driver
openstackgerritVincent Hou proposed openstack/cinder: Use cinder internal tenant to create the destination volume
*** salv-orlando has quit IRC02:41
*** IlyaG has joined #openstack-cinder02:43
*** hakimo_ has joined #openstack-cinder02:52
*** hakimo has quit IRC02:54
openstackgerritVincent Hou proposed openstack/python-cinderclient: Volume status management for volume migration
*** EinstCrazy has quit IRC02:57
openstackgerritVincent Hou proposed openstack/python-cinderclient: Volume status management for volume migration
*** vincent_hou has joined #openstack-cinder02:58
vincent_houIs there any Peter Wang here?02:59
*** lixiaoy1 has joined #openstack-cinder02:59
*** wilson1 has quit IRC02:59
vincent_houAnyone knows his IRC name?02:59
openstackgerritPatrick East proposed openstack/cinder: Generic image-volume cache
*** haomaiwang has quit IRC03:03
*** haomaiw__ has joined #openstack-cinder03:03
*** IlyaG has quit IRC03:23
*** vilobhmm11 has quit IRC03:26
*** vilobhmm1 has joined #openstack-cinder03:27
*** I has joined #openstack-cinder03:35
*** I is now known as Guest2505903:36
*** longguang has joined #openstack-cinder03:36
*** links has joined #openstack-cinder03:40
*** davechen has joined #openstack-cinder03:41
*** salv-orlando has joined #openstack-cinder03:43
*** jwcroppe_ has joined #openstack-cinder03:45
*** jwcroppe has quit IRC03:47
*** Guest25059 has quit IRC03:49
*** haigang has joined #openstack-cinder03:49
*** samwan has joined #openstack-cinder03:49
*** vilobhmm1 has quit IRC03:50
*** vilobhmm1 has joined #openstack-cinder03:51
*** takedakn has joined #openstack-cinder03:53
*** garthb has joined #openstack-cinder03:55
*** salv-orlando has quit IRC03:56
openstackgerritThang Pham proposed openstack/cinder: Conversion to volume object
openstackgerritThang Pham proposed openstack/cinder: Sync volume versionedobject to ORM
*** vilobhmm1 has quit IRC04:02
*** vilobhmm1 has joined #openstack-cinder04:04
openstackgerritVincent Hou proposed openstack/python-cinderclient: Volume status management for volume migration
*** takedakn has quit IRC04:04
openstackgerritThang Pham proposed openstack/cinder: Register RPC and object versions
*** hw__chenying has quit IRC04:09
*** subhadeep_de has quit IRC04:21
*** subhadeep_de has joined #openstack-cinder04:22
*** sgotliv has joined #openstack-cinder04:29
longguangdoes cinder implement chain-snapshot, but  images (stored in glance)  when create snapshot?04:29
openstackgerritLei Li proposed openstack/cinder: Adjust to right exception for chunkeddriver
*** zhangjn has joined #openstack-cinder04:37
*** shyama has joined #openstack-cinder04:41
*** sgotliv has quit IRC04:43
*** takedakn has joined #openstack-cinder04:47
*** deepakcs has joined #openstack-cinder04:49
*** salv-orlando has joined #openstack-cinder04:52
*** theanalyst has quit IRC04:54
*** theanalyst has joined #openstack-cinder04:56
*** salv-orlando has quit IRC04:57
*** salv-orlando has joined #openstack-cinder04:57
*** ig0r__ has joined #openstack-cinder04:59
*** ig0r_ has quit IRC04:59
*** shausy has joined #openstack-cinder05:01
*** samwan is now known as sam_wan05:06
*** salv-orlando has quit IRC05:10
*** takedakn has quit IRC05:11
*** sam_wan is now known as sam_wan_05:12
*** vincent_hou_ has joined #openstack-cinder05:13
*** sam_wan_ is now known as wanshan05:14
*** wanshan is now known as wan_shan05:14
*** vincent_hou has quit IRC05:14
*** vincent_hou_ is now known as vincent_hou05:15
openstackgerritLisaLi proposed openstack/cinder: Minor optimization
*** wan_shan has quit IRC05:17
*** takedakn has joined #openstack-cinder05:18
*** haigang has quit IRC05:23
*** takedakn has quit IRC05:32
openstackgerritXinXiaohui proposed openstack/cinder: Calculate virtual free capacity and notify
*** garthb has quit IRC05:35
*** shausy has quit IRC05:36
*** shausy has joined #openstack-cinder05:36
*** haigang has joined #openstack-cinder05:47
*** markvoelker has joined #openstack-cinder05:51
*** markvoelker_ has joined #openstack-cinder05:53
*** yrabl has joined #openstack-cinder05:56
*** markvoelker has quit IRC05:56
*** sam_wan has joined #openstack-cinder05:57
*** hw__chenying has joined #openstack-cinder06:01
*** ankit_ag has joined #openstack-cinder06:06
*** salv-orlando has joined #openstack-cinder06:10
*** vilobhmm1 has quit IRC06:16
*** vilobhmm1 has joined #openstack-cinder06:18
openstackgerritLisaLi proposed openstack/cinder: Raise HTTPBadRequest for invalid filters
*** ishant has joined #openstack-cinder06:18
*** afazekas_ has joined #openstack-cinder06:19
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Transifex
*** markvoelker has joined #openstack-cinder06:22
*** vincent_hou has quit IRC06:23
*** vincent_hou has joined #openstack-cinder06:25
*** markvoelker_ has quit IRC06:25
*** markvoelker_ has joined #openstack-cinder06:27
*** haigang has quit IRC06:27
*** haigang has joined #openstack-cinder06:28
*** markvoelker has quit IRC06:29
*** haigang has quit IRC06:29
*** vilobhmm1 has quit IRC06:30
*** anshul has joined #openstack-cinder06:34
*** haigang has joined #openstack-cinder06:36
openstackgerritzhaohua proposed openstack/cinder: Huawei: fix multi REST-URLs bug
*** zhangjn has quit IRC06:45
*** zhangjn has joined #openstack-cinder06:45
*** stefan_amann has joined #openstack-cinder07:03
*** hw__chenying has quit IRC07:24
*** hw__chenying has joined #openstack-cinder07:24
*** marcusvrn_ has quit IRC07:25
*** ronis has joined #openstack-cinder07:25
openstackgerritVincent Hou proposed openstack/cinder: LVM: add the exception handling to volume copy
*** alexschm has joined #openstack-cinder07:28
*** zhangjn has quit IRC07:28
*** kaisers has joined #openstack-cinder07:33
*** takedakn has joined #openstack-cinder07:37
*** baojg has joined #openstack-cinder07:40
*** takedakn has quit IRC07:47
*** sgotliv has joined #openstack-cinder07:47
*** ndipanov has joined #openstack-cinder07:47
*** takedakn has joined #openstack-cinder07:47
*** yrabl has quit IRC07:48
*** hw__chenying has quit IRC07:48
*** takedakn has quit IRC07:49
*** yrabl has joined #openstack-cinder07:50
*** shyama has quit IRC07:50
*** hw__chenying has joined #openstack-cinder07:52
*** ircuser-1 has joined #openstack-cinder08:08
*** zhangjn has joined #openstack-cinder08:10
*** jistr has joined #openstack-cinder08:15
openstackgerritVincent Hou proposed openstack/cinder: LVM: add the exception handling to volume copy
*** lpetrut has joined #openstack-cinder08:23
*** salv-orlando has quit IRC08:24
*** salv-orlando has joined #openstack-cinder08:25
*** shyama has joined #openstack-cinder08:28
*** hw__chenying has quit IRC08:28
*** markvoelker_ has quit IRC08:28
*** lpetrut has quit IRC08:29
*** zhonghua-lee has quit IRC08:30
*** lpetrut has joined #openstack-cinder08:30
*** aix_ has quit IRC08:32
*** aix has joined #openstack-cinder08:33
*** haigang has quit IRC08:44
*** haigang has joined #openstack-cinder08:44
*** zhangjn has quit IRC08:45
*** haigang has quit IRC08:50
*** haigang has joined #openstack-cinder08:52
*** e0ne has joined #openstack-cinder08:58
vincent_houI have logged one bug This is something I am not quite certain. Need folks to verify it.09:04
openstackLaunchpad bug 1490445 in Cinder "Migration: Unnecessary destination volume check during delete_volume" [Undecided,New]09:04
*** IanGovett has joined #openstack-cinder09:08
*** e0ne has quit IRC09:09
*** flip214 has joined #openstack-cinder09:09
*** e0ne has joined #openstack-cinder09:09
*** bluex-pl has joined #openstack-cinder09:10
*** bluex-pl has quit IRC09:14
*** bluex-pl has joined #openstack-cinder09:14
*** haigang has quit IRC09:20
*** zhangjn has joined #openstack-cinder09:23
*** baojg has quit IRC09:23
*** markvoelker has joined #openstack-cinder09:23
*** salv-orlando has quit IRC09:24
*** markvoelker_ has joined #openstack-cinder09:26
*** zhangjn has quit IRC09:27
openstackgerritVincent Hou proposed openstack/cinder: Remove the unnecassary volume_api.get(context, volume_id)
*** markvoelker has quit IRC09:30
*** vincent_hou has quit IRC09:31
*** markvoelker_ has quit IRC09:31
*** haigang has joined #openstack-cinder09:33
*** ig0r_ has joined #openstack-cinder09:37
*** markvoelker has joined #openstack-cinder09:37
*** baojg has joined #openstack-cinder09:38
*** ig0r__ has quit IRC09:39
*** zhangjn has joined #openstack-cinder09:39
*** zhangjn has quit IRC09:41
*** shyama has quit IRC09:48
*** aix has quit IRC09:49
*** amann has joined #openstack-cinder09:49
*** s_amann has joined #openstack-cinder09:49
*** amann has quit IRC09:49
*** stefan_amann has quit IRC09:52
*** lixiaoy1 has quit IRC09:53
*** davechen has left #openstack-cinder09:54
*** baojg has quit IRC10:00
*** takedakn has joined #openstack-cinder10:11
*** takedakn has quit IRC10:14
*** haigang has quit IRC10:14
*** takedakn has joined #openstack-cinder10:14
*** haigang has joined #openstack-cinder10:15
*** takedakn has quit IRC10:17
*** takedakn has joined #openstack-cinder10:18
*** aix has joined #openstack-cinder10:18
*** bluex-pl has quit IRC10:34
*** sam_wan has quit IRC10:41
*** DTadrzak has joined #openstack-cinder10:44
*** bluex-pl has joined #openstack-cinder10:49
*** ganso_ has joined #openstack-cinder10:57
*** jaypipes has joined #openstack-cinder10:58
*** takedakn has quit IRC11:01
*** bluex-pl has quit IRC11:03
*** bluex-pl has joined #openstack-cinder11:03
*** bluex-pl has quit IRC11:09
*** e0ne has quit IRC11:10
*** takedakn has joined #openstack-cinder11:11
*** takedakn has quit IRC11:13
*** takedakn has joined #openstack-cinder11:13
*** haomaiw__ has quit IRC11:15
*** ishant has quit IRC11:25
*** anshul has quit IRC11:29
*** anshul has joined #openstack-cinder11:30
*** sgotliv has quit IRC11:31
*** jaypipes has quit IRC11:35
*** jaypipes has joined #openstack-cinder11:36
*** openstackgerrit has quit IRC11:46
*** openstackgerrit has joined #openstack-cinder11:47
*** sgotliv has joined #openstack-cinder11:48
*** markvoelker_ has joined #openstack-cinder11:51
*** markvoelker has quit IRC11:51
*** markvoelker_ has quit IRC11:55
*** markvoelker has joined #openstack-cinder11:56
*** e0ne has joined #openstack-cinder11:59
openstackgerritDaniel Tadrzak proposed openstack/cinder: Service object
*** dtadrzak_ has joined #openstack-cinder12:03
*** cristicalin has joined #openstack-cinder12:03
openstackgerritVincent Hou proposed openstack/cinder: LVM: add the exception handling to volume copy
openstackgerritDaniel Tadrzak proposed openstack/cinder: CGSnapshot Object
*** IanGovett1 has joined #openstack-cinder12:05
openstackgerritDaniel Tadrzak proposed openstack/cinder: Cleanup for cinder tests with CGSnapshot
*** IanGovett has quit IRC12:05
cristicalinanybody have a good example of python-cinderclient code using connection pooling or the session ?12:07
cristicalinI need to write something that fetches the quotas for tenants and cinder is particularly crappy in performance because I can't figure out how to tell it to reuse the keystone auth12:08
cristicalinit works for nova an neutron, nova has a connection_pool=True you can set in the Client class12:08
cristicalinbut no luck for Cinder12:08
cristicalinany help would be greatly appreciated12:09
*** sgotliv has quit IRC12:09
*** bluex-pl has joined #openstack-cinder12:09
*** Yogi1 has joined #openstack-cinder12:11
* DuncanT finally gets back into the swing of reviewing things12:12
*** raildo-afk is now known as raildo12:16
*** theanalyst has quit IRC12:18
*** theanalyst has joined #openstack-cinder12:19
*** akerr has joined #openstack-cinder12:20
DuncanTcristicalin: The only way I can see to do it is to get a token manually (e.g. keystone token-get) and pass that into cinder-client12:22
*** simondodsley has joined #openstack-cinder12:31
*** edmondsw has joined #openstack-cinder12:32
*** krtaylor has quit IRC12:34
*** haomaiwa_ has joined #openstack-cinder12:37
*** timcl has joined #openstack-cinder12:38
*** takedakn has quit IRC12:41
*** takedakn has joined #openstack-cinder12:42
*** diablo_rojo has joined #openstack-cinder12:42
*** smoriya has quit IRC12:42
*** annegentle has joined #openstack-cinder12:44
openstackgerritDaniel Tadrzak proposed openstack/cinder: Cleanup for SnapshotObject
*** sgotliv has joined #openstack-cinder12:47
*** krtaylor has joined #openstack-cinder12:47
*** afazekas__ has joined #openstack-cinder12:56
*** sgotliv has quit IRC12:57
*** cbader has joined #openstack-cinder12:58
*** afazekas_ has quit IRC12:58
*** merooney has joined #openstack-cinder13:02
*** edmondsw has quit IRC13:04
*** dustins has joined #openstack-cinder13:04
*** anshul has quit IRC13:08
*** bluex-pl has quit IRC13:13
*** sgotliv has joined #openstack-cinder13:14
*** xyang1 has joined #openstack-cinder13:16
*** yrabl has quit IRC13:17
dhellmannhey, folks, who's filling in for thingee for the L3 checkup this week?13:18
