Tuesday, 2015-05-05

*** erlon has quit IRC00:01
*** mtanino has quit IRC00:05
*** emagana_ has quit IRC00:06
*** emagana has joined #openstack-cinder00:06
*** emagana has quit IRC00:10
*** _cjones_ has joined #openstack-cinder00:13
*** geguileo has quit IRC00:26
*** vilobhmm1 has joined #openstack-cinder00:30
*** geguileo has joined #openstack-cinder00:34
*** vilobhmm11 has joined #openstack-cinder00:36
*** vilobhmm1 has quit IRC00:38
*** setmason has quit IRC00:40
*** _cjones_ has quit IRC00:43
*** leeantho has quit IRC00:49
*** IanGovett has joined #openstack-cinder00:52
*** primechuck has joined #openstack-cinder00:52
*** primechuck has quit IRC00:54
*** daneyon_ has quit IRC00:55
*** primechuck has joined #openstack-cinder00:55
*** zhenguo has joined #openstack-cinder00:56
*** vilobhmm11 has quit IRC01:04
*** rmesta has quit IRC01:06
*** Apoorva has quit IRC01:06
*** madskier has joined #openstack-cinder01:15
*** madskier has quit IRC01:19
*** IanGovett has quit IRC01:22
*** heyun has joined #openstack-cinder01:26
*** Lee1092 has joined #openstack-cinder01:28
nikeshmasselin__ patrickeast: thanks setting up CI,one small doubt, is PUBLISH_HOST means log server?01:31
patrickeastnikeshm: yep01:32
*** _cjones_ has joined #openstack-cinder01:39
nikeshmpatrickeast : export GIT_NAME in vars.sh do we have to give full name of gerrit CI account here?01:40
nikeshmfull name in https://review.openstack.org/#/settings/ for my CI account01:41
patrickeastnikeshm: yea, so like mine has...01:42
patrickeastexport GIT_NAME='Pure Storage Cinder CI'01:42
*** _cjones_ has quit IRC01:44
*** Yogi1 has joined #openstack-cinder01:44
*** Longgeek has joined #openstack-cinder01:46
openstackgerritEric Brown proposed openstack/cinder: VMware: insecure option should be exposed  https://review.openstack.org/17999101:48
nikeshmpatrickeast: thanks, i am setting up a jenkins master on a VM on different server and openstack cloud on different server,now i want to give openstack cloud info to jenkins master node,but in vi nodepool/nodepool.yaml.erb we are using localhost01:49
*** Yogi1 has quit IRC01:49
nikeshmdo i have to change localhost to the ip of cloudprovider ip01:49
nikeshmin vi nodepool/nodepool.yaml.erb01:49
patrickeastnikeshm: yea change the ‘auth-url’ for the provider01:49
nikeshmdata repository is on jenkins master slave01:50
patrickeastnikeshm: as an example, here is what mine looks like http://paste.openstack.org/show/214915/01:50
*** Yogi1 has joined #openstack-cinder01:52
openstackgerritEric Brown proposed openstack/cinder: VMware: insecure option should be exposed  https://review.openstack.org/17999101:54
*** Longgeek has quit IRC01:58
*** heyun has quit IRC01:59
*** julim has joined #openstack-cinder02:03
*** patrickeast has quit IRC02:04
openstackgerritJamie Lennox proposed openstack/cinder: Catch additional type conversion errors  https://review.openstack.org/18000302:14
*** akerr has joined #openstack-cinder02:31
*** akerr_ has joined #openstack-cinder02:35
nikeshmpatrckeast: in images section we have to give image-name which should be present in glance image-list,02:39
nikeshmcan i create it manually in cloud provider02:39
*** akerr has quit IRC02:39
nikeshmglance image-create --name trusty --disk-format qcow2 \   --container-format bare --is-public True --file ./trusty02:40
nikeshmand give name in image section02:40
*** Yogi1 has quit IRC02:44
*** dobson has quit IRC02:48
*** Longgeek has joined #openstack-cinder02:56
*** dobson has joined #openstack-cinder03:01
*** Longgeek has quit IRC03:08
*** dobson has quit IRC03:18
*** dobson has joined #openstack-cinder03:22
*** coolsvap|afk is now known as coolsvap03:23
*** dims_ has quit IRC03:23
*** setmason has joined #openstack-cinder03:23
*** MentalRay has joined #openstack-cinder03:23
*** timcl has quit IRC03:25
*** madskier has joined #openstack-cinder03:26
*** _cjones_ has joined #openstack-cinder03:28
*** links has joined #openstack-cinder03:31
*** _cjones_ has quit IRC03:32
*** akerr_ has quit IRC03:34
openstackgerritEric Brown proposed openstack/cinder: VMware: insecure option should be exposed  https://review.openstack.org/17999103:34
*** MentalRay has quit IRC03:36
*** rongze has joined #openstack-cinder03:39
*** avishay_ has joined #openstack-cinder03:43
*** MentalRay has joined #openstack-cinder03:43
*** MentalRay has quit IRC03:43
*** setmason has quit IRC03:46
*** sks has joined #openstack-cinder03:48
*** anshul has joined #openstack-cinder03:51
openstackgerritJohn Griffith proposed openstack/python-cinderclient: Add CLI read-only functional tests  https://review.openstack.org/17928003:51
*** ebalduf has quit IRC03:54
*** Longgeek has joined #openstack-cinder04:05
*** vilobhmm1 has joined #openstack-cinder04:12
nikeshmdo we have to cp scp.jpi /var/lib/jenkins/plugins/ in https://github.com/rasselin/os-ext-testing04:14
nikeshmpatrickeast asselin: ?04:15
nikeshmin document mentioned to download scp.jpi,but after that i didnot get04:16
*** shyama has joined #openstack-cinder04:19
*** dims has joined #openstack-cinder04:23
*** _cjones_ has joined #openstack-cinder04:28
*** dims has quit IRC04:28
*** Longgeek has quit IRC04:29
*** bkopilov has joined #openstack-cinder04:32
*** _cjones_ has quit IRC04:33
*** heyun has joined #openstack-cinder04:39
*** sks has quit IRC04:41
*** anuragpalsule has joined #openstack-cinder04:54
*** ebalduf has joined #openstack-cinder04:55
*** sks has joined #openstack-cinder04:56
*** rongze has quit IRC04:58
nikeshmi am getting this error on running nodepoold -d $DAEMON_ARGS    of https://github.com/rasselin/os-ext-testing-data     IOError: [Errno 2] No such file or directory: '/var/log/nodepool/image/image.log'04:58
*** ebalduf has quit IRC05:00
*** rongze has joined #openstack-cinder05:00
*** sks has quit IRC05:01
nikeshmDuncanT: i am getting one error in setting up https://github.com/rasselin/os-ext-testing-data05:02
*** madskier has quit IRC05:03
*** nkrinner has joined #openstack-cinder05:04
*** lpetrut has joined #openstack-cinder05:12
*** emagana has joined #openstack-cinder05:16
*** BharatK has joined #openstack-cinder05:17
*** rongze has quit IRC05:45
*** rongze has joined #openstack-cinder05:50
*** rongze has quit IRC05:52
*** beecee has quit IRC05:52
*** nlevinki has joined #openstack-cinder05:58
*** Longgeek has joined #openstack-cinder05:59
*** yrabl_ has joined #openstack-cinder06:00
*** yrabl_ has quit IRC06:01
*** nikeshm has quit IRC06:02
*** yrabl has quit IRC06:04
*** yrabl has joined #openstack-cinder06:04
*** SamYaple_ is now known as SamYaple06:13
*** ankit_ag has joined #openstack-cinder06:14
*** Longgeek_ has joined #openstack-cinder06:14
*** Longgeek has quit IRC06:17
*** _cjones_ has joined #openstack-cinder06:17
*** _cjones_ has quit IRC06:22
*** openstackgerrit has quit IRC06:23
*** vilobhmm1 has quit IRC06:23
*** openstackgerrit has joined #openstack-cinder06:24
*** sgotliv_ has quit IRC06:24
*** svasheka has quit IRC06:25
*** vilobhmm1 has joined #openstack-cinder06:27
*** vilobhmm11 has joined #openstack-cinder06:29
*** vilobhmm1 has quit IRC06:31
*** emagana has quit IRC06:32
*** emagana has joined #openstack-cinder06:33
*** svasheka has joined #openstack-cinder06:37
*** emagana has quit IRC06:38
*** zhenguo has quit IRC06:47
*** dulek has joined #openstack-cinder06:48
*** alecv has joined #openstack-cinder06:50
*** e0ne has joined #openstack-cinder06:51
*** zhenguo has joined #openstack-cinder06:53
*** lpetrut has quit IRC06:54
*** nlevinki has quit IRC06:57
*** e0ne has quit IRC07:06
*** ronis has joined #openstack-cinder07:08
*** sandywalsh has quit IRC07:11
*** markus_z has joined #openstack-cinder07:12
*** sandywalsh has joined #openstack-cinder07:12
*** tshefi has joined #openstack-cinder07:12
*** ebalduf has joined #openstack-cinder07:14
*** agarciam has joined #openstack-cinder07:16
*** ebalduf has quit IRC07:19
*** vilobhmm11 has quit IRC07:23
*** lpetrut has joined #openstack-cinder07:24
DuncanTnikeshm: Try mkdir that directory07:35
*** dulek has quit IRC07:43
*** dulek has joined #openstack-cinder07:44
*** chlong has quit IRC07:44
dulekthingee: Hi, I just noticed that session about task resumptions is scheduled before Active/Active c-vol.07:47
dulekthingee: I think it should be the other way, because task resumption approach depends on TaskFlow decision which will be made on Active/Active c-vol session.07:49
dulekthingee: Also it would be beneficial to know if tooz will be get into Cinder - this will be decided on the same session.07:49
*** btran has quit IRC07:50
*** jwang_ has quit IRC07:51
*** ndipanov has joined #openstack-cinder07:52
*** jwang has joined #openstack-cinder07:53
*** aarefiev has joined #openstack-cinder07:54
openstackgerritEdwin Wang proposed openstack/cinder: Add iSCSI protocol support for IBM FlashSystem  https://review.openstack.org/15620607:55
*** dulek has quit IRC07:55
*** dulek has joined #openstack-cinder07:55
*** ebalduf has joined #openstack-cinder07:58
*** dulek has quit IRC08:02
*** ebalduf has quit IRC08:03
*** _cjones_ has joined #openstack-cinder08:06
*** haomaiw__ has joined #openstack-cinder08:08
*** jordanP has joined #openstack-cinder08:10
*** haomaiwang has quit IRC08:10
*** _cjones_ has quit IRC08:12
*** sgotliv has joined #openstack-cinder08:13
openstackgerritJulien Danjou proposed openstack/cinder: tests: replace mox by mox3, clean out mox usage  https://review.openstack.org/14747608:16
