Wednesday, 2015-04-29

*** Tross has quit IRC00:01
*** melwitt has quit IRC00:03
*** breitz has quit IRC00:08
*** breitz has joined #openstack-cinder00:08
*** tsekiyama has quit IRC00:18
*** mtanino_ has quit IRC00:21
*** jwcroppe_ has quit IRC00:30
*** jwcroppe has joined #openstack-cinder00:30
*** heyun has joined #openstack-cinder00:31
*** david-lyle has quit IRC00:32
*** jwcroppe has quit IRC00:35
*** theanalyst has quit IRC00:39
*** IanGovett has quit IRC00:40
*** adurbin_ has quit IRC00:43
*** theanalyst has joined #openstack-cinder00:43
thingeeDuncanT: can you reply back to
*** leeantho has quit IRC00:51
*** bswartz has quit IRC00:53
*** bswartz has joined #openstack-cinder00:54
*** xyang1 has quit IRC01:05
*** emagana has joined #openstack-cinder01:06
*** drjones has quit IRC01:07
*** fanyaohong has quit IRC01:09
*** emagana has quit IRC01:10
*** emagana has joined #openstack-cinder01:11
*** david-lyle has joined #openstack-cinder01:20
*** Yogi1 has joined #openstack-cinder01:24
*** Yogi1 has quit IRC01:29
*** Yogi1 has joined #openstack-cinder01:32
*** Lee1092 has joined #openstack-cinder01:34
*** markvoelker has quit IRC01:36
*** zhenguo has joined #openstack-cinder01:38
*** annashen has joined #openstack-cinder01:39
*** Apoorva has quit IRC01:41
*** annashen has quit IRC01:42
*** vilobhmm1 has quit IRC01:51
*** dims_ is now known as dims01:52
*** vilobhmm1 has joined #openstack-cinder01:53
*** vilobhmm1 has quit IRC01:53
*** vilobhmm1 has joined #openstack-cinder01:54
openstackgerritThang Pham proposed openstack/cinder: Switch to oslo_versionedobjects
*** bnemec has quit IRC01:56
*** bnemec has joined #openstack-cinder01:59
*** dannywilson has joined #openstack-cinder02:02
*** harlowja is now known as harlowja_away02:03
*** vilobhmm1 has quit IRC02:04
*** dannywilson has quit IRC02:16
*** dannywilson has joined #openstack-cinder02:17
*** dannywilson has quit IRC02:21
openstackgerritwuyuting proposed openstack/cinder: Admin extends tenant's volume but change admin's quota
*** david-lyle has quit IRC02:30
*** david-lyle has joined #openstack-cinder02:31
*** patrickeast has quit IRC02:32
*** jwcroppe has joined #openstack-cinder02:32
*** david-lyle has quit IRC02:36
*** emagana has quit IRC02:40
*** markvoelker has joined #openstack-cinder02:41
*** jwcroppe has quit IRC02:45
openstackgerritEdwin Wang proposed openstack/cinder: Add iSCSI protocol support for IBM FlashSystem
*** ebalduf has joined #openstack-cinder03:00
*** heyun has quit IRC03:01
openstackgerritwanghao proposed openstack/cinder: Implement function to import/export snapshots
*** Yogi1 has quit IRC03:08
openstackgerritEdwin Wang proposed openstack/cinder: Add iSCSI protocol support for IBM FlashSystem
*** coolsvap|afk is now known as coolsvap03:27
*** david-lyle has joined #openstack-cinder03:32
*** changbl has joined #openstack-cinder03:33
*** ebalduf has quit IRC03:41
*** jwcroppe has joined #openstack-cinder03:42
*** heyun has joined #openstack-cinder03:43
*** bswartz has quit IRC03:43
*** bswartz has joined #openstack-cinder03:45
openstackgerritEdwin Wang proposed openstack/cinder: Add iSCSI protocol support for IBM FlashSystem
*** ho has joined #openstack-cinder04:00
*** vilobhmm1 has joined #openstack-cinder04:08
*** xyang has quit IRC04:17
*** buster_604 has joined #openstack-cinder04:18
*** rushiagr_away is now known as rushiagr04:21
*** buster_604 has quit IRC04:32
*** buster_604 has joined #openstack-cinder04:32
*** xyang has joined #openstack-cinder04:37
*** buster_604 has quit IRC04:44
*** rushiagr is now known as rushiagr_away04:45
*** buster_604 has joined #openstack-cinder04:48
*** nkrinner has joined #openstack-cinder04:57
openstackgerritEdwin Wang proposed openstack/cinder: Add iSCSI protocol support for IBM FlashSystem