*** ankit_ag has quit IRC13:19
*** kbyrne has quit IRC13:19
dhellmannDuncanT: ^^ ?13:20
*** anshul has joined #openstack-cinder13:20
*** e0ne has quit IRC13:22
scottdadhellmann: maybe jgriffith13:23
DuncanTdhellmann: jgriffith I believe13:23
simondodsleylooks like the cores are going to be busy today - a lot of things out there waiting to get into L-313:25
*** nkrinner has quit IRC13:27
*** eharney has joined #openstack-cinder13:27
*** anshul has quit IRC13:28
*** e0ne has joined #openstack-cinder13:29
*** mtanino has joined #openstack-cinder13:30
*** links has quit IRC13:30
*** kbyrne has joined #openstack-cinder13:32
*** edmondsw has joined #openstack-cinder13:33
*** IanGovett1 has quit IRC13:35
*** superdan is now known as dansmith13:35
*** salv-orlando has joined #openstack-cinder13:36
*** dims has joined #openstack-cinder13:36
*** rushil has joined #openstack-cinder13:36
geguileoe0ne: ping - file I/O volume migration patch13:37
*** rushil_ has joined #openstack-cinder13:37
*** thangp has joined #openstack-cinder13:39
e0negeguileo: hi13:39
*** annasort has joined #openstack-cinder13:39
geguileoe0ne: I saw you +2 that patch but didn't +1 WF, any specific reason?13:39
e0negeguileo: i just wanted to be sure that it's everything ok with infra13:40
*** anshul has joined #openstack-cinder13:40
*** rushil has quit IRC13:41
e0negeguileo: afaik, a lot of patches were affected with new netaddr package13:41
e0negeguileo: may be we need to +A on it and recheck if CI will failed13:41
geguileoe0ne: So you think this may fail CI when we try to merge it?13:42
e0negeguileo: i'm not sure. but it's not related to the pacht13:42
geguileoe0ne: But if it fails it just won't merge  ;-)13:43
e0negeguileo: you're right. just put my +A on it. thanks:)13:43
geguileoe0ne: Ok13:43
openstackgerritVincent Hou proposed openstack/cinder: Remove the unnecassary volume_api.get(context, volume_id)
*** yuriy_n17 has joined #openstack-cinder13:45
*** esker has quit IRC13:47
*** esker has joined #openstack-cinder13:48
*** bill_az has joined #openstack-cinder13:49
openstackgerritwanghao proposed openstack/cinder: Add pagination to backups
*** annegentle has quit IRC13:49
*** breitz has quit IRC13:50
*** breitz has joined #openstack-cinder13:50
*** amoturi has joined #openstack-cinder13:51
*** jungleboyj has quit IRC13:54
*** agarciam has joined #openstack-cinder13:57
*** anshul has quit IRC13:58
*** zhangjn has joined #openstack-cinder13:59
*** zhangjn has quit IRC13:59
*** mriedem has joined #openstack-cinder13:59
*** zhangjn has joined #openstack-cinder14:00
*** IanGovett has joined #openstack-cinder14:01
*** deepakcs has quit IRC14:03
*** bnemec has joined #openstack-cinder14:05
*** afazekas__ has quit IRC14:08
*** zhangjn has quit IRC14:09
openstackgerritwanghao proposed openstack/cinder: Add updated_at into response of listing detail
*** markvoelker has quit IRC14:09
*** edtubill has joined #openstack-cinder14:11
*** eharney has quit IRC14:11
*** boris-42 has joined #openstack-cinder14:11
*** timcl has quit IRC14:11
*** anshul has joined #openstack-cinder14:11
*** rushil_ has quit IRC14:12
*** sidbhatt13 has joined #openstack-cinder14:13
*** sidbhatt13 has quit IRC14:15
*** zhangjn has joined #openstack-cinder14:17
*** shausy has quit IRC14:17
*** peter-wang has joined #openstack-cinder14:18
*** yrabl has joined #openstack-cinder14:19
*** peter-wang has left #openstack-cinder14:19
*** annegentle has joined #openstack-cinder14:19
*** zhangjn has quit IRC14:21
*** r-daneel has joined #openstack-cinder14:22
*** peter-wang has joined #openstack-cinder14:22
openstackgerritJoel Coffman proposed openstack/cinder: Clean up volume_type_update method
*** timcl has joined #openstack-cinder14:24
openstackgerritTom Barron proposed openstack/cinder: Scheduler-based over-subscription for NFS drivers
*** lpetrut has quit IRC14:32
*** sidbhatt13 has joined #openstack-cinder14:33
*** mriedem is now known as mriedem_away14:39
*** andreykurilin__ has joined #openstack-cinder14:41
*** lpetrut has joined #openstack-cinder14:43
openstackgerritDaniel Tadrzak proposed openstack/cinder: Cleanup for cinder tests with CGSnapshot
*** garthb has joined #openstack-cinder14:47
*** jkremer has joined #openstack-cinder14:48
*** eharney has joined #openstack-cinder14:48
*** zhangjn has joined #openstack-cinder14:49
jgriffithmriedem_away: on patch
jgriffithmriedem_away: I was wondering if you'd mind terribly moving that to the existing log statement here:
*** dims has quit IRC14:51
*** IlyaG has joined #openstack-cinder14:53
*** markvoelker has joined #openstack-cinder14:53
*** gouthamr has joined #openstack-cinder14:54
*** markvoelker has quit IRC14:54
*** markvoelker has joined #openstack-cinder14:55
*** crose has joined #openstack-cinder14:55
openstackgerritDaniel Tadrzak proposed openstack/cinder: Cleanup for SnapshotObject
*** jistr is now known as jistr|call14:57
*** cristicalin has quit IRC14:58
*** rhagarty__ has joined #openstack-cinder14:58
*** zhipeng has joined #openstack-cinder14:58
openstackgerritTom Barron proposed openstack/cinder: NetApp DOT block driver over-subscription support
*** lcurtis has joined #openstack-cinder15:00
*** rhagarty__ has quit IRC15:02
*** s_amann has quit IRC15:02
*** dims has joined #openstack-cinder15:03
*** s_amann has joined #openstack-cinder15:03
amoturiCould someone from the core team please review this patchset
*** lyanchih has joined #openstack-cinder15:04
*** hemnafk has quit IRC15:04
*** bnemec has quit IRC15:04
*** rhagarty__ has joined #openstack-cinder15:08
openstackgerritDaniel Tadrzak proposed openstack/cinder: Cleanup for cinder tests with CGSnapshot
*** bnemec has joined #openstack-cinder15:09
lyanchihjaypipes: Hi, recently I report a bug . It is talking about swap volume will stay in incorrect status. When I try to fix, I found nova expect cinder’s migration_volume_complete api should attach and detach volume. But cinder will only return new volume’s id. I have no idea about such situation. Nova’s member said I can come here to discuss.15:11
openstackLaunchpad bug 1489744 in OpenStack Compute (nova) "Swapping volume will not result in correct status" [Undecided,New] - Assigned to Chung Chih, Hung (lyanchih)15:11
openstackgerritOleksii Butenko proposed openstack/python-cinderclient: Add functional tests for python-cinderclient
*** hodos has joined #openstack-cinder15:18
*** peter-wang has quit IRC15:20
*** ronis has quit IRC15:21
*** jaypipes has quit IRC15:26
*** agarciam has quit IRC15:28
*** haomaiwa_ has quit IRC15:29
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** david-ly_ has joined #openstack-cinder15:31
*** david-ly_ is now known as david-lyle_15:33
*** harlowja_at_home has joined #openstack-cinder15:33
*** david-lyle has quit IRC15:33
openstackgerritEric Harney proposed openstack/cinder: Remove executable bits on files
*** david-lyle_ is now known as david-lyle15:35
*** anshul has quit IRC15:35
*** crose has quit IRC15:35
*** jungleboyj has joined #openstack-cinder15:36
*** lyanchih has quit IRC15:38
*** rhagarty__ has quit IRC15:43
*** IanGovett1 has joined #openstack-cinder15:43
*** IanGovett has quit IRC15:43
openstackgerritEric Harney proposed openstack/cinder: Detect addition of executable files
*** rushil has joined #openstack-cinder15:48
*** shyama has joined #openstack-cinder15:49
*** jistr|call is now known as jistr15:53
*** zhangjn has quit IRC15:54
*** sgotliv has quit IRC15:55
*** diegows has joined #openstack-cinder15:58
*** annegentle has quit IRC15:59
*** markvoelker_ has joined #openstack-cinder16:00
*** markvoel_ has joined #openstack-cinder16:01
*** timcl has quit IRC16:03
*** Apoorva has joined #openstack-cinder16:03
*** markvoelker has quit IRC16:03
*** dannywilson has joined #openstack-cinder16:04
*** sayali has quit IRC16:04
*** IanGovett1 has quit IRC16:05
*** timcl has joined #openstack-cinder16:05
*** markvoelker_ has quit IRC16:05
*** IanGovett has joined #openstack-cinder16:06
*** sayali has joined #openstack-cinder16:06
*** zhipeng has quit IRC16:07
openstackgerritMichael Price proposed openstack/cinder: Add additional SSC extra specs to E-Series driver
openstackgerritMichael Price proposed openstack/cinder: Implement thin provisioning support for E-Series
*** mriedem_away is now known as mriedem16:08
*** IanGovett1 has joined #openstack-cinder16:10
*** IanGovett has quit IRC16:10
mriedemjgriffith: is there a reason for moving it from the volume manager to the volume api?16:10
mriedemoh i see the comment now16:10
mriedemjgriffith: i'm assuming you're referring to sdague's oslo blueprint/rules on logging16:11
*** leeantho has joined #openstack-cinder16:11
*** jistr has quit IRC16:13
mriedemjgriffith: i could change from info to debug in - i think that makes sense16:14
mriedemi wanted to keep the log message before the call to the volume driver though16:14
jgriffithmriedem: correct16:14
jgriffithmriedem: fair enough16:14
jgriffithmriedem: I still gave it a +2 :)16:14
mriedemyeah, i'll update quick16:14
jgriffithmriedem: I would prefer it demoted to debug in that case, but leaving it up to you16:15
jgriffithmriedem: if you don't like any of those options I'll go ahead and approve it16:15
jgriffithmriedem: but our logging is kind of a mess and inconsistent in Cinder :(16:15
openstackgerritGorka Eguileor proposed openstack/cinder: Remove API races from delete methods
openstackgerritGorka Eguileor proposed openstack/cinder: Add atomic conditional updates to objects
openstackgerritGorka Eguileor proposed openstack/cinder: WIP: Remove more API races
openstackgerritGorka Eguileor proposed openstack/cinder: Move get_by_id to CinderObject
openstackgerritGorka Eguileor proposed openstack/cinder: Improve metadata update operations
openstackgerritGorka Eguileor proposed openstack/cinder: Remove API races from attach and detach methods
openstackgerritGorka Eguileor proposed openstack/cinder: Rename db.api methods to match models' names
jgriffithI see everybody has been busy creating all sorts of branches and using stash this week-end16:16
patrickeastjgriffith: hey, let me know if/when you have a couple min to chat about the image cache review, just saw your comments and wanted to make sure I understand everything16:16
jgriffithpatrickeast: I'm working on some stuff now16:17
jgriffithpatrickeast: with the caching16:17
jgriffithpatrickeast: so the biggest problem is the context stuff16:17
jgriffithpatrickeast: it's also "kinda fat"16:17
jgriffithpatrickeast: we could've done this much more easily with a new of modified create_from_image I think.. but that's not a big deal16:18
jgriffithpatrickeast: I'm not sure why your db calls are puking all over the place though16:18
patrickeastjgriffith: yea so thats new, i haven't seen that16:18
jgriffithpatrickeast: it's sort of disturbing16:18
patrickeastjgriffith: ill have to go try and repo this morining, it might be something from one of the rebases16:18
jgriffithpatrickeast: and what's WORSE, is thanks to tflow retries I end up with 3 image volumes16:18
patrickeastjgriffith: definitely want to get that fixed16:19
jgriffithpatrickeast: but none in the db and none that ever get used16:19
patrickeastjgriffith: shiit16:19
patrickeastjgriffith: thats not good16:19
jgriffithpatrickeast: at the very least we need to mark these as non-retryable for taskflow16:19
patrickeastjgriffith: do you know how to do that?16:19
* patrickeast doesn't16:19
jgriffithpatrickeast: yeah... there's a list of exceptions you can add in and say "don't try these again"16:20
jgriffithpatrickeast: but that means we also have to translate and raise something appropriate in that case16:20
*** e0ne has quit IRC16:20
patrickeastjgriffith: ah, ok, gotcha16:20
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Add new Connector APIs for path validation
jgriffithpatrickeast: the tenant thing is easier to get around, drivers can pull in the config themeselves16:20
jgriffithpatrickeast: but that's kind of a PITA as it's just using the existing clone... it *should* be able to just use the owner context like we already do in the clone call16:21
patrickeastjgriffith: yea so i wasn't sure about the tenant, imo its weird to have each backend be able to use their own internal tenant... so i'm not really sure the problem with that config16:21
jgriffithpatrickeast: I agree, I'm certainly not proposing each has their own tenant16:21