*** mdbooth has quit IRC08:16
*** shyama has quit IRC08:19
*** mdbooth has joined #openstack-cinder08:22
*** dulek has joined #openstack-cinder08:23
*** e0ne has joined #openstack-cinder08:28
*** e0ne is now known as e0ne_08:28
*** alecv has quit IRC08:40
*** e0ne_ is now known as e0ne08:44
*** e0ne is now known as e0ne_08:45
*** e0ne_ is now known as e0ne08:46
*** jistr has joined #openstack-cinder08:56
*** aix has quit IRC08:59
*** deepakcs has joined #openstack-cinder08:59
*** dulek has quit IRC09:00
*** shyama has joined #openstack-cinder09:08
*** dobson has quit IRC09:11
*** agarciam1 has joined #openstack-cinder09:14
*** agarciam has quit IRC09:14
*** alecv has joined #openstack-cinder09:20
*** heyun has quit IRC09:21
openstackgerritYuriy Nesenenko proposed openstack/cinder: Fix unit tests spam output  https://review.openstack.org/17878309:25
*** alexpilotti has joined #openstack-cinder09:25
*** aix has joined #openstack-cinder09:26
*** markvoelker has quit IRC09:48
openstackgerritPetrut Lucian proposed openstack/cinder: Windows iSCSI: remove ensure_export  https://review.openstack.org/18008009:49
openstackgerritPetrut Lucian proposed openstack/cinder: Windows iSCSI: Add CHAP authentication support  https://review.openstack.org/18008109:49
*** ociuhandu has joined #openstack-cinder09:51
*** dobson has joined #openstack-cinder09:51
*** _cjones_ has joined #openstack-cinder09:56
*** _cjones_ has quit IRC10:01
*** tshefi has quit IRC10:01
*** bkopilov has quit IRC10:01
*** dobson has quit IRC10:03
*** shyama has quit IRC10:09
*** shyama has joined #openstack-cinder10:10
*** aix has quit IRC10:12
*** dobson has joined #openstack-cinder10:13
*** tshefi has joined #openstack-cinder10:17
*** bkopilov has joined #openstack-cinder10:17
*** tobe has joined #openstack-cinder10:18
*** kbyrne has quit IRC10:18
*** dulek has joined #openstack-cinder10:19
*** IanGovett has joined #openstack-cinder10:19
*** dims has joined #openstack-cinder10:20
*** kbyrne has joined #openstack-cinder10:20
*** dims_ has joined #openstack-cinder10:22
*** aix has joined #openstack-cinder10:25
*** dims has quit IRC10:26
*** e0ne is now known as e0ne_10:35
*** e0ne_ is now known as e0ne10:36
*** e0ne is now known as e0ne_10:56
*** ebalduf has joined #openstack-cinder11:00
*** lpabon has joined #openstack-cinder11:03
*** swamireddy1 has joined #openstack-cinder11:04
swamireddy1: Is ´volume forec detach´ supported?11:04
*** dobson has quit IRC11:04
*** ebalduf has quit IRC11:04
*** IanGovett1 has joined #openstack-cinder11:06
*** fanyaohong has joined #openstack-cinder11:08
*** IanGovett has quit IRC11:09
*** ankit_ag has quit IRC11:12
*** sandywalsh has quit IRC11:12
*** fanyaohong has quit IRC11:13
*** fanyaohong has joined #openstack-cinder11:13
*** sandywalsh has joined #openstack-cinder11:14
DuncanTswamireddy1: Not yet, the nova part isn't done11:17
DuncanTswamireddy1: Depending on your situation, state reset might help, but it might leave you with a broken system11:18
*** dobson has joined #openstack-cinder11:19
*** dencaval has joined #openstack-cinder11:22
*** alecv has quit IRC11:25
*** alecv has joined #openstack-cinder11:26
*** dobson has quit IRC11:36
*** dobson has joined #openstack-cinder11:37
*** ctina_ has joined #openstack-cinder11:44
*** agarciam1 has quit IRC11:44
*** _cjones_ has joined #openstack-cinder11:45
*** marcusvrn1 has joined #openstack-cinder11:46
*** marcusvrn has quit IRC11:48
*** HoloIRCUser1 has joined #openstack-cinder11:49
*** _cjones_ has quit IRC11:50
*** hodos has joined #openstack-cinder11:50
hodostrying to do cinder upload-to-image; getting broken pipe in cinder after a while; the process hangs11:51
*** markvoelker has joined #openstack-cinder11:51
*** e0ne_ is now known as e0ne11:51
hodosglance shows: ValueError: invalid literal for int() with base 16: ''11:51
hodosimage-list shows 'saving'11:51
hodoscinder list shows 'uploading'11:52
hodosall on devstack11:52
*** Tross has joined #openstack-cinder11:52
*** Tross1 has quit IRC11:54
*** Tross has quit IRC11:56
*** Tross1 has joined #openstack-cinder11:56
deepakcsDuncanT, Hi, when u get time, pls look at https://review.openstack.org/#/c/172373/ (cinder part). The nova part of spec is done (https://review.openstack.org/#/c/134517/) but john garbutt wanted the cinder part to get in first. thanks11:56
*** ganso has joined #openstack-cinder12:01
*** zhenguo has quit IRC12:07
*** agarciam has joined #openstack-cinder12:07
*** agarciam has quit IRC12:08
DuncanTdeepakcs: sure12:08
deepakcsDuncanT, thanks, ttyl12:15
*** deepakcs has quit IRC12:15
*** sgotliv has quit IRC12:22
*** primechuck has quit IRC12:23
*** sgotliv has joined #openstack-cinder12:24
*** HoloIRCUser1 has quit IRC12:27
*** HoloIRCUser has joined #openstack-cinder12:27
*** HoloIRCUser1 has joined #openstack-cinder12:28
*** HoloIRCUser has quit IRC12:28
openstackgerritPetrut Lucian proposed openstack/cinder: Windows iSCSI: remove ensure_export  https://review.openstack.org/18012312:30
*** bkopilov has quit IRC12:31
swamireddy1DuncanT:  Thank you....my case - even after detaching a volume of a instance, it shown as volume is in use state.12:32
swamireddy1DuncanT:  is this known issue in cinder?12:33
DuncanTswamireddy1: I'm guessing something failed in the detach.... have you checked to logs?12:33
openstackgerritPetrut Lucian proposed openstack/cinder: Windows iSCSI: remove ensure_export  https://review.openstack.org/18008012:34
openstackgerritPetrut Lucian proposed openstack/cinder: Windows iSCSI: Add CHAP authentication support  https://review.openstack.org/18008112:34
*** hodos has quit IRC12:36
swamireddy1DuncanT: OK, sure will check the log and confirm the status. Any other reason for this issue??12:37
DuncanTswamireddy1: The logs should at least give you a good hint12:38
swamireddy1DuncanT: Sure, will analyse the log fist... Thank you.12:38
swamireddy1DuncanT:  Any one working on volume force detach option? or its open to pickup?12:39
*** dobson has quit IRC12:39
DuncanTScottDA is working on it...12:40
DuncanTI'm sure he'd welcome help12:40
*** timcl has joined #openstack-cinder12:40
*** dims_ has quit IRC12:41
*** dims has joined #openstack-cinder12:41
*** IanGovett1 has quit IRC12:43
*** dobson has joined #openstack-cinder12:46
*** charlesr has joined #openstack-cinder12:48
openstackgerritVictor Stinner proposed openstack/cinder: Replace suds test dependency with suds-jurko  https://review.openstack.org/18013012:51
*** links has quit IRC12:52
scottdaswamireddy1: I've this BluePrint for fixing Cinder force-detach: https://blueprints.launchpad.net/cinder/+spec/implement-force-detach-for-safe-cleanup12:52
*** tshefi has quit IRC12:53
scottdaAnd this spec, please review and give feedback: https://review.openstack.org/#/c/172213/12:53
scottdaYou could also add 'affects me' to Nova side: https://bugs.launchpad.net/python-novaclient/+bug/1449223 https://bugs.launchpad.net/python-novaclient/+bug/144922112:54
openstackLaunchpad bug 1449223 in python-novaclient "Nova volume-detach lacks '--force' command for cleanup " [Undecided,New] - Assigned to Andrea Rosa (andrea-rosa-m)12:54
openstackLaunchpad bug 1449221 in OpenStack Compute (nova) "Nova volume-detach lacks '--force' command for cleanup" [Undecided,In progress] - Assigned to Andrea Rosa (andrea-rosa-m)12:54
*** agarciam has joined #openstack-cinder12:55
*** lpabon has quit IRC12:55
*** ctina_ has quit IRC13:03
*** eharney has joined #openstack-cinder13:05
*** changbl has quit IRC13:08
*** dustins has joined #openstack-cinder13:11
*** shyama has quit IRC13:12
*** Ctina has joined #openstack-cinder13:13
*** Yogi1 has joined #openstack-cinder13:14
*** rushil has joined #openstack-cinder13:15
*** BharatK has quit IRC13:17
*** fanyaohong has quit IRC13:19
*** primechuck has joined #openstack-cinder13:19
*** bswartz has quit IRC13:23
*** charlesr has quit IRC13:24
*** marcusvrn has joined #openstack-cinder13:25
*** marcusvrn1 has quit IRC13:26
*** BharatK has joined #openstack-cinder13:31
*** _cjones_ has joined #openstack-cinder13:34
*** BharatK has quit IRC13:36
*** _cjones_ has quit IRC13:39
*** tobe has quit IRC13:40
*** hodos has joined #openstack-cinder13:40
j_kingyay, talked to ayoung yesterday about implied trusts and he was happy. will finally have a nice solution for https://etherpad.openstack.org/p/implied-trusts13:42
swamireddy1DuncanT: Thanks..Will sync with scottda on this13:42
swamireddy1scottda: Hi13:43
j_kinger https://bugs.launchpad.net/cinder/+bug/129813513:43
openstackLaunchpad bug 1298135 in Cinder "Cinder should handle token expiration for long ops" [Medium,Confirmed] - Assigned to j_king (james-agentultra)13:43
swamireddy1scottda: I have updated the bug report as äffected me¨13:43
swamireddy1scottda:  so spec is pending for review13:44
swamireddy1scottda: will review the same13:44
*** rooneym has quit IRC13:45
*** BharatK has joined #openstack-cinder13:49
*** mriedem has joined #openstack-cinder13:51
*** ebalduf has joined #openstack-cinder13:55
DuncanTj_king: I don't understand that etherpad, and the whole thing seems complicated... not to mention an API change13:55