*** dannywilson has joined #openstack-cinder05:22
openstackgerritwuyuting proposed openstack/cinder: Dell EQLX: CHAP config option has no value
*** nlevinki has joined #openstack-cinder05:25
*** dannywilson has quit IRC05:26
*** vilobhmm1 has quit IRC05:28
*** wanghao has quit IRC05:35
*** annashen has joined #openstack-cinder05:45
*** ebalduf has joined #openstack-cinder05:53
*** tshefi has joined #openstack-cinder05:56
*** annashen has quit IRC05:57
*** ebalduf has quit IRC05:58
*** sgotliv has quit IRC06:01
*** dulek has joined #openstack-cinder06:02
*** Longgeek has joined #openstack-cinder06:03
*** vilobhmm1 has joined #openstack-cinder06:03
*** Longgeek has quit IRC06:08
*** pcaruana has joined #openstack-cinder06:09
*** jdurgin1 has quit IRC06:10
*** ho has quit IRC06:11
*** TobiasE has joined #openstack-cinder06:21
*** nikeshm has joined #openstack-cinder06:39
openstackgerritwanghao proposed openstack/cinder: Implement function to import/export snapshots
*** dulek has quit IRC06:46
*** vilobhmm1 has quit IRC06:46
*** vilobhmm1 has joined #openstack-cinder06:47
*** vilobhmm1 has quit IRC06:48
*** dulek has joined #openstack-cinder06:52
*** BharatK_ has joined #openstack-cinder06:55
*** BharatK has quit IRC06:55
*** Longgeek has joined #openstack-cinder06:56
*** Longgeek has quit IRC07:02
*** hflai has quit IRC07:03
*** Longgeek has joined #openstack-cinder07:06
*** buster_604 has quit IRC07:12
*** jordanP has joined #openstack-cinder07:19
*** buster_604 has joined #openstack-cinder07:23
*** buster_604 has quit IRC07:23
*** buster_604 has joined #openstack-cinder07:24
*** aarefiev has quit IRC07:24
*** hflai has joined #openstack-cinder07:26
*** sgotliv has joined #openstack-cinder07:34
*** ronis has joined #openstack-cinder07:36
*** hflai has quit IRC07:38
*** kaisers has quit IRC07:39
*** aarefiev has joined #openstack-cinder07:40
*** kaisers has joined #openstack-cinder07:41
*** jwcroppe has quit IRC07:44
*** jwcroppe has joined #openstack-cinder07:45
*** hflai has joined #openstack-cinder07:45
*** kaisers has quit IRC07:45
*** heyun has quit IRC07:47
*** kaisers has joined #openstack-cinder07:48
*** jwcroppe has quit IRC07:49
*** anshul has joined #openstack-cinder07:52
*** sgotliv has quit IRC07:57
*** jistr has joined #openstack-cinder07:58
*** Longgeek has quit IRC08:05
*** dims has quit IRC08:10
*** ndipanov has quit IRC08:12
*** ndipanov has joined #openstack-cinder08:12
*** buster_604 has quit IRC08:17
openstackgerritJames.Tsai proposed openstack/cinder: Add volume drivers for Infortrend Storage
*** pschaef has joined #openstack-cinder08:30
*** xyang has quit IRC08:43
*** jwcroppe has joined #openstack-cinder08:46
*** sgotliv has joined #openstack-cinder08:53
*** boichev has joined #openstack-cinder08:57
*** annashen has joined #openstack-cinder08:57
*** deepakcs has joined #openstack-cinder08:59
openstackgerritPranali Deore proposed openstack/cinder: cinder os-attach api returns 500
*** annashen has quit IRC09:02
*** dulek_ has joined #openstack-cinder09:03
*** dulek has quit IRC09:03
*** e0ne has joined #openstack-cinder09:06
*** Longgeek has joined #openstack-cinder09:06
*** e0ne is now known as e0ne_09:06
*** alecv has joined #openstack-cinder09:07
*** Longgeek has quit IRC09:12
*** Longgeek has joined #openstack-cinder09:13
*** jwcroppe has quit IRC09:14
*** e0ne_ is now known as e0ne09:14
openstackgerritAnton Arefiev proposed openstack/cinder-specs: Service healthcheck
openstackgerritPh. Marek proposed openstack/cinder: Re-add DRBD driver.