jgriffithpatrickeast: it looks like you moved some of the configs into manager which I think will help with some of the things I was concerned about so that's good16:22
patrickeastjgriffith: ok cool, but that does mean i don't understand what the issue is :o16:22
jgriffithpatrickeast: so here's the problem....16:22
*** jaypipes has joined #openstack-cinder16:22
*** skylerberg has joined #openstack-cinder16:22
jgriffithpatrickeast: You create the template (create-volume(internal-tenant-context, image=xxxx))16:22
jgriffithpatrickeast: that's all cool16:22
jgriffithpatrickeast: BUT, then the clone_volume call goes down16:23
*** ronis has joined #openstack-cinder16:23
openstackgerritEric Harney proposed openstack/cinder: Detect addition of executable files
openstackgerritMatt Riedemann proposed openstack/cinder: Add debug logging before attaching volume in driver
jgriffithpatrickeast: it's supposed to be "source-context" and vref as params16:23
jgriffithpatrickeast: that way the src context is used to "find" the actual source volume for the clone... and the destination project info is pulled from the vref16:24
patrickeastjgriffith: ok, im following you there16:24
jgriffithpatrickeast: it looks like you're passing in the context for the destination on the clone call16:24
jgriffithat least that's what I'm seeing at the driver layer16:24
* patrickeast pulls up the code16:24
patrickeastlemmie check... i thought i was doing what you just described as the correct way, but totally possible i mixed em up16:25
*** salv-orl_ has joined #openstack-cinder16:25
jgriffithpatrickeast: hmm...16:25
patrickeastso this is in the cache hit clone or in the cache miss creation of the template clone?16:25
jgriffithpatrickeast: could be... I haven't stepped through everything yet (surprisingly kinda busy this morning :) )16:25
patrickeastor both?16:25
patrickeastheh yea i imagine today will be busy for a lot of foks16:26
jgriffithpatrickeast: not sure... all I know is it's in a clone call16:26
patrickeastok i'll take a look16:26
*** ronis has quit IRC16:26
jgriffithmriedem: you're the man!!!16:26
patrickeastone last thing, just to make sure we are on the same page, my comment about documenting it wont work for some drivers wasn't anything to do with the driver configs not working, i agree that needs to be fixed, i was talking specifically about the problem 3PAR has16:27
patrickeastjgriffith: ^16:27
*** harlowja_at_home has quit IRC16:28
jgriffithpatrickeast: understood... but I have issues with that :(16:28
*** salv-orlando has quit IRC16:28
jgriffithpatrickeast: I'm not going to fight it any more, but I STILL firmly believe that this business of "some features work for some backends, and not for others" is the COMPLETE WRONG direction for Cinder16:28
patrickeastjgriffith: i've discussed this a bunch with hemna and the problem is 100% the way they implemented their clone_volume in the driver16:28
patrickeastjgriffith: its a bug in their driver16:28
jgriffithpatrickeast: understood16:29
jgriffithpatrickeast: the good thing is I'm obviously the only one who cares about this any more so it's not really a big deal :)16:29
*** porrua has joined #openstack-cinder16:29
jgriffithpatrickeast: Totally cool with that.16:30
patrickeastjgriffith: :( don't get me wrong i understand what you are saying, i guess i see this as differently... it isn't that they *cant* do it, its just a bug they have with the way its implemented right now16:30
*** jungleboyj has quit IRC16:30
jgriffithpatrickeast: I'm just using it as an opportunity to state my opinion that I think we're making some pretty big mistakes with the project and we should really think about quality and behavior for people using Cinder16:31
patrickeastjgriffith: gotcha16:31
*** markvoel_ has quit IRC16:31
*** ociuhandu has joined #openstack-cinder16:31
jgriffithpatrickeast: and stop worrying so much about the features our company has to offer or what BS our sales people feed us that they "have to have"16:31
jgriffithpatrickeast: anyway, don't get me started :)16:31
*** markvoelker has joined #openstack-cinder16:31
*** yrabl has quit IRC16:32
patrickeastjgriffith: haha, alright, well i think we are on the same page with this stuff... i'll try and figure out whats going on with the db and double check the context stuff16:32
*** markvoelker_ has joined #openstack-cinder16:32
jgriffithpatrickeast: cool... I'm looking as well (on and off)16:32
patrickeastjgriffith: thanks for taking the time to look at this stuff, i appreciate the in-depth review16:33
*** lifeless has quit IRC16:34
*** angela-s has joined #openstack-cinder16:34
*** annegentle has joined #openstack-cinder16:35
*** markvoelker has quit IRC16:36
*** markvoelker_ has quit IRC16:37
sidbhatt13Hello. I'm new here and just getting started developing a new volume driver. Quick question: the minimum features that a driver must implement are listed here ( but the page doesn't mention Juno/Kilo/Liberty. Has anything changed, or is the list on that page still current?16:39
*** crose has joined #openstack-cinder16:39
*** jungleboyj has joined #openstack-cinder16:40
*** yuriy_n17 has quit IRC16:40
*** lifeless has joined #openstack-cinder16:41
jgriffithpatrickeast: ya know, it sure would've been easier to just put all this in the manager :)16:42
jgriffithsidbhatt13: not much has changed... BUT16:43
jgriffithsidbhatt13: you probably will have better luck just looking at the base VD class in openstack/cinder/cinder/volume/driver.py16:44
*** s_amann has quit IRC16:44
patrickeastjgriffith: maybe, I tried a couple of ways but it always ended up worse jumping back and forth between task flow code and manager code... seemed to be best if its all one way or the other16:44
sidbhatt13jgriffith: i'll do that, thanks.16:45
jgriffithpatrickeast: yeah.... but the nice thing about keeping it in manager is you can just effect everything locally as a driver16:45
jgriffithpatrickeast: you don't need to jump back and forth... but admittedly, yes there are a couple of details that end up being a trade-off and might not be much better16:45
*** sghanekar has joined #openstack-cinder16:45
patrickeastjgriffith: we can always look into refactoring it too, ideally this can we changed around and not affect the end user in any way16:46
jgriffithpatrickeast: agreed16:47
jgriffithsidbhatt13: look for these:
jgriffithsidbhatt13: things in base vd with the abc decorator are things you "MUST" impelement16:48
jgriffithsidbhatt13: good place to start at any rate16:48
jgriffithsidbhatt13: and make sure you referece the lvm driver for some details16:48
jgriffithsidbhatt13: things are going to be kinda hectic around here the next day or two but there's usually plenty of helpful people around16:49
sidbhatt13jgriffith: yup, looking at that class now. i'll hang out here and also subscribe to the mailing list in case i have other questions. btw, any idea what the deadline might be for completing a new volume driver for the "M" release?16:50
jgriffithsidbhatt13: for M it's going to likely be the first milestone again (don't know exact date)16:51
jgriffithsidbhatt13: but will probably be around end of Nov/Dec I'm guessing (early)16:51
sidbhatt13jgriffith: ok. thanks.16:52
*** tpsilva has joined #openstack-cinder16:52
*** crose has quit IRC16:53
*** lpetrut has quit IRC16:53
*** dims has quit IRC16:55
*** salv-orl_ has quit IRC16:56
sdaguejgriffith: - openssl now a requirement for the tests, but not specified?17:00
sdague2 cinder patches failing in gate over this17:01
jgriffithsdague: looking...17:01
*** kvidvans has quit IRC17:01
jgriffithsdague: sigh... k17:02
jgriffithsdague: I'll get it added or get xyang1 to check it out (her code introduced it)17:02
sdaguejgriffith: ok, thanks17:02
jgriffithsdague: sure17:02
jgriffithxyang1: you around?17:02
*** earlephilhower has joined #openstack-cinder17:02
xyang1jgriffith: What is this?17:02
*** dannywilson has quit IRC17:02
jgriffithxyang1: looks like vmax added a dep on ssl?17:03
xyang1jgriffith: Let me see17:03
jgriffithxyang1: ./ ssl17:04
xyang1jgriffith: I wonder why it failed now17:04
xyang1It was merged some time ago17:05
jgriffithxyang1: looking....17:06
*** IanGovett1 has quit IRC17:08
*** rhagarty_ has quit IRC17:09
*** rhagarty has quit IRC17:09
sdagueit may be that there is some platform assumption that's not expressed in tox that leaked through?17:10
*** julim has joined #openstack-cinder17:11
jgriffithsdague: which patch ref is that?  I think I'm confused.17:14
*** rhagarty has joined #openstack-cinder17:14
jgriffithxyang1: either way, as your using ssl wouldn't hurt us to update to explicitly adding pyopenssl I don't hink17:14
*** rhagarty_ has joined #openstack-cinder17:14
xyang1I don't know.  This was merged back in Feb. 817:15
*** rhedlind has joined #openstack-cinder17:15
xyang1jgriffith: sdague I'd like to see which patch this is failing on17:16
jgriffithxyang1: regardless, the import of OpenSSL is in the driver code17:16
sdaguejust go to
jgriffithxyang1: and we're not doing anything anywhere ot ensure it's there17:16
sdaguethere are 2 patches failing in gate17:16
jgriffithsdague: yeah, but I'm not seeing the same thing :(17:17
jgriffithsdague: I'm just seeing the gluster fail?17:17
sdagueoh, they just reset17:17
sdaguebecause ironic17:17
jgriffithsdague: hehe.. ok, so I'm not crazy17:17
sdague - those results are still registered17:17
openstackgerrityogeshprasad proposed openstack/cinder: Retype support for CloudByte iSCSI cinder driver
sdague2015-08-31 16:32:35.299 | Triggered by:
sdaguethe review url always exists in the console17:18
openstackgerrityogeshprasad proposed openstack/cinder: Retype support for CloudByte iSCSI cinder driver
*** aix has quit IRC17:21
*** alexschm has quit IRC17:23
*** IanGovett has joined #openstack-cinder17:24
xyang1jgriffith: sdague I can open a bug and fix our test to ignore the import error?17:24
*** e0ne has joined #openstack-cinder17:26
*** dims has joined #openstack-cinder17:29
*** vilobhmm1 has joined #openstack-cinder17:29
*** dims_ has joined #openstack-cinder17:30
*** dims has quit IRC17:30
openstackgerritJoel Coffman proposed openstack/cinder: Clean up volume_type_update method
nikeshmhi, do cinder supports incremental snapshot?17:35
*** e0ne has quit IRC17:36
*** e0ne has joined #openstack-cinder17:39
*** kvidvans has joined #openstack-cinder17:40
*** diablo_rojo has quit IRC17:40
*** lpetrut has joined #openstack-cinder17:43
*** ig0r__ has joined #openstack-cinder17:43
jbernardnikeshm: it depends on the backend being used17:44
jbernardnikeshm: and the cinder driver that enables it (to be more precise)17:44
jbernardnikeshm: but it can, yes17:44
*** ig0r_ has quit IRC17:46
sdaguexyang1 / jgriffith all those patches are failing in the gate again17:46
*** MIDENN_ has joined #openstack-cinder17:46
xyang1sdague: pyOpenSSL is in the requirement17:47
nikeshmjbernard: can you share any link of incremental snapshot?17:47
*** crose has joined #openstack-cinder17:47
jbernardnikeshm: you can look at the rbd driver17:48
sdaguexyang1: where?17:48
xyang1sdague: jgriffith gloabal-requirements.txt17:48
*** aix has joined #openstack-cinder17:48
sdaguexyang1: it's not in cinder's requirements file17:49
xyang1sdague: You mean the test requirement should have this too?17:50
sdaguexyang1: one of the requirements files needs it, if it's going to unconditionally import it17:50
sdaguethere needs to be no assumption on system level packages17:50
sdagueyes, there are now 4 cinder patches failing in the gate17:51
sdaguefor this reason17:51
xyang1sdague: I wonder if something got removed recently as this is an old change17:52
sdaguexyang1: might have been17:52
sdaguemtreinish: any thoughts on why this might have just exposed?17:53
sdaguedid something change with ostestr17:53
mtreinishsdague: is there a link? I don't have any context17:53
mtreinishbut I haven't pushed a ostestr release in a few weeks17:54
sdagueall the cinder tests are now failing on openssl not being there -
nikeshmjbernard: thanks, what about glustefs driver?17:54
sdaguebut apparently, this isn't new17:54
sdagueso, trying to develop a working theory on why17:54
mtreinishsdague: that error is coming from the test runner trying to do discovery17:54
sdaguelike discovery changing17:54
sdaguemtreinish: right17:55
sdaguemtreinish: but it's on patches that don't touch that file17:55
sdaguethat has the discovery issue17:55
jbernardnikeshm: i think you'll find support there as well17:56
openstackgerrityogeshprasad proposed openstack/cinder: Retype support for CloudByte iSCSI cinder driver