DuncanTj_king: What is wrong with cinder just getting the trust on behalf of the user?13:55
*** bswartz has joined #openstack-cinder13:56
hodosis volume_type mandatory?13:56
*** e0ne is now known as e0ne_13:56
j_kingDuncanT: the problem is at cinder-backup. if the backup process takes longer than the configured token expiration it will fail. the trusts API is supposed to delegate trust for this purpose but presently requires the user to create the trust13:57
hodoshow to create a volume with no volume type?13:57
*** HoloIRCUser1 has quit IRC13:57
*** dobson has quit IRC13:57
DuncanThodos: You can set a default type in cinder.conf13:57
*** rooneym has joined #openstack-cinder13:57
j_kingDuncanT: so the proposal allows a background process to create a trust on behalf of a user via a one-time authenticated request13:57
hodosDuncanT: default cannot be None13:58
scottdaswamireddy1: thansk13:58
DuncanTj_king: I am familiar with the problem. Cinder can request a trust right now with the token it is given13:58
DuncanThodos: You can't create a volume with no type, that doesn't make sense13:58
DuncanThodos: You can avoid having to specify one though13:58
DuncanTj_king: No need for anything external13:58
hodosDuncanT: OK13:58
j_kingDuncanT: I think the API required the user to create the trust between services? I'd have to re-check that.13:59
*** e0ne_ is now known as e0ne13:59
*** e0ne is now known as e0ne_13:59
DuncanTj_king: cinder can call keystone to do it.... the bit I couldn't fifure out is what roll(s) the trust needs to talk to swift14:00
*** e0ne_ is now known as e0ne14:00
*** e0ne is now known as e0ne_14:00
*** e0ne_ is now known as e0ne14:00
*** timcl1 has joined #openstack-cinder14:02
j_kingDuncanT: how would you go about setting it up? apparently this isn't the only project asking for this feature from keystone. I hope I don't have it all wrong. :/14:04
*** timcl has quit IRC14:04
*** dobson has joined #openstack-cinder14:04
DuncanTj_king: cinder already has a token, and knows the keystone URL, so it literally just calls the 'create trust' API exactly as a user would14:06
*** esker has joined #openstack-cinder14:06
*** setmason has joined #openstack-cinder14:08
*** vmtrooper has joined #openstack-cinder14:09
*** mdenny has joined #openstack-cinder14:10
*** timcl has joined #openstack-cinder14:11
*** timcl1 has quit IRC14:12
openstackgerritTom Barron proposed openstack/cinder: Add standard QoS spec support to cDOT drivers  https://review.openstack.org/17935214:13
*** anuragpalsule has quit IRC14:14
*** agarciam has quit IRC14:14
*** asselin has joined #openstack-cinder14:15
jdandreaWhen volume creation fails, is there some way to determine the cause programmatically? I'm looking for a fail_reason field in the volume object, as there is in backup objects, but not seeing it. (Might it be something I could add and, if so, where do I get the error info from?)14:15
openstackgerritYucong Feng proposed openstack/cinder: Removing sleep between when a command is sent and 'YES' is sent  https://review.openstack.org/17733514:16
hodosDuncanT: does volume upload-to-image work at all for NFS backend ? I'm using generic driver on devstack and it seems to hang.....14:17
DuncanTjdandrea: No, it isn't there. There's a summit session about the best to fix it14:18
DuncanThodos: Should be, yeah. Can be very slow though14:18
jdandreaDuncanT: That's terrific. I will be in Vancouver. Hopefully I can be there for that session.14:18
*** thangp has joined #openstack-cinder14:18
DuncanTjdandrea: It's a cross-project session I think14:18
jdandreaDuncanT: Thanks. If there's a link, please advise. I'll take note.14:18
hodosDuncanT: I'm not sure how to pass Tempest with thirdparty nfs driver then14:18
* jdandrea also goes looking14:18
*** dulek has quit IRC14:19
DuncanThodos: It passes in the gate AFAIK. I've never tried it14:19
*** lifeless has quit IRC14:20
*** swamireddy1 has quit IRC14:25
openstackgerritGrzegorz Grasza (xek) proposed openstack/cinder: Service object  https://review.openstack.org/16041714:26
*** lifeless has joined #openstack-cinder14:27
*** setmason has quit IRC14:28
*** vmtrooper has quit IRC14:28
*** annegentle has joined #openstack-cinder14:29
*** IanGovett has joined #openstack-cinder14:30
*** mtanino has joined #openstack-cinder14:32
*** dulek has joined #openstack-cinder14:34
openstackgerritEric Harney proposed openstack/cinder: Fix 'driver is uninitialize' typo  https://review.openstack.org/18017714:38
*** rooneym has quit IRC14:38
*** rooneym has joined #openstack-cinder14:39
*** emagana has joined #openstack-cinder14:40
*** MentalRay has joined #openstack-cinder14:45
jdandreaDuncanT: Is there a bug for the missing cinder volume error_reason? (Still looking for it so I can keep apprised and pitch in.)14:46
DuncanTjdandrea: Not aware of a bug, since that is generally considered to be a poor method of passing status back to the user14:47
DuncanTjdandrea: The summit session is about how to pass asyncronous errors back properly14:47
jdandreaDuncanT: That's what I'm talking about, the async errors.14:48
*** coolsvap is now known as coolsvap|afk14:48
jdandreaDuncanT: I ask because I'm assigned to this: https://bugs.launchpad.net/heat/+bug/1450861 ... and for Stack Orchestration there is no sensible error message. It's due to "Unknown" which sends folks off on the proverbial wild goose chase. :)14:48
openstackLaunchpad bug 1450861 in heat "Cinder create error reason not visible" [Medium,Triaged] - Assigned to Joe D'Andrea (joedandrea)14:48
*** madskier has joined #openstack-cinder14:48
*** madskier has quit IRC14:48
openstackgerritJohn Griffith proposed openstack/python-cinderclient: Introduce functional testing to cinderclient  https://review.openstack.org/17551214:49
*** madskier has joined #openstack-cinder14:49
openstackgerritGorka Eguileor proposed openstack/cinder: Preserve usage and reservations on quota deletion  https://review.openstack.org/16272214:49
*** rongze has joined #openstack-cinder14:49
openstackgerritJohn Griffith proposed openstack/python-cinderclient: Add functional post test_hook  https://review.openstack.org/17552114:49
*** madskier has quit IRC14:50
*** shyama has joined #openstack-cinder14:50
*** crose has quit IRC14:50
DuncanTjdandrea: Part of the problem is that different providers have different requirements as to what level of detail they want to pass back to tenants.... private cloud might be fine with lots of details, where as prublic cloud might not want to pass on more than 'something went wrong, try again later'14:51
*** crose has joined #openstack-cinder14:51
* jdandrea nods14:52
*** rooneym_ has joined #openstack-cinder14:52
DuncanTjdandrea: Nobody has come up with a good way to square that circle.14:53
smcginnisAnyone know what's going on with gate?14:53
*** setmason has joined #openstack-cinder14:53
*** agarciam has joined #openstack-cinder14:53
jdandreaDuncanT: Aye. Meanwhile, "Unknown" doesn't help *anyone*. (I know, preaching to the choir.) I wonder if there are other examples of this conundrum elsewhere in OS that have been broached? (Seriously.)14:54
*** cdelatte has quit IRC14:54
*** delattec has quit IRC14:54
* jdandrea might go on a mission :)14:54
*** ronis has quit IRC14:55
DuncanTjdandrea: I'm not aware of anything, no. Nova seems mostly to have said 'screw the sensibilities of public cloud operators', which is one approach, but I'd like to at least consider the alternatives14:56
jdandreaI would too.14:56
*** ebalduf has quit IRC14:56
DuncanTjdandrea: Maybe build a hiarachy of detail in errors, and let the cloud provider pick what levels get passed to the tenants? (Send details events for all errors to the admin queue though, so they can be found somewhere14:57
jdandreaDuncanT: That's where I was headed as well, agreed.14:58
*** reed_ has joined #openstack-cinder14:59
DuncanTjdandrea: Any chance you'll get time to write up an idea before the summit?14:59
*** rooneym_ is now known as mrooney14:59
jdandreaDuncanT: I was reminded of NSError (puts on iOS developer hat momentarily). Note the different Localized Error Description properties: https://developer.apple.com/library/ios/documentation/Cocoa/Reference/Foundation/Classes/NSError_Class/15:00
jdandreaSo imagine a description being visible, but the reason being something that can be optionally visible only to admins (?).15:01
*** delattec has joined #openstack-cinder15:02
*** cdelatte has joined #openstack-cinder15:02
*** ebalduf has joined #openstack-cinder15:02
DuncanTjdandrea: Maybe have a specific role to get the detailed reason, that way private clouds that want to can give that role to all tenants?15:03
jdandreaBasically offer a Description, a Failure Reason, and a Recovery Suggestion. (I sense this construct might be better served by Oslo, but ...)15:03
jdandreaDuncanT: Maybe. I often wonder about creating roles for things like that. I mean it's certainly doable. Is that what a role should be used for? (Perhaps it's the best thing we have going.)15:04
*** _cjones_ has joined #openstack-cinder15:05
jdandreaDuncanT: Where can I contribute my writeup? Is there an Etherpad for that session?15:05
DuncanTRoles are just labels in the end15:05
DuncanTOne sec, I'll check15:06
*** rmesta has joined #openstack-cinder15:06
jdandrea*nods* and perhaps we allow optional association of one or more roles with a given error property as part of the config.15:07
DuncanThttps://etherpad.openstack.org/p/liberty-cinder-async-reporting - not much there yet, feel free to add15:07
jdandreaThanks. :)15:07
*** mrooney has quit IRC15:09
*** _cjones_ has quit IRC15:09
*** MentalRay has quit IRC15:10
openstackgerritRushil Chugh proposed openstack/cinder: Avoid LUN ID collisions in NetApp iSCSI drivers  https://review.openstack.org/17923915:11
*** jaypipes has quit IRC15:13
*** lpetrut has quit IRC15:14
*** asselin has quit IRC15:15
*** anshul has quit IRC15:16
*** mrooney has joined #openstack-cinder15:18
*** nlevinki has joined #openstack-cinder15:26
*** jdurgin1 has joined #openstack-cinder15:27
*** nkrinner has quit IRC15:28
*** vmtrooper has joined #openstack-cinder15:32
*** dims_ has joined #openstack-cinder15:32
*** vmtrooper has quit IRC15:32
openstackgerritTom Barron proposed openstack/cinder: Use nfs_oversub_ratio when reporting pool capacity  https://review.openstack.org/18019915:33
tbarronping mkoderer15:34
*** dims has quit IRC15:34
*** MentalRay has joined #openstack-cinder15:34
thingeeDuncanT, jdandrea: there was a plan to talk about it, but I think scottda said force-delete and force-detach satisfies all cases. So there is no plan to talk about it at the summit.15:35
DuncanTthingee: async error reporting is totally different to the stuff Scott is working on15:35
scottdaWe can talk about this is need be, but I think landing nova + cinder for force-detach will solve most problems15:35
scottdathingee: jdandrea ^^15:36
DuncanTthingee: And we're finding corner cases that are pretty hard to get a good answer to15:36
*** dannywilson has joined #openstack-cinder15:36
thingeeDuncanT: I'm talking about the original question about volumes stuck in creating state.15:36
DuncanTthingee: Fair enough15:37
*** rongze has quit IRC15:38
*** tsekiyama has joined #openstack-cinder15:40
*** nikesh has joined #openstack-cinder15:40
*** rongze_ has joined #openstack-cinder15:40
*** nikesh has quit IRC15:40
*** dannywilson has quit IRC15:41
*** nikeshm has joined #openstack-cinder15:41
*** dannywilson has joined #openstack-cinder15:42
*** rongze has joined #openstack-cinder15:42
*** shakamunyi has joined #openstack-cinder15:43
*** Ctina is now known as ctina15:43
*** reed_ has quit IRC15:45
*** changbl has joined #openstack-cinder15:45
*** rongze_ has quit IRC15:46
*** hemnafk is now known as hemna15:48
*** anshul has joined #openstack-cinder15:50
*** markstur has quit IRC15:50
thingeejungleboyj: ping15:51
openstackgerritYuriy Nesenenko proposed openstack/cinder: Fix unit tests spam output  https://review.openstack.org/17878315:52
*** vmtrooper has joined #openstack-cinder15:52
*** cdelatte has quit IRC15:56
*** delattec has quit IRC15:56
*** cdelatte has joined #openstack-cinder15:57
*** vmtrooper has quit IRC15:57
*** cdelatte has quit IRC15:58
*** cdelatte has joined #openstack-cinder15:58
*** timcl has quit IRC15:59
*** Yogi1 has quit IRC15:59
*** agarciam has quit IRC16:00
*** chlong has joined #openstack-cinder16:01
*** daneyon has joined #openstack-cinder16:02
*** nlevinki has quit IRC16:02
*** daneyon_ has joined #openstack-cinder16:03
*** avishay_ has quit IRC16:04
*** _cjones_ has joined #openstack-cinder16:04
*** daneyon has quit IRC16:06
*** madskier has joined #openstack-cinder16:07
*** Apoorva has joined #openstack-cinder16:09
openstackgerritIvan Kolodyazhny proposed openstack/cinder-specs: Volume get and list should be able get detailed views  https://review.openstack.org/15629216:15
e0nethingee: hi. could you please take a look on spec ^^ when you'll have a time?16:15
*** chlong has quit IRC16:18
*** _cjones_ has quit IRC16:20
*** daneyon_ has quit IRC16:21
*** daneyon has joined #openstack-cinder16:21
jdandreathingee: No plan to talk about volume creation errors?16:22
*** annegentle has quit IRC16:23
thingeejdandrea: Not sure if you read what I said earlier, but scottda was convinced force-delete and force-detach solved the problem.16:23
jdandreaI'm not just talking about volumes stuck in creation state though. I'm talking about something broader.16:23
thingeejdandrea: ok, can you talk to scottda?16:24
jdandreathingee: Ah. That *is* important. What I'm talking about is async error reporting in a general sense (perhaps even something that belongs in oslo longer-term).16:24
thingeejdandrea: oh. Can you explain more? according the etherpad I said this should be cross project16:24
thingeejdandrea: everyone else agreed as well16:25
thingeejdandrea: https://etherpad.openstack.org/p/cinder-liberty-proposed-sessions16:25
jdandreascottda: I'm tasked with tackling this. https://bugs.launchpad.net/heat/+bug/1450861 ...16:25
openstackLaunchpad bug 1450861 in heat "Cinder create error reason not visible" [Medium,Triaged] - Assigned to Joe D'Andrea (joedandrea)16:25
*** annegentle has joined #openstack-cinder16:25
*** _cjones_ has joined #openstack-cinder16:25
jdandrea... and this led to a chat with DuncanT which led me to this, which I'm interested in helping with: https://etherpad.openstack.org/p/liberty-cinder-async-reporting16:25
*** jistr has quit IRC16:26
dulekthingee: Hi, I've seen session about tasks resumptions is scheduled before active/active c-vol. The problem is it should be the other way as first one depends on second one.16:26
jdandrea... but I see it's not accepted at this time. Not sure how to best proceed then. :-o16:26
thingeejdandrea: can we make this a cross project summit session?16:27
jdandreathingee: That's what I thought it was?16:27
*** timcl has joined #openstack-cinder16:27
thingeedulek: thanks, I will fix that16:27
dulekthingee: At active/active c-vol session a decision on TaskFlow will be made which is very relevant to approach on task resumptions. Also use of tooz will be discussed.16:27
dulekthingee: Thanks!16:27
thingeejdandrea: I'm confused now16:27
*** annegent_ has joined #openstack-cinder16:27
*** annegentle has quit IRC16:28
jdandreathingee: Correction: DuncanT thought it *might* be cross-project (not definitively so). I definitely think it deserves to be cross-project. It screams Oslo to me, in fact.16:30
*** Yogi1 has joined #openstack-cinder16:30
thingeedulek: done http://libertydesignsummit.sched.org/type/design+summit/Cinder#.VUjv7d-c2fY16:31
jdandreaAs Duncan stated earlier, different providers have different requirements as to the level of detail they want to pass back to tenants. Squaring that circle has proven tricky. I have some ideas I'd like to float though. Perhaps they'd be helpful in getting a conversation going.16:31
thingeedulek: and etherpad updated16:31
asselin__nikeshm, hi...still having issues?16:31
e0nedulek: thanks for noting this! i missed it:(16:31
thingeejdandrea: right. I was saying it needs to be in the cross project track. It being in Cinder tracks is not cross project, so you'll need to propose it through whatever etherpad has been setup for cross project16:32
jdandreathingee: Ahh!16:32
*** ctina has quit IRC16:32
* jdandrea is learning16:32
jdandreaNow I see the confusion. :)16:32
openstackgerritEric Brown proposed openstack/cinder: VMware: insecure option should be exposed  https://review.openstack.org/17999116:32
* jdandrea wonders where that etherpad is ...16:32
*** fthiagogv has joined #openstack-cinder16:32
*** ctina has joined #openstack-cinder16:32
*** annegent_ has quit IRC16:33
*** dulek has quit IRC16:34
thingeejdandrea: http://lists.openstack.org/pipermail/openstack-dev/2015-April/061070.html16:34
thingeedoh might've passed to suggest sessions16:34
openstackgerritYuriy Nesenenko proposed openstack/cinder: Fix unit tests spam output  https://review.openstack.org/17878316:35
jdandreathingee: tyvm16:35
e0nethingee: http://libertydesignsummit.sched.org/event/57b44975517778e6450613b5c0ca2681#.VUjxRa08aHo can you add some tag to indicate that it's related both to cinde and ironic?16:35
thingeee0ne: yes16:36
thingeee0ne: done16:36
e0nethingee: thanks! i chatted with ironic team yesterday. thay are going to attenf it16:37
jdandreathingee: I see a few proposals already there. Good! Also, Async API Error Reporting (ameade) is already there (but not bolded).16:37
thingeee0ne: http://libertydesignsummit.sched.org/overview/type/design+summit/Ironic#.VUjxuN-c2fY16:37
thingeee0ne: nice16:37
*** leeantho has joined #openstack-cinder16:37
e0nehemna: ^^16:37
*** jaypipes has joined #openstack-cinder16:38
thingeejdandrea: yeah, originally ameade suggested it in cinder tracks and then the cinder community suggested it be cross project. glad to see it's in there.16:38
thingeebut there should be a follow up with ttx16:38
ameadejdandrea, thingee: it is being merged with the talk on row 1016:38
scottdajdandrea: My focus is the APIs to fix volumes stuck in attaching/detaching with a 'nova volume-detach --force' option and better Cinder force_delete(016:38
ameadeas they are really trying to solve the same problem16:38
scottdaSo I'm interested in this Async Error reporting, but I'm not actively doing anything with this.16:39
jdandreaameade: Ah, I see your note on row 10 now.16:39
jdandreaI'm interested on behalf of the Heat team.16:39
thingeescottda: sorry I confused your work with something else apparently16:39