mtreinishsdague: well the last unittest2 release (which is where discovery would be) is from june17:58
sdaguemtreinish: yep, nothing is jumping out at me either17:58
sdaguewhich is why I was throwing it to you as a hail marry17:58
sdaguein case I missed something17:58
*** IlyaG has quit IRC17:59
mtreinishsdague: well, the other thing is maybe it was getting pulled into the nodepool images before18:00
mtreinishbut that would imply site packages was enabled18:00
mtreinishso it it's probably not that18:00
sdagueso here is a failing run -
sdaguehere is a passing run -
sdaguein the passing run, openssl definitely gets installed18:01
mtreinishsdague: yep, it's getting pulled in by something18:02
nikeshmjbernard: when i googled "Incremental Snapshot in openstack cinder", its giving "incremental backup support" in cinder18:03
sdaguein the failing case, no cffi installing either18:03
*** timcl has quit IRC18:03
*** IlyaG_ has joined #openstack-cinder18:03
sdagueno cryptography18:03
jbernardnikeshm: that's a different feature18:03
e0nethingee: hi Mike. are you around?18:04
*** salv-orlando has joined #openstack-cinder18:04
*** IlyaG_ has quit IRC18:05
sdagueah... it's the python-glanceclient release18:06
sdaguepython-glanceclient 1.0 no longer requires pyopenssl directly18:06
sdaguethey removed all of that18:06
mtreinishsdague: oh, ok, that makes sense18:06
sdaguethe cinder code still needs pyopenssl, but didn't specify it18:06
mtreinishsdague: well the fix is the same either way, cinder should be requiring pyopenssl if it is importing it18:07
sdagueand worked because glanceclient dragged it in18:07
openstackgerritMitsuhiro Tanino proposed openstack/cinder-specs: Fix Get Volume Driver Capabilities Spec
sdaguejgriffith / xyang1 ^^^18:07
sdaguethat's why you used to work before18:07
sdagueand are going to fail 100% now18:07
*** salv-orlando has quit IRC18:09
sdagueok, jgriffith / xyang1 someone posting that patch?18:12
xyang1sdague: I can do it18:12
openstackgerritSean Dague proposed openstack/cinder: Add support for file I/O volume migration
xyang1sdague: jgriffith so I add it to cinder test-requirement only?18:13
sdagueok, I'm clearing all the gate queue patches at the moment, as these are jamming it up for others18:13
sdaguexyang1: that should be fine18:13
*** IlyaG has joined #openstack-cinder18:13
openstackgerritSean Dague proposed openstack/cinder: Enhance FC zone support for Huawei driver
xyang1sdague: Ok, thanks18:13
openstackgerritSean Dague proposed openstack/cinder: Add pagination to backups
*** takedakn has quit IRC18:13
openstackgerritSean Dague proposed openstack/cinder: Allow specified backend capabilities to be retrieved
openstackgerritSean Dague proposed openstack/cinder: LVM: add the exception handling to volume copy
sdaguexyang1: - opened a bug for this18:16
openstackLaunchpad bug 1490668 in Cinder "cinder test-requirements does not include pyopenssl" [Critical,New]18:16
xyang1sdague: Ok18:16
tbarronyah, our internal CI is clogged up with this one too ^^^^ :)18:17
sdaguetbarron: would have been helpful to raise that fact :)18:17
eharneyxyang1: i think i've found a way to get the tests to pass w/o adding it as a requirement18:17
*** IlyaG has quit IRC18:17
xyang1eharney: Oh, how18:17
tbarronsdague: agree, but I just realized what it was ...18:17
*** markvoelker has joined #openstack-cinder18:17
*** IlyaG has joined #openstack-cinder18:18
sdagueeharney: it explodes on pep8 for similar reasons18:18
*** julim has quit IRC18:18
nikeshmjbernard: thanks, when i searched "copy on write", then its showing Ceph RBD drivers18:18
eharneyxyang1: we often do this kind of thing for driver reqs:
xyang1eharney: That is ok too, if you guys think that is better18:19
nikeshmjbernard: but not showing about any other driver18:20
xyang1sdague: eharney which way is better18:20
*** markvoelker has quit IRC18:21
*** julim has joined #openstack-cinder18:21
eharneyxyang1: i always prefer this because it means we don't add deps that are only relevant for one driver18:21
sdaguehonestly, I'm not sure what you are testing, I'll leave eharney to decide18:21
xyang1sdague: eharney ok, will do18:21
*** IlyaG has quit IRC18:22
*** diablo_rojo has joined #openstack-cinder18:23
jbernardnikeshm: you'll have to look into the particular drivers you're interested in18:25
openstackgerritEric Harney proposed openstack/cinder: Don't require OpenSSL for unit tests
*** rushil has quit IRC18:25
eharneyxyang1: ^  had some unrelated issues in my test env (i think), hopefully this works18:25
jbernardnikeshm: the incremental logic is implemented in the storage backend, so you're looking for backend support and wether the cinder driver exposes it18:26
xyang1eharney: Ok, let's see18:26
jbernardnikeshm: i think you'll find support for snapshots in most cases; backups are a different matter18:26
*** p0rtal has joined #openstack-cinder18:27
jbernardnikeshm: improvements on that front are underway18:27
jgriffitheharney: xyang1 just my two cents... but IMO that should use pyopenssl, and even if you use the if/else hack you still should probably have it in the requirements file.18:28
eharneyjgriffith: we've opted to not add things like this for a handful of other drivers to the requirements files in the past...18:28
xyang1jgriffith: If we add to requirement, then we still need to change the driver?18:28
eharneyjgriffith: hp reworked their stuff to not require hp3parclient, we have that for rbd/rados too...18:30
*** takedakn has joined #openstack-cinder18:30
eharneyjgriffith: drbd as well... violin... purestorage...18:31
jgriffitheharney: well, I meant the requirements, not test-requirements18:32
openstackgerritxing-yang proposed openstack/cinder: Catch OpenSSL ImportError
eharneyjgriffith: we don't list any of those in either18:32
jgriffitheharney: but I guess you're saying the same thing there18:32
jgriffitheharney: geesh dude, calm down18:33
jgriffitheharney: :)18:33
jgriffithgive me a chance to type already18:33
xyang1jgriffith: eharney sdague patch submitted, let me know if I need to change :)18:33
jgriffithxyang1: cool, does that monkey patch work like that when it's None?18:34
xyang1jgriffith: Test does not require monkey patch18:35
jgriffithxyang1: cool18:35
xyang1jgriffith: I can still add pyopenssl in requirement if you and eharney can agree with each othet:)18:35
jgriffitheharney: I'm certainly not arguing.. but I do want to clarify if that's ok with you?18:36
jgriffitheharney: I personally see a difference between things like hp-3parclient and openssl libraries18:36
jgriffitheharney: which is the only reason I mentioned it18:36
eharneyjgriffith: well.. the drbd driver does this with dbus, which is more like the openssl requirements18:37
jgriffitheharney: I don't think that things like pyopenssl and vendor libraries fall into the same category; however I also don't care.  If it's preferred to if out the import because only one driver is using it that's fine by me18:37
jgriffitheharney: ok, but regardless that doesn't change the statement I made above.  Again, I'm not arguing with you at all18:38
eharneywe can add it as a req if it's easier to fix the gate etc, but i'm not really sure what the benefit is18:38
jgriffitheharney: I don't think there is any benefit and not saying we should do that.  Just pointing out why that was my first response18:38
jgriffiththat's all... no big deal18:38
patrickeastjgriffith: hey, so i can't for the life of me repro that db issue you are seeing with the image cache, got any tips on how you made it break? are you using devstack defaults or anything special with your db config?18:39
patrickeast i can create volumes from images and cache hits, misses, evictions, shared images, private images, different sizes, etc. and everything looks peachy in the logs
eharneythe main reason i don't like adding them is that most people don't need that library at all, but it brings it into every test env install, etc18:39
*** salv-orlando has joined #openstack-cinder18:39
jgriffitheharney: yeah, I agree with you.  I used to give Walt crap about that ALL the time ;)18:40
jgriffitheharney: totally cool18:40
jgriffitheharney: the only thing that's different is that it is in the global reqs already, and used in other projects; but really I'm not arguing one way over the other18:40
eharneyjgriffith: well... it was in there, anyway :)18:41
mriedemxyang1: did an os-brick release get out with the rootwrap filter stuff?18:41
mriedemdep freeze is this week18:41
xyang1mriedem: I don't think so18:41
*** takedakn has quit IRC18:41
xyang1mriedem: Where is hemna18:41
jgriffithmriedem: nope18:41
jgriffithWalt's not around, haven't seen him today18:42
*** sgotliv has joined #openstack-cinder18:42
mriedemi'll push the release request18:42
mriedemand add cinder core18:42
jgriffithmriedem: thanks18:42
xyang1mriedem: Thanks18:42
*** dflorea has joined #openstack-cinder18:42
jgriffithmriedem: I'm looking to see if there's anything queued up that might be *important*18:42
*** lpetrut has quit IRC18:43
xyang1kmartin: Will hemna be back soon?18:44
jgriffithmriedem: might want to merge this one
jgriffithanish: ^^18:44
mriedemno idea18:45
mriedemjgriffith: there are 2 changes with +2s though,n,z18:45
jgriffithmriedem: yeah... and it looks like anish 's change breaks some things18:46
mriedemi'd probably hold off on feature adds at this point18:46
jgriffithmriedem: agreed18:46
jgriffithmriedem: I didn't notice that until just now18:46
*** daneyon_ has joined #openstack-cinder18:47
jgriffithmriedem: ok, those two should be good, other than that I think freeze is fine18:48
jgriffithmriedem: hemna and others can revisit if they want later18:48
*** ig0r__ has quit IRC18:48
*** dflorea_ has joined #openstack-cinder18:48
mriedemcool, posting to the dev ML for now on the plan18:48
mriedemthen i'll push the releases change once those are merged18:49
jgriffithmriedem: perfect18:49
*** daneyon has quit IRC18:49
*** timcl has joined #openstack-cinder18:50
*** dflorea has quit IRC18:51
tbarronxyang1: please see my comment on
tbarronxyang1: hopefully I'm doing something wrong ...18:52
eharneytbarron: xyang1: see my version, it handled that
tbarroneharney: kk, I'll try that18:53
*** gouthamr has quit IRC18:55
xyang1eharney: Ok, thanks18:55
jgriffitheharney: oh, I didn't see you submitted as well18:55
jgriffithxyang1: isn't that the case I specifically asked you about here in IRC ? :)18:56
*** edmondsw has quit IRC18:56
xyang1jgriffith: Sorry18:56
jgriffithxyang1: LOL... no problem18:57
kmartinxyang1, Walt is having computer issues but he is here?18:57
*** hemna has joined #openstack-cinder18:57
tbarroneharney: xyang1: works for me locally18:57
*** edmondsw has joined #openstack-cinder18:57
kmartinxyang1, need me to have him check something?18:57
jgriffitheharney: xyang1 I'll leave it between the two of you which one to merge :)18:57
jgriffithkmartin: stats on releasing os-brick before dependency freeze18:58
hemnasorry, my desktop here at work is hosed, been offline all morning18:58
jgriffithhemna: check dev list18:58
*** ociuhandu has quit IRC18:58
anishmriedem: jgriffith err, what breakage18:58
jgriffithhemna: mriedem posted question WRT os-brick and dependency freeze18:59
jgriffithanish: looks like all the 3'rd party CI's failed your os-brick patch18:59
anishcoz last I discussed there was a breakage with iser being called from outside of nova, but that is fixes now18:59
hemnasomehow lvm upgrade on my ubuntu changed the UUID of my / partition and now it can't boot.  bleh.18:59
anishjgriffith: yeah, that I noticed. weirdly it's the fc CIs18:59
anishI do not touch fc at all19:00
mriedemhemna: jgriffith:
mriedemxyang1: ^19:00
jgriffithanish: well... you know what I say about FC :)  But regardless if it breaks them that's no-bueno19:00
*** gouthamr has joined #openstack-cinder19:00
*** gouthamr has quit IRC19:01
hemnamriedem, sounds ok to me.19:01
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
anishjgriffith: makes sense19:03
anishI'd be happy to fix things if anyone who knows the FC code could talk to me19:03
*** gouthamr has joined #openstack-cinder19:04
anishthe purestorage ci seems to fail in code that is completely in the FC connector. there is no code change there so very weird19:04
jgriffithanish: could be a *coincidence*19:04
jgriffithanish: you need to figure that out with patrickeast and hemna I guess19:05
SwansonIs there a global FC breakage that explains my broken FC CI?19:05
patrickeastjgriffith: anish: which review?19:05
anishiono, IBM is passing FC19:05
patrickeastew, that looks like a rootwrap problem?