hodosvolume upload-to-image doesn't seem to ever end on 1Gib NFS volume16:39
hodosany1 has experience ?16:40
jdandreaIn particular as it relates to Cinder (because that's where I first noticed it, with volume creation errors having a reason of "Unknown"), but it has broader reach, for certain.16:40
hodosglance api shows some auth exceptions...16:40
jdandreascottda: Got it. That's good too!16:40
jdandrea(actually extremely good)16:40
ameadejdandrea: good, i'm working with Angus16:41
scottdajdandrea: Please feel free to review and add Heat to Nova !16:41
ameadejdandrea:  this is something i've been trying to push for solving a couple of releases now16:41
ameadejdandrea: now i have time to actually do the work16:41
jdandreascottda: Add Heat to Nova?16:42
thingeetbarron or eharney can you help hodos?16:42
scottdaYes, nova side are bugs, so it you click "affects me" it will help Nova team to see this as a priority16:43
e0nehodos: i've got similar issue with swift backend for glance. i didn't find glance issue now:(16:43
jdandreascottda: Context lost. Not sure which bug ... ?16:43
eharneythingee: hodos: i haven't tried that in a while, let me look after i get off of a call16:44
hodoseharney: thanks. i waited for couple of hours; eventually volume status changed from 'uploading' to 'available' but image is still 'saving'16:45
hemnae0ne, ?16:45
e0nehemna: it was fyi about design session16:46
scottdajdandrea: https://bugs.launchpad.net/python-novaclient/+bug/1449223 and https://bugs.launchpad.net/python-novaclient/+bug/144922116:46
openstackLaunchpad bug 1449223 in python-novaclient "Nova volume-detach lacks '--force' command for cleanup " [Undecided,New] - Assigned to Andrea Rosa (andrea-rosa-m)16:46
openstackLaunchpad bug 1449221 in OpenStack Compute (nova) "Nova volume-detach lacks '--force' command for cleanup" [Undecided,In progress] - Assigned to Andrea Rosa (andrea-rosa-m)16:46
hemnae0ne, ok I guess I need to sign up for that one :P16:46
e0nehemna: or find a good pub:)16:47
*** jordanP has quit IRC16:47
jdandreascottda: tx16:47
hemnastill sitting there with a +216:47
hemnaI pinged infra about it this morning, no joy.16:47
*** anuragpalsule has joined #openstack-cinder16:48
e0nehemna: i'll update my flask PoC tomorrow. i rethink it after looking on sahara. it will very simple16:50
jdandreaameade: I checked those bugs out, but I'm not sure if heat is affected (meaning there *might* be but I'm just dense and not seeing it yet). I do think it's needed though.16:50
tbarronhodos: reading backlog now16:50
hemnae0ne, sahara?16:50
e0nes/will/will be16:50
e0nehemna: yep. they use flask for api too16:50
hemnaah ok cool.16:51
hemnaI think for now we can just ignore the keystone part16:51
e0neflask+keystonemidlware ~= 3 line of code16:51
hemnaand just an API service running in front of the 3 main pieces of os-brick16:51
hemnaget_connector, connect_volume, disconnect_volume16:51
hemnaI'm just not sure keystone is the right thing though16:51
e0negot your point. i'll provide you update tomorrow16:52
*** IlyaG has joined #openstack-cinder16:53
*** mk has joined #openstack-cinder16:54
*** mk is now known as Guest3303916:54
*** sgotliv has quit IRC16:54
*** markus_z has quit IRC16:56
*** Guest33039 has quit IRC16:56
*** muratk has joined #openstack-cinder16:56
*** dustins has quit IRC16:56
*** BharatK has quit IRC16:57
*** yuriy_n17 has quit IRC16:58
*** deepakcs has joined #openstack-cinder16:59
*** patrickeast has joined #openstack-cinder17:00
*** ebalduf has quit IRC17:02
*** dustins_ has joined #openstack-cinder17:02
*** ebalduf has joined #openstack-cinder17:03
*** dustins_ is now known as dustins17:03
-openstackstatus- NOTICE: zuul has been restarted to troubleshoot an issue, gerrit events between 15:00-17:00 utc were lost and changes updated or approved during that time will need to be rechecked or have their approval votes readded to trigger testing17:03
*** dustins has quit IRC17:04
*** e0ne has quit IRC17:05
*** anuragpalsule1 has joined #openstack-cinder17:06
*** jdurgin1 has quit IRC17:06
tbarronhodos: I haven't been running upload-to-image with the generic nfs driver, but if you post c-vol and g-reg logs I'll take a look.17:06
*** madskier has quit IRC17:06
*** anuragpalsule has quit IRC17:07
hodostbarron: is it OK to remove some tests from thirdparty CI testing17:08
*** harlowja has quit IRC17:08
hodostbarron: cause i see that some providers do not run all 304 tests17:08
*** harlowja has joined #openstack-cinder17:09
*** swamireddy1 has joined #openstack-cinder17:09
tbarronhodos: I'm not authoritative on your question, but my understanding is that we can't just remove tests from our CI because they don't work with our drivers :-)17:10
*** anuragpalsule1 has quit IRC17:10
hodostbarron: well I can see that some of the NFS-extended drivers run only 241 tests and upload is not among them :)17:11
*** madskier has joined #openstack-cinder17:11
*** rongze has quit IRC17:12
*** dustins has joined #openstack-cinder17:13
*** vilobhmm1 has joined #openstack-cinder17:14
*** vilobhmm1 has quit IRC17:15
DuncanThodos: Some people did disable tests, they aren't supposed to. If you spot any missing, please file a bug17:15
*** vilobhmm1 has joined #openstack-cinder17:16
hodosDuncanT: OK17:16
hodostbarron: http://paste.openstack.org/show/215022/ this is c-api.log17:17
hodostbarron: http://paste.openstack.org/show/215022/ this is c-vol.log17:17
hodostbarron: http://paste.openstack.org/show/215023/ this is g-api17:18
tbarronhodos: I'm only seeing 285 lines, which gets part way through the print out of the config options :-)17:23
*** IlyaG has quit IRC17:25
tbarronhodos: also pls post your cinder.conf and glance*.conf files17:25
*** avishay_ has joined #openstack-cinder17:31
*** bkopilov has joined #openstack-cinder17:32
*** dims_ has quit IRC17:33
*** dims has joined #openstack-cinder17:34
*** shyama has quit IRC17:38
*** shyama has joined #openstack-cinder17:39
*** melwitt_ has joined #openstack-cinder17:42
*** melwitt_ has quit IRC17:42
*** bkopilov has quit IRC17:45
*** bkopilov has joined #openstack-cinder17:45
ameadejdandrea: you mean to ping scottda?17:48
*** e0ne has joined #openstack-cinder17:50
*** e0ne is now known as e0ne_17:50
*** e0ne_ is now known as e0ne17:51
ameadehodos: I believe the plan is to have everyone run all 304 tests but it is not a hard requirement yet17:51
ameadehodos: if something doesn't work on a backend then it is okay to add a skip to that test and file a launchpad bug17:51
ameadehodos: but it should be addressed promptly17:52
tbarronhodos: ameade knows much more about our CI than I do.17:52
*** vmtrooper has joined #openstack-cinder17:52
smcginnisIf any cores have time - would love to get this through and move on to the next one in the series: https://review.openstack.org/#/c/173368/17:53
asselin__ameade, tbarron keep in mind there are some bugs that prevent running all 304. e.g. https://review.openstack.org/#/c/174656/17:53
ameadeasselin__: good to know thanks17:53
hodosameade: is there a config to skip a test?17:54
tbarronhodos: and asselin is your guy for CI in general!17:54
ameadehodos: we've been doing it like this export DEVSTACK_GATE_TEMPEST_REGEX='tempest.api.volume.(?!.*(test_volume_upload|test_volume_backup))'17:55
ameadehodos: i've seen others turn features off in tempest.conf17:55
ameadeasselin__: curious how you do it17:55
asselin__ameade, we use regex17:55
jdandreaameade: Possibly. Might have my wires crossed. :)17:56
asselin__ameade, how do you turn off in temest.conf?17:56
hodosameade: thanks17:56
ameadebackup = False17:56
ameadeapi_extensions = all17:56
*** vmtrooper has quit IRC17:57
ameadeyou can do '17:57
ameadesnapshot = False17:57
asselin__ameade, interesting...17:57
ameadei have everything runnin on the FC setup except nova ec2 tests17:57
ameadeec2_api = False17:57
hodostbarron: here's cinder.conf http://paste.openstack.org/show/215035/17:57
ameadewith just a REGEX of 'volume'17:58
hodostbarron: glance is default17:58
*** timcl has quit IRC17:59
*** deepakcs has quit IRC17:59
tbarronhodos: so you see the issue with both ns_nfs backend and nfsv4 backend ?17:59
hodostbarron: both i tried nexenta first then went with generic nfs18:00
tbarronhodos: and they are using the same export: /home/ubuntu/shares.cfg18:00
*** ganso_ has joined #openstack-cinder18:00
hodostbarron: yes, the same share18:00
asselin__ameade, are those the boto thirdparty tests?18:00
ameadeasselin__: yeah18:01
asselin__ameade, we skip those too...will try your way18:02
*** dannywilson has quit IRC18:03
ameadeasselin__: I saw datera and solidfire dont seem to have it set to skip them but it skips them anyways, i think they are pointing to ami files that arent actually there so tempest skips18:03
mtreinishasselin__: btw, I would love to see tempest patches to cleanup the skip logic on those dumb boto tests :)18:05
mtreinishheh, I personally can't bring myself to look at that code to figure out how to fix it. I just get angry every time I do18:05
DuncanTsmcginnis: Done18:07
*** patrickeast_ has joined #openstack-cinder18:09
smcginnisDuncanT: Thanks!18:10
*** ganso has quit IRC18:11
jgriffithameade: the test is run, but it's not run with the extra image-repo declaration18:11