*** shyama has quit IRC19:06
*** Apoorva has quit IRC19:06
anishpatrickeast: yeah, same things with IBM storwize19:07
anishrootwrap problems19:07
asselin_we've been having fc issue for our cinder patches  but hemna found the issue this morning19:07
asselin_we were inadvertently using os-brick/master instead of the version in pip19:08
patrickeastanish: so its either something in that change, or a brick problem (maybe fixed later on?)19:08
patrickeastanish: that isn't something that would show up in the intermittent failures (which we do have)19:08
anishexact same problem with HP storage as well19:08
asselin_anish, if you're having the same issue as us, look at your pip-freeze19:09
anishasselin_: nope, I'm just trying to figure out if my patch is breaking fc19:09
*** dflorea has joined #openstack-cinder19:10
*** dflorea_ has quit IRC19:11
asselin_hemna, ^^19:11
*** dflorea has quit IRC19:12
*** jwcroppe has joined #openstack-cinder19:13
openstackgerritTom Barron proposed openstack/cinder: Scheduler-based over-subscription for NFS drivers
anishjgriffith: well, they all seems to be failing with the exact same root wrap issue except for HP iscsi which has a failure on tempest.scenario.test_volume_boot_pattern.TestVolumeBootPattern.test_create_ebs_image_and_check_boot19:14
*** jwcroppe_ has quit IRC19:15
anishat least that I think could not be related to the patch19:15
asselin_test_create_ebs_image_and_check_boot should have been fixed by an upstream tempest updated19:19
*** subhadeep_de has quit IRC19:20
*** Apoorva has joined #openstack-cinder19:21
angela-shemna: Brocade CI has reported successful run now:
*** hodos|2 has joined #openstack-cinder19:29
*** lpetrut has joined #openstack-cinder19:29
*** hodos has quit IRC19:31
openstackgerritxing-yang proposed openstack/cinder: ScaleIO driver should use os-brick connector
jgriffithpatrickeast: If you want to repro the issue; just hack in a 'raise' in your clone_volume method in your driver19:36
jgriffithpatrickeast: then you should see the same thing19:37
*** sidbhatt13 has quit IRC19:37
*** salv-orlando has quit IRC19:37
patrickeastjgriffith: ok cool19:37
patrickeastjgriffith: wait a sec, that means the exception was coming from your driver code, what was causing that? the context thing?19:39
patrickeastjgriffith: or you put in a raise just to test things?19:39
jgriffithYes, the context thing19:40
jgriffithwhich I still haven't had a chance to look at, but firing it back up now19:40
patrickeastjgriffith: gotcha19:40
patrickeastjgriffith: i was looking through the code and as far as i can tell it *should* be working, but I think there might be some weirdness introduced with the dual context stuff i'm not fully accounting for19:41
openstackgerritSean Dague proposed openstack/cinder: Port image_utils to Python 3
jgriffithpatrickeast: LOL.. well of course it *should* be working19:47
patrickeastjgriffith: :D19:47
sdagueI'm going to keep sniping out cinder changes that I see until the unit test fix lands in the gate19:47
patrickeastjgriffith: what i mean is the db calls for volume_get work with either context, both don't get seem to actually get filtered by project19:48
jgriffithsdague: I noticed...  not sure others caught that we shouldn't be +2/A 'ing for the time being19:48
jgriffithhemna: smcginnis_afk DuncanT patrickeast e0ne ^^19:49
jgriffithsdague: is this something we can help with or easier for you to just pick them off as you come across them?19:49
e0nejgriffith: no +A for tonight. i've got it!:)19:49
sdaguehonestly, I'm about to call it a day, so I'll leave it for others19:49
jgriffithe0ne: well... just until we get the EMC import patch merged19:50
*** salv-orlando has joined #openstack-cinder19:50
sdagueI'm dealing with a well pump replacement and it's time for beer19:50
jgriffithsdague: cool... you just yanking anything that fails?19:50
jgriffithsdague: Ohh... ouch!19:50
sdaguejgriffith: honestly, right now, I'm just yanking any cinder patch, because they'll all fail19:50
e0nejgriffith: anyway, it's almost 11pm. good time to stop reviewing:)19:50
sdagueI'm just checking that it's not the fix patch first before I do19:50
jgriffithsdague: alright... good luck with the pump;  I'll watch the gate for a bit19:50
sdaguejgriffith: thanks19:51
jungleboyjjgriffith: So, nothing else from the cinder-liberty-3 review list is going to make it today?19:51
jgriffithjungleboyj: NO19:51
jgriffithjungleboyj: we just need to get the one patch in before anything else at this point19:51
jungleboyjjgriffith: Ok.19:52
jungleboyjStill cutting off items from that list at mdnight tonight?19:52
jgriffithjungleboyj: yes19:52
jgriffithand it's midnight UTC19:52
*** DericHorn-HP has joined #openstack-cinder19:52
*** sidbhatt13 has joined #openstack-cinder19:53
jgriffithand well... I guess given the time it takes for things to get through, maybe the answer to your question will in fact be yes L3 is done19:53
jungleboyjjgriffith: Yeah, that was what I was trying to get it.19:53
jgriffithjungleboyj: sorry.... lost track of how slow things are at the moment19:53
jungleboyjMidnight UTC is 4 hours away.19:54
jungleboyjShould I go through everything and -219:54
jgriffithOf course we'll be waiting a long time it appears19:54
jgriffithjungleboyj: maybe so19:54
* jungleboyj volunteers to become very unpopular19:54
jgriffithjungleboyj: we can consider where things are at tonight if/after this merges:
e0nejungleboyj: you're "Capitan Oslo", not "Capitan -2" ;)19:55
openstackgerritJoel Coffman proposed openstack/cinder: Enforce constraints at DB layer rather than API
jungleboyjjgriffith: Up to you, you are the boss right now.19:56
hemnaso we're done19:56
jgriffithhemna: well, the queue is at 5 hours19:56
jungleboyjOooh, we got hemna's attention.19:56
hemnawell, I had to eat lunch.19:56
jgriffithhemna: doing the math and based on the "rules" that were set in place it kinda looks like features are done19:56
hemnastill no working desktop.19:57
jgriffithhemna: note that doesn't apply to bug fixes, and honestly IMO features should have pretty much been done already anyway19:57
jgriffithhemna: What's up with that!!!19:57
*** bswartz has quit IRC19:57
jungleboyjhemna: What happened to your desktop?19:57
diablo_rojoI think you should not let jungleboyj do all the -2's. He just wants his +1/-1 score to be lower than mine :P19:57
hemnaok, there is just 1 more os-brick patch that I'd like to get in, but haven't been able to test it yet.19:57
hemnanot sure what the 'rules' are wrt os-brick19:58
kmartinI have a new driver, is it to late to make liberty?19:58
jgriffithhemna: os-brick doesn't count here19:58
*** Yogi1 has quit IRC19:58
* hemna cheers19:58
jgriffithkmartin: every release you do that!!!19:58
*** rushil has joined #openstack-cinder19:58
jgriffithjungleboyj: not this is features, not bug fixes19:58
kmartinyep, had to keep the tradition19:58
jgriffithjungleboyj: I'm happy to help out19:59
jungleboyjjgriffith: Yes ... I know this is the just the features.19:59
jgriffithjust making sure Initially I thought "not that many" but now that I look again OI-Vai19:59
jgriffiththere's a lot of crap dumped in over the week-end20:00
jungleboyjjgriffith: Yeah.20:00
*** david-lyle has quit IRC20:01
hemnaso bug fixes can still +A ?20:01
jungleboyjhemna: Yes.20:02
hemnaok coolio.  thanks20:02
jgriffithtbarron: appears this is done... yes?20:02
jungleboyjI am mainly concerned with this:
e0nejgriffith: what about "moving something to objects" patches?20:03
jgriffithe0ne: sigh...20:03
jgriffithe0ne: technically those should be out, but they're in progress; and I hate the half in half out thing20:03
jgriffithe0ne: I'm just starting by cleaning up the BP's... then I'll go from there20:03
jgriffithe0ne: jungleboyj the idea being that if it's not on the L3 list in launchpad it's out20:05
jgriffithe0ne: jungleboyj I'm moving everything that's not submitted or not needs review out of L3 now20:05
*** markvoelker has joined #openstack-cinder20:06
*** david-lyle has joined #openstack-cinder20:06
jungleboyjOk, you want me to block the reviews for now then?20:06
jgriffithjungleboyj: so if it's a bp and it's NOT on the L3 list it gets a -2 or just ignored20:08
*** subhadeep_de has joined #openstack-cinder20:08
xyang1jgriffith: What about our vote in last meeting? Not valid?:)20:08
jgriffithjungleboyj: speaking of which:
jgriffithjungleboyj: not sure what the current state of that one is ^^20:08
jgriffithjungleboyj: looks done?20:09
*** sidbhatt13 has left #openstack-cinder20:09
jungleboyjjgriffith:  Yeah, we could mark that Done.  The support is in and we have a 3rd Party CI for it.  I had thought we needed to get unit tests going on it, but with the CI we are good.20:10
jungleboyjSo, lets close that one out.20:10
*** dflorea has joined #openstack-cinder20:10
jgriffithjungleboyj: so those are all the easy ones, items that had nothing associated or had no reviews20:10
jgriffithjungleboyj: now the "hard" part is all the "Needs Code Review ones"20:10
jgriffithwhich will make people cry20:10
jgriffithjungleboyj: k, marking it as Implemented20:11
*** dims has joined #openstack-cinder20:11
*** dims has quit IRC20:12
jgriffithjungleboyj: now for the 27 "needs review" items :(20:12
jungleboyjSheesh, 27 BPs.20:12
jgriffithjungleboyj: yeah... remaining!20:12
jgriffithjungleboyj: fairest thing I can think of is to first cut all that are "low" priority20:12
xyang1I think some are merged, but status not updated20:12
jgriffithjungleboyj: e0ne eharney xyang1 thoughts on that?20:13
patrickeastseems like the right way to do it, thats what the priority is there for, right?20:13
mtaninoCan I change the status of this to implemented?