*** patrickeast has quit IRC18:11
*** patrickeast_ is now known as patrickeast18:11
*** madskier has quit IRC18:11
*** shyama has quit IRC18:11
*** barra204 has joined #openstack-cinder18:12
*** bkopilov has quit IRC18:14
*** vmtrooper has joined #openstack-cinder18:14
*** bkopilov has joined #openstack-cinder18:15
*** timcl has joined #openstack-cinder18:15
*** MentalRay has quit IRC18:16
*** ebalduf has quit IRC18:17
thingeejgriffith: from the discussions at the last midcycle meetup, do we want to stop with reset state? I just saw this https://review.openstack.org/#/c/164309/18:18
*** vmtrooper has quit IRC18:20
tbarronhodos: are you running swift in your devstack?18:21
*** bkopilov has quit IRC18:21
*** IlyaG has joined #openstack-cinder18:22
*** vmtrooper has joined #openstack-cinder18:24
*** avishay_ has quit IRC18:26
*** barra204 has quit IRC18:27
*** logan2 has quit IRC18:28
*** jdandrea has quit IRC18:30
*** logan2 has joined #openstack-cinder18:31
hodostbarron: nope18:32
tbarronhodos: so glance is backed by local storage?18:32
hodostbarron: yes18:33
*** vilobhmm1 has quit IRC18:33
tbarronhodos: so I was expecting some problem with swift since cinder eventually gives up and marks the volume available if the glance upload times out18:35
*** vilobhmm1 has joined #openstack-cinder18:35
tbarronhodos: why it would time out with local storage I dunno18:35
*** Tross has joined #openstack-cinder18:36
tbarronhodos: your glance log stops at line 326, within a second of when then upload started.  still doing update_image_metadata.18:36
*** Tross1 has quit IRC18:37
*** ebalduf has joined #openstack-cinder18:38
e0nehemna: changes to requirements.txt is almost merged! but we've got ~330 jobs in zuul's queue18:38
hemnaheh yah, hurry up and wait.18:38
nikeshmasselin__ : bash install_master.sh is done,but on running nodepoold -d $DAEMON_ARGS, i was getting error "IOError: [Errno 2] No such file or directory: '/var/log/nodepool/image/image.log'"18:39
nikeshmasselin__ : then created that file18:40
*** j_king has quit IRC18:40
asselin__nikeshm, I haven't seen this error before18:40
tbarronhodos: your cinder log snippet starts a minute later (!), stops at line 285, and has nothing in it about the upload-to-image18:40
asselin__echo $DAEMON_ARGS18:40
asselin__also, which user are you? sudo su - nodepool18:40
*** j_king has joined #openstack-cinder18:40
nikeshmasselin__ : echo $DAEMON_ARGS                                                 -c /etc/nodepool/nodepool.yaml -l /etc/nodepool/logging.conf18:41
nikeshmsudo su - nodepool18:41
*** avishay_ has joined #openstack-cinder18:41
*** barra204 has joined #openstack-cinder18:42
asselin__nikeshm, I don't have that folder /var/log/nodepool/image/18:43
asselin__maybe it's new18:43
patrickeastnikeshm: asselin__ : I had some trouble running nodepool with that command, i ended up using ‘nodepoold -d 2>&1 | tee /var/log/nodepool/nodepool.log’ instead, was too lazy to debug why the logging config wasn’t working18:43
*** vmtrooper has quit IRC18:44
nikeshmasselin__ : after creating file manually when i am running nodepoold -d $DAEMON_ARGS, i am getting below error http://paste.openstack.org/show/215042/18:45
asselin__nikeshm, ok those are errors in your nodepool.yaml file18:46
*** muratk has quit IRC18:46
*** fthiagogv has quit IRC18:46
nikeshmasselin__ : this is my yaml file http://paste.openstack.org/show/215044/18:47
asselin__nikeshm, delete 24-2818:47
*** sgotliv has joined #openstack-cinder18:47
asselin__b/c you don't have a provider for that. but you can add it later.18:47
nikeshmasselin__ : what about 18-20, any change there18:48
nikeshmmy image name is trusty18:49
nikeshmso what should i give in name,image and min-ready18:50
asselin__those are fine18:50
asselin__the only other thing to mention is to use DIB to create the image....18:50
asselin__I found it simpler...but didn't update the sample18:51
*** vmtrooper has joined #openstack-cinder18:51
*** swamireddy1 has quit IRC18:51
ameadejgriffith: oh ur right, i hadn't looked a few weeks18:52
asselin__nikeshm, http://paste.openstack.org/show/215048/18:52
ameadejgriffith: did you have to tweak anything in particular for that test to work?18:53
asselin__nikeshm, strongly recommend that...espeically if you're just starting...and avoid the future switching costs18:53
asselin__nikeshm, and it's much simpler and easier that the old way18:54
*** nikeshm has quit IRC18:54
*** mrooney has quit IRC18:57
*** merooney has joined #openstack-cinder18:58
*** annegentle has joined #openstack-cinder18:59
*** IlyaG has quit IRC19:00
*** nikeshm has joined #openstack-cinder19:00
*** vmtrooper has quit IRC19:01
*** lpetrut has joined #openstack-cinder19:01
*** Longgeek_ has quit IRC19:02
*** rooneym has quit IRC19:03
lpetrutthingee: Hi19:04
*** IlyaG has joined #openstack-cinder19:05
*** jistr has joined #openstack-cinder19:08
*** sgotliv has quit IRC19:08
*** lpetrut has left #openstack-cinder19:09
*** avishay_ has quit IRC19:10
*** anshul has quit IRC19:10
*** Yogi11 has joined #openstack-cinder19:11
*** vilobhmm1 has quit IRC19:12
*** vilobhmm1 has joined #openstack-cinder19:13
*** sgotliv has joined #openstack-cinder19:13
*** Yogi1 has quit IRC19:13
*** Yogi12 has joined #openstack-cinder19:13
*** EmilienM is now known as EmilienM|afk19:15
*** Yogi11 has quit IRC19:16
*** sjcherry has joined #openstack-cinder19:26
*** annegentle has quit IRC19:28
nikeshmasselin__ : i updated the file in my git data repo "https://github.com/nikesh-mahalka/os-ext-testing-data",now do i have to again run "bash install_master.sh"19:32
nikeshmin first run it asked data repo url19:32
nikeshmin other runs,its not asking19:32
asselin__nikeshm, yes. run that again...it's idepempotent19:33
asselin__it only asks if the repo is not available locally19:34
*** mtanino has quit IRC19:34
asselin__nikeshm, be sure not to put your private key in github19:34
*** jungleboyj has quit IRC19:34
jgriffithameade: I did not, I just happened to never specify the image location, so never hit the problem19:37
jgriffiththingee: I dunno, seems scottda and others would really like to clean it up / keep it, so that's cool with me19:37
jgriffiththingee: I do think it's pretty abused at this point though19:37
*** ebalduf has quit IRC19:38
thingeejgriffith, scottda: yeah I thought we were in agreement at the mid cycle meetup to stop these and actually document that we don't recommend it.19:38
*** madskier has joined #openstack-cinder19:39
scottdaIt is documented in the reset-state docstring and help....19:39
*** ebalduf has joined #openstack-cinder19:40
scottdaIdeally, we find ways to fix things everywhere, i.e. force-detach and force-delete...19:42
*** emagana has quit IRC19:42
jgriffithscottda: yeah, personally I prefer the "force-xxx" over the reset business19:42
scottdaBut the question remains: Do we want an API for cleaning up the Cinder DB? Even if that is dangerous?19:42
*** emagana has joined #openstack-cinder19:42
scottdaSo how about this issue with changing reset-state for migration status? What does the fix for that look like?19:43
*** markstur has joined #openstack-cinder19:43
scottdaI guess there could be a 'cleanup-migration'19:43
*** vmtrooper has joined #openstack-cinder19:43
jgriffithscottda: or we could throw away the broken migration code19:43
jgriffithscottda: and actually make things "work" instead of band-aiding everything with reset-commands :)19:44
scottdaIt seems the patch could/should be: #1 Provide some tools/APIs to cleanup serious problems that exist today19:44
e0nescottda: we already have some functions to clean db in cinder-manage. i don't think that rest api for such actions is a good idea19:44
*** vmtrooper has quit IRC19:45
scottda#2 Fix things we know are wrong to prevent these issues19:45
jgriffithe0ne: +1 for cinder-manage19:45
scottda#3 re-factor and possible change the architecture as hemna and duncant are looking into for Cinder <-> Nova API19:45
guitarzanit's much easier to give access to the api than it is to cinder-manage19:46
scottdae0ne: I'm ok with getting rid for reset state. My current pet project is 'nova volume-detach --force' which would call cinder force_detach19:46
guitarzanis totally getting rid of reset-state a serious talk?19:47
jgriffithguitarzan: probably not19:47
guitarzanok, good19:47
jgriffithguitarzan: but whether it should continue to proliferate to every resource/possibility is more the question I think19:47
guitarzanwe don't really use it much19:47
nikeshmasselin__ : after running nodepoold -d $DAEMON_ARGS now its not giving error but its waiting to something19:47
guitarzanjgriffith: agreed19:47
jgriffithguitarzan: my opinion was that it's become abused, and shouldn't allow "any state" to be set19:48
guitarzanthat's an interesting question though19:48
e0nejgriffith, scottda: i'm not on the same page with you:(. are you going to remove reset-state features to add --force everywhere?19:48
jgriffithshould be very limited19:48
scottdaguitarzan: people use reset-state for a volume stuck in detaching, and then wonder why they cannot re-attach the volume. It is because Nova still sees it as attached. So that is broken.19:48
asselin__nikeshm, first thing is does is createan image19:48
scottdae0ne: No19:48
scottdabut Nova needs a force-detach19:49