jgriffithxyang1: yeah... I'm looking at those as I go through... but as I've ALWAYS said in the past it's submitters responsibility to make sure LP is updated and reflects reality!20:13
*** crose has quit IRC20:13
jgriffithmtanino: if it's implemented YES you should20:13
jgriffithpatrickeast: yeah... in theory :)20:14
*** sidbhatt13 has joined #openstack-cinder20:14
xyang1jgriffith: I can't change others bp status any way20:14
e0nejgriffith: agree. we can drop low priorities BPs from l-320:14
jungleboyjOk, starting to update those reviews.20:14
*** sidbhatt13 has left #openstack-cinder20:14
angela-sjgriffith: addressing the elephant in the room, when can we start requesting exceptions? :)20:14
*** dims_ has quit IRC20:14
jgriffithangela-s: no exceptions this time around :)20:14
* jungleboyj hears the trumpet of the elephant.20:14
mtaninojgriffith: I thought we shouldn't change that status of the BP. DONE.20:15
jgriffithangela-s: so we can probably discuss tomorrow or Wed in meeting20:15
jgriffithmtanino: well... I don't know how Mike does things, but IMO if the system didn't pick it up the way it was supposed to you should change it20:15
jgriffithor notify somebody else to change it20:15
angela-sjgriffith: ok, will wait until tomorrow or in meeting, thanks20:15
mtaninojgriffith: I understand. Thank you.20:16
xyang1jgriffith: We don't really have an exception process20:16
jgriffithPersonally LP is the only thing I have to go off of for status, so if it's wrong, I'm kinda screwed20:16
jgriffithxyang1: correct.... that needs to be discussed with the PTL20:16
jgriffithxyang1: who informed me that there was no such thing20:16
jgriffithxyang1: so I just follow orders20:16
xyang1jgriffith: :). Not completely true20:16
jgriffithxyang1: what's not completely true?20:17
*** dflorea_ has joined #openstack-cinder20:17
xyang1jgriffith: But I don't want to get anyone into trouble20:17
jgriffithxyang1: is there a process I don't know about?  OOops20:17
jgriffithxyang1: no no... if I'm doing something wrong, please let me know20:17
* jgriffith slams on the brakes20:17
jgriffithstops what he's doing20:17
jgriffithxyang1: LOL... or you mean about "me following orders" :)20:18
xyang1jgriffith: No:)20:18
patrickeastany chance we could get a follow up email to explaining the FFE policy (or lack thereof) would make it easier for people like me who may have to explain why things don't go in20:19
*** rushil has quit IRC20:20
*** dflorea has quit IRC20:20
*** pv__ has joined #openstack-cinder20:21
pv__hey im debugging my attach-volume functionality on my cinder driver, its an FC driver and im running devstack on an HBA enabled machine20:21
pv__im getting in my n-cpu logs20:21
pv__after my c-vol log reports "initialize connection completed successfully"20:21
*** Apoorva has quit IRC20:22
pv__when i do a systool -c fc_host -v it shows both the HBA cards with  their respective WWNs but apparently openstack doesnt see the same thing, i dont understand why20:22
pv__any help?20:22
jungleboyjjgriffith: xyang1 hemna Any concerns with this approach:
jgriffithjungleboyj: that seems reasonable and pretty clear IMO20:23
hemnajungleboyj, lgtm20:23
jungleboyjjgriffith: Thank you.  Moving forward in the low proprity reviews then.20:23
*** kvidvans has quit IRC20:24
*** nzrf has left #openstack-cinder20:24
*** Apoorva has joined #openstack-cinder20:25
pv__has anyone encountered the same sort of issue when testing the attach method for their FC driver?20:26
jgriffithtbarron: what's up with this one:
jgriffithtbarron: I don't see any code for it?20:27
jgriffithtbarron: ahh... wait, the second link has a bunch of stuff20:27
jgriffithtbarron: all listed as either merged or abandoned... assuming done?20:28
*** thangp has quit IRC20:32
*** dflorea_ has quit IRC20:33
angela-spv__: may i ask what vendor and what OS?20:33
pv__both devices are emulex20:33
pv__and OS is ubuntu 14.04 server20:34
jungleboyjjgriffith: patrickeast What did we decide on this one?  Does it have longer or not?
angela-spv__: ok, just ruling out 1 possibility20:34
xyang1jgriffith: That one is WIP. Https://
xyang1jgriffith: tbarron's patch20:34
*** dflorea has joined #openstack-cinder20:35
jgriffithxyang1: thanks20:35
angela-spv__: do the HBAs show up in the nameserver on the FC switches?20:35
jgriffithxyang1: so sounds like that's out then20:35
xyang1jgriffith: Yes20:35
*** svilgelm has joined #openstack-cinder20:35
pv__im not sure sorry, what exactly do you mean by that?20:35
*** simondodsley has quit IRC20:36
*** diogogmt has joined #openstack-cinder20:36
angela-spv__: on the CLI of the FC switch, if you issue 'nsshow', are the WWNs showing up?20:36
*** lpetrut has quit IRC20:36
jgriffithhemna: I'm deferring this one to you:
jgriffithpatrickeast: smcginnis_afk xyang1 e0ne jungleboyj other than that, the LP listing should now be somewhat accurate:
hemnajgriffith, we need that one in L20:38
*** dflorea_ has joined #openstack-cinder20:38
jgriffithhemna: cool, that's why I left it alone and checked with you :)20:38
pv__one second, im getting access to the FC switch to check20:38
*** dflorea has quit IRC20:38
jgriffithhemna: you might want to bump the priority to reflect that20:38
*** svilgelm has left #openstack-cinder20:39
e0nejgriffith: thanks you your job!20:39
hemnaI can't change the priority :(20:39
jungleboyjjgriffith: Thanks.  I have -2's on all the low priority items.  Let me know how you would like to on the higher priority ones.20:39
tbarronjgriffith: sorry, was afk.  I got assigned other work to do for L3 and will target for M20:39
jungleboyj*to do20:39
jgriffithhemna:  now worries, I'll do it20:39
jgriffithhemna: that actually sounds like a backward compatbility bug to me20:40
*** kvidvans has joined #openstack-cinder20:40
hemnajgriffith, yah it kinda is.20:40
*** gouthamr has quit IRC20:40
*** dflorea has joined #openstack-cinder20:41
*** _cjones_ has joined #openstack-cinder20:41
*** _cjones_ has quit IRC20:42
jgriffithOk, so FWIW; that brings our BP backlog to 1420:42
jgriffithThere's still one or two in there that I don't think have a prayer20:42
jgriffithbut, the idea being that those should be the priority for reviews at this point20:43
*** Lee1092 has quit IRC20:43
*** dflorea_ has quit IRC20:45
*** dflorea has quit IRC20:45
tbarronjgriffith: is indeed done (I was AFK)20:45
pv__yeah the HBAs are showing up when i do an nsshow20:45
pv__from the FC switch cli20:45
jgriffithtbarron: good :)20:45
*** dflorea has joined #openstack-cinder20:45
pv__the WWNs match the ones I get when i do a systool -c fc_host -v20:46
jgriffithtbarron: If my adjustments were correct no worries, just let me know if I missed something20:46
pv__so im not sure why nova keeps telling me it cant find any FC devices20:46
*** rushil has joined #openstack-cinder20:46
patrickeastjungleboyj: not sure if it was answered, but that replication one imo falls into the same category as the rest of the low priority bp's20:46
patrickeastjgriffith: ^ correct?20:46
patrickeastthere was the vote to give more time for driver impls... but i had assumed that was just the 1 week from the deadline to post one gerrit/pass jenkins and the actual feature freeze20:47
jgriffithpatrickeast: I was avoiding that one intentionally :(20:48
*** bkopilov has quit IRC20:48
jgriffithpatrickeast: but the reality is if you look at the back-log even if it were +2/A right now, it likely wouldn't merge until Wed some time20:49
jgriffithpatrickeast: I'm not making the call on that one though20:49
jgriffithat least not today20:49
jgriffiththere's enough that's obvious right now to work through IMO20:49
angela-spv__: ok, i just wanted to confirm that the FC part is working.  unfortunately, i do not know what the root of the issue you are facing is.20:49
*** dflorea has quit IRC20:50
pv__alright thanks! ill keep at it20:50
*** timcl has quit IRC20:51
openstackgerritJohn Griffith proposed openstack/cinder: Fix a merge problem in VMAX driver
*** ebalduf has joined #openstack-cinder20:52
*** bkopilov has joined #openstack-cinder20:53
*** xyang1 has quit IRC20:55
*** jwcroppe_ has joined #openstack-cinder20:55
*** markvoelker has quit IRC20:56
*** jwcroppe has quit IRC20:57
*** dims has joined #openstack-cinder20:57
tbarronjgriffith: all our driver feature patches were of course in Low and only started getting review attention recently, as everyone expected given their priority.  But we had bps approve and working code up before the deadlines.21:00
*** Apoorva_ has joined #openstack-cinder21:00
tbarronjgriffith: jungleboyj: so all these remaining are getting -2s at this point and retargeted for Mitaka?21:00
*** bkopilov has quit IRC21:00
*** xyang1 has joined #openstack-cinder21:01
*** esker has quit IRC21:01
tbarronjgriffith: jungleboyj: what I'm trying to figure out is how we as code/bp submitters could be doing anything differently on this?21:01
*** jwcroppe has joined #openstack-cinder21:02
*** jwcroppe_ has quit IRC21:04
*** Apoorva has quit IRC21:04
*** annegentle has quit IRC21:04
jgriffithtbarron: submit early, submit often21:06
jgriffithtbarron: and expect things like gate congestion on the last day of feature acceptance21:06
tbarronjgriffith: we did - but didn't get review cycles till just very recently.  I'm not blaming, just describing.21:06
jgriffithtbarron: I hear ya21:07
tbarronjgriffith: these are changes where we met the deadlines that I'm talking about.  I told my people *not* to ask for reviews on IRC, that there was21:07
*** jwcroppe_ has joined #openstack-cinder21:07
jgriffithtbarron: understood21:08
tbarrona prioritization of core features first and that reviewers would get to them.  Keep up with review comments, stay rebased, don't bug people.21:08
jgriffithtbarron: just to be clear though, that doesn't mean that the review process goes on indefinitely21:08
guitarzantbarron: maybe you can get them on a meeting agenda?21:08
jgriffithguitarzan: +121:08
guitarzanmuch more focused than random irc pestering21:09
jgriffithtbarron: I have a long list of things that I never got to this time around21:09
jgriffithand it sucks, but it's also happened to me every release for the last 4 years so I'm kinda used to it21:09
jgriffithlife goes on21:09
tbarronguitarzan: are you saying I'm doing "random irc pestering" now?21:09
guitarzantbarron: no, the opposite21:09
tbarronguitarzan: kk, sorry, I may be a bit twitchy :-)21:10
jgriffithand every time I tell myself "Ok, get all your crap done in the first milestone... but I never do21:10
*** jwcroppe has quit IRC21:10
guitarzantbarron: I'm saying you were having people *not* pester, and instead maybe you could just get a batch of reviews on a meeting agenda21:10
*** hemna has quit IRC21:10
SwansonI've had luck with random irc pestering.21:10
guitarzansure, squeaky wheel and all that, but it's hard to recommend that :)21:10
tbarronguitarzan: last time I put something on the agenda (about getting our bps approved) I got slapped down pretty hard :-)21:10
jgriffithas guitarzan traditionally irc meetings are the best place to raise awareness and get input21:10
jgriffithtbarron: by who?21:10
guitarzantbarron: ah, that sucks...21:10
jgriffithtbarron: that what weekly meetings are for21:11
jgriffithtbarron: if nothing else, raise it during Open-Discussion (when we have them)21:11
*** svilgelm has joined #openstack-cinder21:11
jgriffithtbarron: so if you want to put a prioritized list together and present things as needed for consideration feel free.  I suspect I'll be getting a flood of those types of things the next 48 hours :)21:12
SwansonI'm pretty sure thingee said to not put such things on the agenda.  But I suppose that doesn't include open discussion.21:12
jgriffithSwanson: ahh.. he may have21:12
jgriffithtbarron: just to be clear though:
tbarron has eight of our reviews that made the bp and code + jenkins deadlines.21:13
*** akerr has quit IRC21:14
jgriffithtbarron: even if you follow the guidelines and meet the date, that doesn't change when freeze happens21:14
tbarronseveral have no review comments.21:14
Swansonjgriffith: I'm trying to remember if he redirected people to the mailing list for that21:14
*** julim has quit IRC21:14
tbarronmost have "add a period to a message" or "please refactor a bit" type remarks.21:14
tbarronnothing fundamental.21:14
jgriffithtbarron: yeah, I've always felt we were out of control with that business21:14
tbarronjgriffith: I understand that freeze happens anyways.  How can we increase review throughput so that given the lack of any21:15
tbarronfundamental objections it is still impossible for submitters to make the freeze.  We don't have control or influence on the21:15
tbarronreview throughput except to do our part to stay up to date, respond right away, etc.21:16
SwansonI've always felt that having two spaces after a period was worth a -1 and has nothing to do with the involuntary eye twitch I've developed.21:16
jgriffithtbarron: better focus?  More reviewers?  Less code change?21:16
jgriffithI dunno21:16
jgriffithSwanson: :)21:16
*** svilgelm has quit IRC21:16
tbarronI personally have nothing against nitpick -1s.21:16
* jgriffith does21:16
jgriffithI'd rather come back and fix little things if need be.. but that's not the point21:17
jgriffithtbarron: so look, here's the thing21:17
tbarronI just pointed out that that is what we have on the outstanding reviews in that  queue.  Or we have no comments at all.21:17
*** lprice has joined #openstack-cinder21:17
jgriffithtbarron: I'm not the PTL. This information and process was communicated quite clearly BY the PTL on numerous occasions21:17
tbarronjgriffith: understood, and I'm not asking you to violate the process.21:18
jgriffithtbarron: you were present, both at the mid-cycle meetup, and in the weekly meeting 2 weeks ago when it was reiterated and clarified21:18
tbarronjgriffith: I am not quarelling with sticking to the process.21:18
tbarronI *am* going on record that we followed the process and it isn't working.21:18
jgriffithtbarron: well, sometimes process for the sake of process is broken21:18
jgriffithtbarron: and I agree completely21:18
*** esker has joined #openstack-cinder21:19
jgriffithevery release, folks come up with great process ideas to "solve" the throughput problems, and every release people are upset21:19
Swansontbarron: I really do think that if you can't bring these things up at the meeting then you need to pester on IRC or on the ML.  I don't think we should be doing that but if you need something pushed through that is your only option.21:19
jgriffiththis is ALL projects BTW, not just Cinder21:19
tbarronjgriffith: I admire thingee introducing clear process and setting clear deadliines.21:19
tbarronjgriffith: I *do not* want to undermine that.21:19
jgriffithtbarron: well then.. you have them21:19
jgriffithtbarron: you don't get it both ways21:20
*** bkopilov has joined #openstack-cinder21:20
jgriffithtbarron: you admire and appreciate them... except when it doesn't work out for you?21:20
jgriffithtbarron: :)21:20
tbarronI *do* think that we need to somehow going forwards modify something so we don't get messed up when we play by the rules.21:20
tbarronjgriffith: that's *not* what I'm saying.21:20
jgriffithtbarron: ok... so look;21:21
tbarronI am saying that we played by the rules and didn't get review cycles.21:21
jgriffithtbarron: first of all, you can certainly raise an appeal/request for something21:21
tbarronNot in time.21:21
jgriffithtbarron: I can guarantee that asking for all 8 would be a wast of your time and mine21:21
jgriffithtbarron: I would also say that yeah... life is hard in OpenStack21:21
jgriffithtime is short21:21
jgriffiththings never work out21:21
jgriffithtbarron: AND, to somehow thing that "Hey... if I submit everything 1 week before the deadline I'll be fine" is stupid21:22
jgriffithit's not going to work, because you and EVERYBODY else is trying to do the same thing21:22
jgriffithit's a simple math problem21:22
jgriffithit doesn't work21:22
jgriffiththe gate gets overloaded, reviewers get overloaded, the change rate is just too high21:22
jgriffithtbarron: now, as I said before; I'm perfectly open to folks brining items to Wed's meeting that we can discuss as a team21:23
tbarronjgriffith: so I guess I should learn to treat L3 deadlines as L2 deadlines :-)21:23
jgriffithtbarron: I'm also working on proposals for the summit in Tokyo to maybe modify some of the ways we work currently in the hopes that it helps with this sort of thing21:24
*** esker has quit IRC21:24
tbarronjgriffith: I'll talk to our people about priorities, triage, etc.21:24
jgriffithtbarron: and finally... one last thing I'd like to point out!!!