*** jistr has quit IRC19:49
guitarzanscottda: the fight between nova and cinder is exactly why we (at RAX) need to keep reset-state19:49
guitarzanour servers dudes are already very gung ho at managing the novadb, but they don't have access to the cinder side19:49
scottdaIt seems reasonable to keep reset-state for Cinder DB only cleanup, to prevent admin from having to use SQL on the DB directly19:49
scottdaBut there should be an admin command for 'nova volume-detach --force' to cleanup everywhere: Nova, CinderDB , storage backend19:50
nikeshmasselin__ : image dir is empty  ls /opt/nodepool_dib/19:50
scottdaPlus I've filed Nova bugs for the '--force' option19:50
asselin__nikeshm, nodepool image-build dpc19:51
asselin__as nodepool user19:51
guitarzanI won't get into that... we fought battles with citrix about force removing a volume :)19:51
scottdahttps://bugs.launchpad.net/python-novaclient/+bug/1449223 and https://bugs.launchpad.net/python-novaclient/+bug/144922119:51
openstackLaunchpad bug 1449223 in python-novaclient "Nova volume-detach lacks '--force' command for cleanup " [Undecided,New] - Assigned to Andrea Rosa (andrea-rosa-m)19:51
openstackLaunchpad bug 1449221 in OpenStack Compute (nova) "Nova volume-detach lacks '--force' command for cleanup" [Undecided,In progress] - Assigned to Andrea Rosa (andrea-rosa-m)19:51
* scottda should probably send this all to ML, as he has posted this 3-4X today19:51
nikeshmasselin__ : ok i run nodepool image-build dpc as nodepool user,BTW it should be done by "nodepoold -d $DAEMON_ARGS"19:55
nikeshmis it not19:55
e0nescottda: i'm definitely not on the same page with you in scope of your issue because i'm not very fimilar with it.19:55
*** crose has quit IRC19:56
scottdaJust in the last couple days there have been several questions from people about volumes stuck in detaching, and how to clean them up. It seems to be a common problem.19:56
e0nescottda: just 2c from my side: imo (it's only my opinion) we need to re-thing and update docs what  attach/dettach is from a cinder point of view19:56
scottdaAnd thingee got feedback at OPs meetup that they want something to help with this as well.19:56
jgriffithscottda: for the record I hear "feedback from OPs meetup" an awful lot these days, but I'd sure like to see more of this "feedback" public and in context19:57
scottdae0ne: I agree completely that the entire Cinder <-> Nova API needs scrutiny, fixing, and docs could use a cleanup as well.19:57
jgriffithscottda: I'm certainly not arguing one way or the other at this point, just asking that we don't continue abusing the hell out of the call19:57
scottdajgriffith: OK, that's one source. But it seems there have been numerous patches proposed, IRC questions, and ML questions about volumes stuck in detaching.19:58
guitarzanthe alternate view is to please leave the tools necessary to use this stuff in place :)19:58
guitarzaneven if you think they're ugly19:58
jgriffithscottda: yeah, not disagreeing there19:58
jgriffithguitarzan: yeah, not saying they should be removed or anyhting like that19:58
jgriffithguitarzan: I was just saying it shouldn't be a "free for all, set anything you want"19:59
*** Rockyg has joined #openstack-cinder19:59
jgriffithbecause it turns out that doesn't solve anyhting either19:59
scottdaguitarzan: reset-state for cleaning up Cinder DB is one tool, but it doesn't cleanup Nova BlockDeviceMapping table, so it is an incomplete solution19:59
guitarzanscottda: totally agreed19:59
scottdaPlus storage backend needs to terminate_connection19:59
jgriffithscottda: for your case yeah, I'd love to see a force-detach type of thing19:59
jgriffithscottda: +1 to both of your last statements :)19:59
jgriffithscottda: but I still say we should avoid writing shitty code too19:59
scottdaYes. And plenty of work to cleanup old, crappy code.20:00
e0nejgriffith, scottda: +120:00
*** mriedem has quit IRC20:02
openstackgerritVilobh Meshram proposed openstack/cinder: Driver get_stats refresh arg is useless  https://review.openstack.org/17499320:03
scottdaThere's a spring topic for Summit on this: "The API and contracts between nova and cinder"20:04
*** annegentle has joined #openstack-cinder20:05
openstackgerritVilobh Meshram proposed openstack/cinder: Driver get_stats refresh arg is useless  https://review.openstack.org/17499320:08
openstackgerritAdriano Freires Rosso proposed openstack/cinder: Fix HNAS driver parser error  https://review.openstack.org/18030020:09
*** merooney has quit IRC20:11
*** rushiagr_away is now known as rushiagr20:12
*** ctina_ has joined #openstack-cinder20:14
*** breitz has quit IRC20:16
*** breitz has joined #openstack-cinder20:16
*** ctina has quit IRC20:18
*** ctina_ has quit IRC20:19
*** mtanino has joined #openstack-cinder20:20
openstackgerritAdriano Freires Rosso proposed openstack/cinder: HDS HNAS Driver fails when FS unhealthy  https://review.openstack.org/18030020:20
*** bkopilov has joined #openstack-cinder20:21
*** timcl has quit IRC20:22
*** jkremer has joined #openstack-cinder20:24
*** vilobhmm1 has quit IRC20:24
*** bkopilov has quit IRC20:26
*** ganso_ has quit IRC20:27
*** madskier has quit IRC20:32
*** madskier has joined #openstack-cinder20:34
*** juzuluag has joined #openstack-cinder20:35
*** barra204 has quit IRC20:37
*** e0ne has quit IRC20:38
*** thangp has quit IRC20:38
*** aix has quit IRC20:38
*** alecv has quit IRC20:38
nikeshmasselin__ : there?20:39
*** Trozz has quit IRC20:39
*** Trozz has joined #openstack-cinder20:39
asselin__sorry...yes, nodepoold normally creates the image.20:39
asselin__however It's better to do it manually the first time to debug any issues20:40
*** rushiagr is now known as rushiagr_away20:40
*** ebalduf has quit IRC20:40
*** uberjay has quit IRC20:41
nikeshmasselin__ : ok once nodepool image-build dpc is done ,we should run "nodepoold -d $DAEMON_ARGS" right20:42
nikeshmasselin__ : its taking long time for "nodepool image-build dpc"20:42
asselin__nodepool image-upload ? ?20:42
asselin__that's also done automatically, but can do manually the first time20:43
nikeshmafter nodepool image-build dpc is complete,nodepool image-upload20:43
nikeshmthen finally nodepoold -d $DAEMON_ARGS"20:44
asselin__nikeshm, yes.....20:45
*** uberjay has joined #openstack-cinder20:45
*** ebalduf has joined #openstack-cinder20:46
asselin__nikeshm, keep in mind normally you do nodepoold - d right away. but until you know your image creation works and upload works, it's simpler to do it step by step20:46
*** r-daneel has joined #openstack-cinder20:47
*** Lee1092 has quit IRC20:51
*** bill_az has joined #openstack-cinder20:52
*** patrickeast has quit IRC20:52
*** sjcherry has quit IRC20:52
*** patrickeast has joined #openstack-cinder20:53
patrickeastasselin__: hey so i’m getting started on setting up fc support for our CI system, i have some pastes from you with helper scripts http://paste.openstack.org/show/180608/ http://paste.openstack.org/show/180612/ and http://paste.openstack.org/show/180613/20:57
*** IlyaG has quit IRC20:57
patrickeastasselin__: do you install the driver as part of the image prep? and then do the passthrough as a ready script?20:58
openstackgerritSean McGinnis proposed openstack/cinder: Logging not using oslo.i18n guidelines (zonemgr)  https://review.openstack.org/18031920:59
asselin__patrickeast, I use the same image. I do everything as part of the jenkins job21:00
*** vilobhmm1 has joined #openstack-cinder21:00
patrickeastasselin__: oh ok, so just another shell builder step that does the driver and passthrough?21:00
*** vilobhmm1 has quit IRC21:01
*** vilobhmm1 has joined #openstack-cinder21:02
*** juzuluag has quit IRC21:03
*** IlyaG has joined #openstack-cinder21:03
*** annegentle has quit IRC21:11
*** annegentle has joined #openstack-cinder21:12
*** bswartz has quit IRC21:12
*** anshul has joined #openstack-cinder21:13
*** merooney has joined #openstack-cinder21:19
*** madskier has quit IRC21:24
*** Yogi12 has quit IRC21:28
*** boris-42 has quit IRC21:28
*** madskier has joined #openstack-cinder21:30
*** merooney has quit IRC21:31
*** merooney has joined #openstack-cinder21:31
*** ociuhandu has quit IRC21:32
*** dustins has quit IRC21:33
*** merooney has quit IRC21:38
*** merooney has joined #openstack-cinder21:41
*** ebalduf has quit IRC21:46
*** mriedem has joined #openstack-cinder21:51
*** ebalduf has joined #openstack-cinder21:51
*** merooney has quit IRC21:52
*** rushil has quit IRC21:52
*** annegent_ has joined #openstack-cinder21:55
*** esker has quit IRC21:58
*** annegentle has quit IRC22:00
*** changbl has quit IRC22:04
*** vilobhmm1 has quit IRC22:04
*** vilobhmm1 has joined #openstack-cinder22:05
*** vilobhmm1 has quit IRC22:05
*** vilobhmm1 has joined #openstack-cinder22:05
*** bnemec has quit IRC22:05
*** bnemec has joined #openstack-cinder22:08
*** vilobhmm1 has quit IRC22:10
*** vilobhmm1 has joined #openstack-cinder22:14
*** merooney has joined #openstack-cinder22:15
*** hodos has quit IRC22:15
*** sgotliv has quit IRC22:18
*** bkopilov has joined #openstack-cinder22:25
*** Zhongjun_ has joined #openstack-cinder22:26
vilobhmm1hemna : ping22:27
vilobhmm1jgriffith : ping22:27
jgriffithvilobhmm1: hey22:28
vilobhmm1hello need some input jgriffith22:28
vilobhmm1for https://bugs.launchpad.net/cinder/+bug/144675022:28
openstackLaunchpad bug 1446750 in Cinder "cinder-manage service list shows happy for uninitialized driver" [Low,New] - Assigned to Vilobh Meshram (vilobhmm)22:28
*** yrabl_ has joined #openstack-cinder22:29
vilobhmm1do we want to make sure that if the driver is uninitialized we show manager as disabled or :( XXX22:29
*** sweston_ has joined #openstack-cinder22:29
jgriffithvilobhmm1: no22:29
*** Zhongjun has quit IRC22:29
*** rwsu_ has quit IRC22:30
*** yrabl has quit IRC22:30
*** sweston has quit IRC22:30
*** ndipanov has quit IRC22:30
*** uberjay has quit IRC22:30
*** ndipanov has joined #openstack-cinder22:30
*** uberjay has joined #openstack-cinder22:30
*** Rockyg has quit IRC22:30
*** Zhongjun_ is now known as Zhongjun22:30
jgriffithvilobhmm1: absolutely not, because for example multi-backend22:30
*** rwsu_ has joined #openstack-cinder22:30
vilobhmm1hemna..since you filed the bug and i was about to start working on it…wanted to get an idea as in what is expected here....22:30
vilobhmm1as it was not clear from the bug https://bugs.launchpad.net/cinder/+bug/144675022:31
openstackLaunchpad bug 1446750 in Cinder "cinder-manage service list shows happy for uninitialized driver" [Low,New] - Assigned to Vilobh Meshram (vilobhmm)22:31
*** sweston_ is now known as sweston22:31
hemnavilobhmm1, so the service is up, but isn't usable for that host22:31
hemnabecause the driver isn't marked as initialized.22:31
hemnacurrently the service entry is listed as happy22:31
hemnabut it's not22:32
hemnaso...I'm not sure what we want to put there22:32
jgriffithhemna: did you look at "cinder service-list"22:32
hemnaI think it's something between XXX and :-)22:32
vilobhmm1jgriffith, hemna : https://github.com/openstack/cinder/blob/master/cinder/cmd/manage.py#L44722:32
jgriffithhemna: no offense but I find that kinda crazy22:32
vilobhmm1is where cinder-manage will fetch the services details from cinder.services22:33
jgriffithhemna: it's either up or down, I think introducing an inbetween state is just asking for trouble22:33
hemnaso XXX is fine with me22:33
jgriffithvilobhmm1: yeah, they're the same call for the most part22:33
hemnaas it's not usable22:33
vilobhmm1right now https://github.com/openstack/cinder/blob/master/cinder/volume/manager.py#L305 does not throw any exception22:33
jgriffithhemna: +122:33
hemnajust didn't know if we wanted to have some finer grained information22:33
jgriffithvilobhmm1: hemna the key column that needs adjusted is 'state'22:34
vilobhmm1ok so the final conclusion is to change the state = 'XXX' in such a scenario22:34
hemnavilobhmm1, +1 from me22:34
vilobhmm1does that make sense  jgrffith, hemna : ^^22:34
jgriffithhemna: we don't, and I personally think it should be black or white... it either is working or it isn't22:34
vilobhmm1jgriffith : agree22:34
hemnayah cool with me.22:34
hemnaXXX just means, look in the c-vol logs, figure it out :)22:35
hemnaand if it's this issue, then they'll see it22:35
jgriffithhemna: what do you want it to say?22:35
hemnaI'm ok with XXX22:35
jgriffithhemna: Ok, I can't tell what you want/expect22:36
vilobhmm1jgriffith : are you ok with the above conclusion ?22:36
jgriffithvilobhmm1: I believe so22:37
jgriffithvilobhmm1: but honestly I'm not completely clear why it isn't doing that already :)22:37
vilobhmm1let me verify it22:37
vilobhmm1will update the bug22:37
*** primechuck has quit IRC22:40
*** vilobhmm1 has quit IRC22:45
*** vilobhmm1 has joined #openstack-cinder22:47
*** vilobhmm1 has quit IRC22:47
*** vilobhmm1 has joined #openstack-cinder22:48
*** merooney has quit IRC22:51
*** ebalduf has quit IRC22:51
*** EmilienM|afk is now known as EmilienM22:52
*** emagana has quit IRC22:57
*** emagana has joined #openstack-cinder22:58
*** bswartz has joined #openstack-cinder23:02
*** annegent_ has quit IRC23:02
*** emagana has quit IRC23:02
*** emagana has joined #openstack-cinder23:09
hemnais there a trick to seeing LOG.XXX output while running tox unit tests?23:12
hemna-v doesn't seem to do it23:12
*** emagana has quit IRC23:14
thingeescottda: I read the backlog on https://review.openstack.org/#/c/164309/ ... not sure if there was a conclusion23:17
*** ebalduf has joined #openstack-cinder23:22
*** setmason has quit IRC23:23
*** rmesta has quit IRC23:29
*** ebalduf_ has joined #openstack-cinder23:31
*** shakamunyi has quit IRC23:31
jgriffithhemna: the point is kinda to NOT see that output23:31
scottdaWell, I'm not sure either...If we are keeping reset-state as a DB-only solution, there could be scope for fixing migration status and attach status.23:31
jgriffithhemna: in fact there have been a number of cleanup patches to make sure all of that is not shown23:32
mtaninosmcginnis: hi,23:32
jgriffithhemna: I'm sure there's an option to do it via captured stdout/stderr settings23:32
scottdaThe problems come from the naive admin believing that reset-state will fix things at a system-wide level23:32
hemnawell I want to see the output to help debug a failure I'm dealing w/23:32
jgriffithhemna: yeah, makes sense23:32
hemnaI get that tox in general runs we don't want to see any LOG.XXX output23:32
jgriffithhemna: you could just attach pdb and use that23:32
hemnaah yah, I guess I could go that route.23:33
jgriffithand use run_tests23:33
hemnajust thought there was some config magic to turn it on23:33
jgriffithhemna: might be faster at least for now23:33
scottdaBut is seems there is plenty of resistance to getting rid of reset-state altogether. And that review ^^^ doesn't seem to affect other components (i.e. nova or backend)23:33
hemnaseems like something simple enough23:33
*** jkremer has quit IRC23:33
hemnajgriffith, thanks23:33
jgriffithhemna: there is IIRC but I've never really messed with it23:33
jgriffithhemna: sure23:33
scottdathingee: ^^^23:33
mtaninosmcginnis: I read this bug. Does this only affect specific driver? https://bugs.launchpad.net/cinder/+bug/145064923:34
openstackLaunchpad bug 1450649 in Cinder "Retyping to new backend not renaming volume on array" [High,Confirmed]23:34
scottdaIt seems that the conclusion will have to come from the community deciding if reset-state should stay as a DB only API for changing state.23:34
*** madskier has quit IRC23:34
jgriffithmtanino: IIRC that's actually retype with migration23:35
jgriffithmtanino: it's the migration piece that's really the bug IIRC23:35
*** ebalduf_ has quit IRC23:35
thingeescottda, jgriffith: we're repeating what we already discussed at the midcycle meetup. we agreed it was confusing people into doing something that it doesn't do. We decided that we should make the help text obvious of what it's doing23:36
thingeescottda, jgriffith: I'll dig up the wording we came up with and propose it in a review23:36
*** Apoorva has quit IRC23:36
mtaninojgriffith: I tried retype using LVM backend1 to LVM backend2, but this was succeeded23:36
jgriffithmtanino: yeah, it's only a problem if/when something fails I think23:38
jgriffithmtanino: might need to get smcginnis and tbarron as I think they discussed it at length23:38
mtaninojgriffith: ah, there is additional conditions that is not mentioned on the bug.23:38
jgriffithmtanino: oh... I remember more now23:38
mtaninojgriffith: remember?23:39
jgriffithmtanino: so there's never a call to the backend to do the rename portion (I think)23:39
thingeescottda: actually I can't find the wording... but I thought we worked on some https://etherpad.openstack.org/p/cinder-meetup-winter-201523:41
*** Apoorva has joined #openstack-cinder23:41
mtaninojgriffith: you mean rename of destination volume id?23:42
*** Apoorva has quit IRC23:42
jgriffithmtanino: correct (I think)23:42
mtaninojgriffith: OK! let me ask more detail from smcginnis. I need reproduction steps...23:43
jgriffithmtanino: So here's the problem....23:43
jgriffithmtanino: the issue is that *some* backends use the volume-id as a lookup on their backend23:43
jgriffithmtanino: the migration does the id swap when it completes23:43
jgriffithbut doesn't send any notification or request to the backend to update on the backend23:44
jgriffithso the lookup is broken23:44
*** Apoorva has joined #openstack-cinder23:44
jgriffithThat bug needs a lot more explanation :(23:44
jgriffithmtanino: the fix *I think* is actually to implement a migration-complete/update calll to the driver to let it make the same uuid swap23:45
jgriffithmtanino: that it did in the DB23:45
jgriffithmtanino: see what I mean?23:45
mtaninojgriffith: aah, I got it.23:45
jgriffithewwwwwww..... http://paste.openstack.org/show/215148/23:47
jgriffithmtanino: you want me to update the bug description?23:47
*** annegentle has joined #openstack-cinder23:47
*** annegentle has quit IRC23:47
mtaninojgriffith: If you have a chance to do it, it's helpful to understand the detail :)23:48
jgriffithmtanino: sure, I'll update it here in a second23:48
mtaninojgriffith: Thank you for the explanation.23:48
jgriffithmtanino: certainly23:48
mtaninojgriffith: I added this issue into the Vincent's migration improvement items. https://etherpad.openstack.org/p/volume-migration-improvement23:52
jgriffithmtanino: oh, nice!23:56
jgriffithmtanino: I updated the bug as well23:56
mtaninojgriffith: Thank you :) We can discuss it at design summit.23:57
jgriffithlooking forward to it :)23:57

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