tbarronjgriffith: I'll look for those proposals with great interest.21:24
*** esker has joined #openstack-cinder21:24
jgriffith^^ I submitted that SPECIFICALLY because of this very situation!!!21:24
SwansonI thought I was going to drop dead when I had to merge by a kilo-1 to get my driver in.  Could not believe how long it was taking to get through gate.21:25
jgriffithtbarron: it went over like a led balloon21:25
jgriffithSwanson: exactly!21:25
*** dustins has quit IRC21:26
jgriffithlanding strips, interim deadlines, driver freeze, feature freeze... doesn't matter; they're all deadlines and they all incur the same challenges in the end21:26
*** hodos|2 has quit IRC21:26
Swansonjgriffith: I think we talked about this friday.  I think drivers need to trail the release so that they can implement all the features in a release.21:26
jgriffithSwanson: yes, although when I read back through the notes I don't think I presented that whole thing very well21:27
*** jwcroppe has joined #openstack-cinder21:28
patrickeastjust to add my 2c it does seem like maybe we could at least improve throughput of reviews a bit more with better automated tests (maybe things like in-tree functional tests) and increasing core reviewers... both have some risk though to decrease quality if not done well though21:28
patrickeastthe idea that folks need to download my patch, spin up their own system, and make sure it doesn't blow up seems like a lot of time wasted21:29
patrickeastfor some things totally makes sense21:29
*** edtubill has quit IRC21:29
patrickeastbut if its going to just barf immediately its not really valuable21:29
jgriffithpatrickeast: sure... although when it comes to "core" things that's what CI is for no?21:29
patrickeastjgriffith: yea but there isn't an easy way to just flip on some new feature or add new tempest tests along side a patch21:30
jgriffithpatrickeast: and BTW, part of the challenge here is the "vendor-specific" features/methods that become optional21:30
patrickeastjgriffith: i guess my point was there are probably some things we could do to help at least a little bit with it21:30
jgriffithpatrickeast: there is if we go back to doing things the way we used to, where that was part of the process21:30
patrickeastjgriffith: yea definitely a problem with vendor specific things21:30
jgriffithpatrickeast: and YES, there are always thing that we can do to improve... for sure!21:30
tbarronwe could consider a random ordering of reviews in the backlog so that one or two vendors aren't at the end (irrespective of being ready as early or earlier than others).21:30
asselin_patrickeast, you could add cross-project dependency to tempest, no?21:30
*** jwcroppe_ has quit IRC21:31
jgriffithpatrickeast: more cores is fine, but it doesn't solve the problem IMO21:31
*** jkremer has quit IRC21:31
patrickeastasselin_: yea but its then requiring the cinder reviewers to go look at the tempest review and review that21:31
jgriffithtbarron: so personally that's why I don't use any of the etherpads, dashboards or whiz-bang toys for review21:31
tbarronthe random ordering suggestion is built on agreeing - which I'm not sure I should - to the idea that there will always be more backlog than review throughput.21:31
jgriffithtbarron: just randomly pick anything that hasn't failed or been down-voted already21:31
patrickeastjgriffith: yea i agree, its a bandaid approach and doesn't scale21:31
jgriffithtbarron: have you ever looked at the backlog?21:32
*** mudassirlatif has joined #openstack-cinder21:32
jungleboyjtbarron: Catching up.  So are there particular ones you feel we failed you on?21:32
tbarronjgriffith: yeah, I've actually been doing some reviews from time to time, so I do look at it.21:33
jgriffithtbarron: go here:,n,z21:33
tbarronjungleboyj: no one personally has failed me.  But we have 8 reviews on the launchpad (bascially almost all we submitted) that21:33
tbarronare getting -2ed here.21:33
jgriffithtbarron: tell me how many pages you go back before finding the end :)21:33
tbarronjgriffith: I'm well familiar with that page21:34
*** merooney has quit IRC21:34
tbarronjgriffith: I think that's why the etherpad was set up.  We were towards the bottom of the Low priority.21:34
tbarronjgriffith: I didn't try to change that.  Big mistake.21:34
SwansonIs every cinder patch is going to fail today?21:34
*** e0ne has quit IRC21:34
jgriffithSwanson: yes until the ssl patch merges21:35
Swansonjgriffith: which one is that?21:35
*** annegentle has joined #openstack-cinder21:35
tbarronI think I've made my point.  Am going to have a beer and move on.21:35
eharneyjgriffith: grenade just failed that patch for the second time, i think with an unrelated error :/21:36
jgriffitheharney: darn it!!21:36
eharneypkg_resources.ContextualVersionConflict: (pbr 0.11.0 (/usr/local/lib/python2.7/dist-packages), Requirement.parse('pbr<2.0,>=1.3'), set(['sqlalchemy-migrate']))21:36
* eharney dunno21:36
Swansonjgriffith: that one is toast21:36
*** hemna has joined #openstack-cinder21:36
jgriffitheharney: so there's a bunch of stuff going on in gate right now21:37
jgriffitheharney: hmm... but I don't see that anywhere else21:37
patrickeasteharney: i saw an email a few min ago that mentioned a new sqlalchemy-migrate release breaking the gate21:37
guitarzan"555 fails in 24 hrs" impressive one21:37
patrickeasti thought just for kilo though21:37
jgriffithpatrickeast: well, that would puke grenade then :)21:38
*** annegent_ has joined #openstack-cinder21:38
jgriffithpatrickeast: well, not always, but usually21:38
patrickeastohh derp, yea grenade would be broken21:38
eharneyyeah, grenade does kilo->liberty21:38
*** tpsilva has quit IRC21:38
eharneypatrickeast: and yes it was the sqlalchemy-migrate issue21:39
jgriffithtbarron: so the point here is "shit happens", even if you meet all deadlines etc; if the gate pukes, or something goes wrong in the process there's not much can be done21:39
jgriffithtbarron: find rank the items in priority and "why" and post to dev list21:39
jgriffithwe'll go from there21:39
tbarronjgriffith: I hear ya, thanks.21:39
*** annegentle has quit IRC21:40
jungleboyjWith regards to the 'IRC Pestering' from earlier I don't mind people asking me for reviews in IRC at all.  I think Mike prefers they not happen in the meetings.21:40
*** sgotliv has quit IRC21:41
jgriffitheharney: so how hard do you think it would be to add an automated check for imports against requirements file?21:43
kmartinjgriffith, in grizzly I remember them all fitting on two pages, the good old days21:43
jgriffitheharney: so we could check that third party section maybe and say if it doens't have the "if xxxx/else" block and it's not in the requirements file fail21:44
*** porrua has quit IRC21:44
*** gouthamr has joined #openstack-cinder21:44
jgriffithkmartin: haha.. and keep in mind that I think auto-abandon is running occasionally now so that's not even all of them LOL21:44
eharneyjgriffith: i started thinking about that, i'm not too sure21:44
eharneyjgriffith: but, i think it also can be done with a job that runs cinder-volume on a node w/ no other services installed (though it wouldn't have helped in this case)21:44
guitarzanactually mapping to requirements sounds really hard21:46
jgriffitheharney: OH... snap!!21:46
jgriffitheharney: that reminds me, I never finished my functional test stuff21:47
*** ebalduf has quit IRC21:47
jgriffitheharney: so cinder/tests/functional would be perfect for that21:47
eharneyjgriffith: hmmm, quite possible21:47
*** kvidvans has quit IRC21:47
jgriffitheharney: something that just walked through the tree and did an init on every driver even?21:48
eharneythough i'm not sure it solves the "oops now it broke because one of our deps changed their deps" case which is what we hit here21:48
jgriffitheharney: true21:48
eharneyjgriffith: well, the unit tests basically do that now21:48
jgriffitheharney: true, but this would eliminate mocks/stubs and other various magic21:48
eharneyah, that's true21:49
jgriffitheharney: well.. until people mocked things there I suppose21:49
eharneyjgriffith: you could pull it off if you added a new driver method that was pre-check_for_setup_error, like check_for_environment_error21:49
*** cbader has quit IRC21:49
jgriffitheharney: yeah... I like that idea21:50
eharneyi dunno how popular that would be, but it's worth considering21:50
jgriffitheharney: I think it's a pretty worthwhile investment personally21:50
jgriffitheharney: if we communicate it early enough, and give everybody time to implement that "one" method21:51
jgriffitheharney: it's not overly burdensome21:51
jgriffithsummit topic perhaps21:51
eharneyjgriffith: yeah, and fortunately that's something we could implement for drivers if needed, too21:51
eharneyseems like a good idea21:51
eharney(to have a summit topic, that is)21:51
jgriffitheharney: +121:52
jgriffitheharney: I'm also playing with the idea of a contrib directory21:53
guitarzanhmm, it doesn't even really have to be implemented in the driver no?21:53
guitarzanyou just need something to instantiate their class21:53
guitarzaner, parse the driver even21:53
jgriffithguitarzan: well the idea would be to get results not just barf I think21:54
eharneyguitarzan: i think that's true only if we decided to go the route of not allowing soft deps21:54
guitarzanjgriffith: ah... that sounds hard too :)21:54
guitarzaneharney: not sure I follow21:55
jgriffithguitarzan: so I almost have this model of a "contrib" manager as a sub-class of the base manager... and that's where "experimental" or non-required methods ends up living21:55
guitarzanbut that's ok21:55
eharneyguitarzan: currently we don't list a lot of driver library dependencies in requirements.txt / test-requirements.txt21:55
guitarzaneharney: gotcha21:55
guitarzanbecause they shouldn't be there21:55
jgriffithguitarzan: I'm also playing around with a similar model for drivers to extend functionality the same sort of way21:55
guitarzanso I'm not sure how this is possible21:55
guitarzanjgriffith: not another mass driver refactor! :D21:56
jgriffithguitarzan: nope, no refactor...21:56
guitarzanI thougth abc won :)21:56
*** garthb_ has joined #openstack-cinder21:56
jgriffithguitarzan: well, yeah :)21:57
*** garthb_ has quit IRC21:57
jgriffithguitarzan: but it just moves those optional things out of the base21:57
jgriffithI'm sure everyone will hate it even more than my last proposal, but I'll keep trying :)21:57
*** IanGovett has quit IRC21:57
jgriffithguitarzan: BTW, this is an example
jungleboyjjgriffith: I don't feel good -2'ing this one:
jgriffithjungleboyj: well, I'm the one who blocked it :(22:04
jgriffithjungleboyj: I'm fine with merging it... it's a doc update that goes with a feature that already merged22:05
jgriffithjungleboyj: IMO those sorts of things should be allowed all the way up to the end22:05
jgriffithdocs are a completely different category I think22:05
jgriffithjungleboyj: good catch22:05
jungleboyjOk, I will look at that one.  I will not -2 that one.22:06
jungleboyjjgriffith: Thank's for the input.22:06
jgriffithjungleboyj: I'm going to just +2/A it because I blocked it, and after reading the response agree that documenting is good22:06
*** eharney has quit IRC22:07
jungleboyjjgriffith: Good deal.22:07
*** markvoelker has joined #openstack-cinder22:11
mtaninojungleboyj: this one was pulled from gate queue due to the gate test error...22:12
patrickeast:*( rip image cache... guess we try again in M22:12
jungleboyjpatrickeast: Yeah, that was hard to do.  I wanted that in.22:12
jgriffithpatrickeast: same as I told tbarron propose it22:12
openstackgerritJoel Coffman proposed openstack/cinder: Enforce constraints at DB layer rather than API
jgriffithpatrickeast: ie exception and the core team can all vote on it22:12
patrickeastjgriffith: yea i'll bring it up22:13
jgriffithjungleboyj: assuming that's ok with you as well22:13
jgriffithseems fair to at least discuss22:13
jungleboyjjgriffith: I am fine with that.22:13
jgriffithjungleboyj: also to mtanino 's point, watch those suckers that were pulled with the comment "(Pulled from gate, cinder can no longer pass unit tests)22:13
jungleboyjOh, mtanino I am sorry.  Fixing.22:14
jgriffithjungleboyj: those were items that were +2/A  already22:14
jgriffithso should get a reset22:14
jgriffithjungleboyj: my fault, I should've pointed that out22:14
mtaninojgriffith: jungleboyj thank you.22:14
jungleboyjFixed.  Were there others.22:15
jgriffithjungleboyj: hmm... there's a few22:15
jgriffithjungleboyj: owner will be either sdague or myself22:15
*** markvoelker has quit IRC22:16
jgriffithjungleboyj: and go figure, I just added one in when I KNEW I was removing anything that went over there :(22:16
openstackgerritJohn Griffith proposed openstack/cinder: Add instructions on how to deploy API under Apache
*** garthb has quit IRC22:17
*** garthb has joined #openstack-cinder22:17
*** garthb_ has joined #openstack-cinder22:17
*** garthb_ has quit IRC22:17
jungleboyjThis one had been +A:
*** garthb has quit IRC22:18
*** garthb_ has joined #openstack-cinder22:18
*** garthb has joined #openstack-cinder22:19
jungleboyjLooks like I should remove my -2 here as well:
xyang1jgriffith: This one is already merged:
jungleboyjI need to drop off for a few.  Then back on.  Will clean up the etherpad.22:21
xyang1jgriffith: This one is also merged:
jgriffithxyang1: thanks!22:23
jgriffithpatrickeast: hey!  I believe I found the problem22:24
patrickeastjgriffith: \o/22:24
patrickeastjgriffith: so what was it?22:24
jgriffithpatrickeast: so it looks like the create-from image goes down, BUT it creates it with the regular tenant (creates the cache volume with the normal tenant)22:25
jgriffithpatrickeast: verifying now22:25
patrickeastjgriffith: oops, so it does do that first22:25
patrickeastjgriffith: then clones that volume to make the cached one22:25
jgriffithpatrickeast: hmm... something's funky22:26
*** jungleboyj has quit IRC22:26
patrickeastflows supposed to be something like: create minimal sized volume from image like normal -> clone and cache the clone -> extend original (if needed) -> return back original as the requested one22:26
jgriffithpatrickeast: ohhhhh22:27
jgriffithpatrickeast: so it clones it AFTER the first create!22:27
jgriffithpatrickeast: DERP I've been looking at this all wrong then22:27
patrickeastjgriffith: yea, that got around a ton of the permission issues, the internal tenant doesn't have to do anything with glance anymore22:27
*** lprice has quit IRC22:27
patrickeastit *used* be the other way around22:27
patrickeastto be*22:27
jgriffithpatrickeast: ahh... that's how you fixed the glance issue I was having22:27
patrickeastjgriffith: yea we keep all our dirty tricks in-house now22:28
*** edmondsw has quit IRC22:28
guitarzanpatrickeast: hmm, can we ask for it to work the other way? :)22:29
* guitarzan runs22:29
patrickeastguitarzan: haha sure, but you have to give the internal tenant access to all your images22:29
patrickeastguitarzan: or the ones you want cached at least22:30
guitarzanI'm actually 100% ok with that22:30
guitarzanothers might not be though22:30
patrickeastguitarzan: its a problem for like private user images, imo it makes things really tricky to handle the permissions22:30
guitarzanI don't understand why though, you still have to do all the validation on future creates22:31
patrickeastguitarzan: yea sure, but we validate based on the caller context who is trying to create the volume22:31
patrickeastguitarzan: not on the cinder internal tenant22:31
guitarzanoh you mean the context of the initial create22:31
*** jaypipes has quit IRC22:32
*** annegent_ has quit IRC22:32
jgriffithpatrickeast: well shit!  That breaks everything for me :(22:32
guitarzananyway, it feels better to me to make the readonly copy first22:32
patrickeastas-is the permissions work the same way they do without the cache, if you do something like 'cinder create --image-id 123 10' you have to have permission to access image 12322:32
*** annegentle has joined #openstack-cinder22:32
guitarzanpatrickeast: sure, I expect that22:32
patrickeastinstead of having to *also* give the cinder internal tenant access to 12322:33
guitarzanbut that doesn't prevent us from fetching the image with an admin tenant22:33
patrickeastto get caching22:33
patrickeastjgriffith: :o why does that break it?22:33
guitarzanpatrickeast: anyway, just tossing that out there22:33
guitarzanOR even just storing it on a volume with the internal tenant using the customer token22:33
*** ociuhandu has joined #openstack-cinder22:33
tbarronso, I had that beer and am going to give an example of a change set that was not well served by our process:
patrickeastjgriffith: does the clone/extend thing not work for you? what happened?22:35
tbarronThis was *not* a NetApp change.22:36
tbarronIt was posted on 8/21 and not reviewed until 8/28.22:36
tbarronOn that day it received a -1 asking for a modification to the commit msg.22:36
tbarronThat mod was done and posted back in < 1.5 hours.22:36
tbarronIt then received threee +1s, almost right away.22:37
jgriffithcinderclient doesn't even apply here FWIW22:37
tbarronWasn't reviewed again till 8/28, which was a -2 because the deadline had been hit.22:37
mtaninooh, it's my patch.22:38
jgriffithtbarron: so when jungleboy gets back he can remove that22:38
*** hemna has quit IRC22:38
jgriffithtbarron: cinderclient is a completely different story22:38
jgriffithtbarron: although with dep freeze now I guess that's the same... hmm.22:38
mtaninojgriffith: I thought so too.22:38
jgriffithtbarron: FWIW it could have 5 +1's and that's not the same as 1 +1 and two +2's22:39
Swanson+1s are pretty much meaningless.22:41
tbarronjgriffith: sure, my point is that it just illustrates someone doing it right, there being nothing anybody has brought up wrong with the code, and it still getting axed.22:42
tbarronjgriffith: which is I guess your point that s*t happens22:43
jgriffithtbarron: dude, you're killin me.  It's not my process, I don't like it either.  The door isn't closed, what more can I say to you?22:43
*** resker has joined #openstack-cinder22:43
* jgriffith remember why he didn't want to be PTL any more :)22:43
tbarronjgriffith: :-)22:44
Swansonjgriffith: Not running for PTL again, I take it?  :)22:44
*** esker has quit IRC22:47
*** resker has quit IRC22:48
*** esker has joined #openstack-cinder22:48
*** jungleboyj has joined #openstack-cinder22:55
jungleboyjOk, I am back briefly.22:55
jungleboyjxyang1: jgriffith I am going to update the etherpad with what I did tonight.  Did you see anything that I -2'd that I shouldn't have?22:55
jgriffithjungleboyj: there was some questions about mtanino cinderclient change:
jgriffithjungleboyj: but other than that I think what you did was awesome22:56
jgriffithjungleboyj: thanks for stepping up and doing that22:56
jgriffithjungleboyj: I figure tomorrow and Wed we'll do the whole "group discussion" thing22:56
mtaninojgriffith: jungleboyj I think this one is also proposed cinderclient
jungleboyjOh, we shouldn't block that if it's dependent patch is going to be merged.  Thanks.22:57
xyang1jungleboyj: jgriffith are you -2 everything or just low priority ones22:58
jungleboyjxyang1: Everything at this point unless it is something that is stuck on the gate.22:59
xyang1jungleboyj: Ok22:59
*** gouthamr_ has joined #openstack-cinder22:59
xyang1jungleboyj: This one was pulled from gate
jungleboyjmtanino: I will have to follow up on that one with Vincent.23:00
mtaninojungleboyj: I undestand.23:01
jungleboyjxyang1: Thank you.23:01
*** gouthamr has quit IRC23:02
jungleboyjjgriffith: Agreed.  I think we should discuss the image cache work.  If there is anything else that comes up we can discuss.  I will make sure to stay on top of IRC tonight and tomorrow.23:03
jungleboyjjgriffith: You are welcome.  Thanks for taking the BP side of things.23:03
jgriffithjungleboyj: thanks!!23:03
*** annegentle has quit IRC23:04
mtaninoonly 7 in gate queue...23:07
jgriffithmtanino: yeah... getting there23:08
jungleboyjOk, I will be logged in if anyone else needs me to fix something.  I have some family commitments but then will be online tonight getting the etherpad updated.23:09
jgriffithjungleboyj: note23:09
jgriffithjungleboyj: we need this to merge:; so we can then recheck and merge this:
jgriffithjungleboyj: *before* anything goes in to verify stage in the gate23:10
jgriffithanything from Cinder that is23:10
jungleboyjOh man, what a mess.23:11
jgriffithI'm not completely sure how keeping new changes to Cinder impacts all of that right now, but that's what we've been doing at any rate23:11
jungleboyjjgriffith: Ok, so we need to watch for the first to merge, recheck/merge 218977 then merge the other ones that were +A.23:12
jungleboyjPiece of cake.23:12
jgriffithjungleboyj: :)23:12
jungleboyjjgriffith: Thanks for the roadmap.  I will keep an eye on it.23:13
jgriffithand that's why folks are bent out of shape, but honestly there's not much I can do about it23:13
jgriffithif the gate stops, the gate stops... it happens23:13
*** lcurtis has quit IRC23:13
jungleboyjOk, back in a couple hours.  Family is calling.  I think we are being fair about it by getting the other ones that had gotten +A a chance.23:13
*** ganso_ has quit IRC23:15
*** daneyon has joined #openstack-cinder23:16
*** Apoorva has joined #openstack-cinder23:17
*** daneyon_ has quit IRC23:17
*** Apoorva has quit IRC23:17
*** Apoorva has joined #openstack-cinder23:18
*** daneyon_ has joined #openstack-cinder23:18
*** Apoorva_ has quit IRC23:20
*** gouthamr has joined #openstack-cinder23:20
*** Zhongjun has joined #openstack-cinder23:21
*** daneyon has quit IRC23:22
*** gouthamr_ has quit IRC23:23
*** rushil has quit IRC23:24
*** lprice has joined #openstack-cinder23:30
*** rushil has joined #openstack-cinder23:32
*** rushil has quit IRC23:36
*** rushil has joined #openstack-cinder23:36
jgriffithpatrickeast: FYI, got the image-caching thing squared away23:44
jgriffithpatrickeast: it actually exposed a bug in my code that's been around *forever*23:45
jgriffithpatrickeast: thanks for your patience23:45
* jgriffith has a long list of bugs to fix in the next week or so :(23:45
patrickeastjgriffith: woot, always happy to help break things23:46
patrickeastfind bugs*23:46
patrickeastjgriffith: thanks for testing it out23:47
patrickeastspeaking of bugs... anyone seen issues with nova in their ci systems like so
*** smoriya has joined #openstack-cinder23:49
patrickeastonly seems to happen with my FC driver (which is actually a concerning theme with my ci failures...)23:49
jgriffithpatrickeast: that's yummy23:49
jgriffithpatrickeast: tried an ER search?23:49
patrickeastjgriffith: yea, nothing turned up23:50
patrickeastalthough i'm not a logstash ninja, so maybe am searching for the wrong thing23:50
jgriffithhmm... I got some results23:50
jgriffithwell... one ;)23:51
jgriffithpatrickeast: bumping out the time frame on it23:51
jgriffithpatrickeast: it's hit a few times... 4 in the last 7 days23:52
patrickeasthmm interesting, so it does seem to be the same error23:52
*** IlyaG has joined #openstack-cinder23:52
patrickeasthave to go dig into them more to see if its the same tests or not23:52
jgriffithpatrickeast: I wonder if this has anything to do with what mtreinish was looking at earlier?23:52
patrickeastjgriffith: oh maybe, that was the issue with the device not showing up on the partition tables, right?23:53
jgriffithpatrickeast: roger23:53
*** diegows has quit IRC23:53
*** DericHorn-HP has quit IRC23:54
*** andreykurilin__ has quit IRC23:55

Generated by 2.14.0 by Marius Gedminas - find it at!