*** afazekas has quit IRC09:32
*** afazekas has joined #openstack-cinder09:32
flip214thingee: ^^ failed gerrit with "dsvm-tempest-xio-fc : NOT_REGISTERED" ??09:39
*** aswadr has joined #openstack-cinder09:44
openstackgerritPranali Deore proposed openstack/cinder: cinder os-detach api returns 500
*** dims has joined #openstack-cinder09:51
openstackgerrityogeshprasad proposed openstack/cinder: Add chap support to CloudByte cinder driver
*** sgotliv has quit IRC09:55
*** reed has quit IRC09:56
*** IanGovett has joined #openstack-cinder10:00
*** e0ne is now known as e0ne_10:03
*** e0ne_ is now known as e0ne10:05
openstackgerritAnton Arefiev proposed openstack/python-cinderclient: Add search_opts into the method list() for qos specs
*** dulek_ has quit IRC10:06
*** mtreinish has quit IRC10:07
flip214DuncanT: THANK YOU, at least a few people around here try to help!10:08
DuncanTflip214: You're welcome10:08
openstackgerritPh. Marek proposed openstack/cinder: Re-add DRBD driver.
flip214see you in Vancouver, I hope!10:09
*** sgotliv has joined #openstack-cinder10:13
*** ociuhandu has joined #openstack-cinder10:15
*** arif-ali has quit IRC10:20
*** reed has joined #openstack-cinder10:28
*** pschaef has quit IRC10:31
*** ociuhandu has quit IRC10:31
*** ociuhandu has joined #openstack-cinder10:34
*** dulek has joined #openstack-cinder10:47
*** jwcroppe has joined #openstack-cinder11:00
*** dulek_ has joined #openstack-cinder11:02
*** shyama has joined #openstack-cinder11:02
*** dulek has quit IRC11:05
*** jwcroppe has quit IRC11:08
*** ociuhandu has quit IRC11:13
*** dims has quit IRC11:18
*** kashyap has joined #openstack-cinder11:29
*** e0ne is now known as e0ne_11:30
*** lpabon has joined #openstack-cinder11:32
*** kashyap has left #openstack-cinder11:33
*** dulek_ has quit IRC11:33
*** coolsvap is now known as coolsvap|afk11:33
*** aix has joined #openstack-cinder11:35
*** dulek has joined #openstack-cinder11:39
*** arif-ali has joined #openstack-cinder11:40
*** cdelatte has joined #openstack-cinder11:42
*** delattec has joined #openstack-cinder11:42
*** delatte has joined #openstack-cinder11:44
*** marcusvrn1 has joined #openstack-cinder11:45
*** cdelatte has quit IRC11:47
*** delattec has quit IRC11:47
*** cdelatte has joined #openstack-cinder11:47
*** marcusvrn has quit IRC11:47
*** delatte has quit IRC11:52
*** cdelatte has quit IRC11:52
*** haomaiwa_ has quit IRC11:52
*** dims has joined #openstack-cinder11:56
*** Adriano_ has joined #openstack-cinder12:00
*** dulek has quit IRC12:02
*** openstackgerrit has quit IRC12:07
*** ganso has joined #openstack-cinder12:07
*** openstackgerrit has joined #openstack-cinder12:07
*** ganso has quit IRC12:08
*** ganso has joined #openstack-cinder12:08
mkodererDuncanT: hi, I proposed a spec for the ABC driver porting
mkodererif you have a sec for a review12:08
*** dencaval has joined #openstack-cinder12:11
*** timcl has joined #openstack-cinder12:13
*** akerr has joined #openstack-cinder12:18
*** bswartz has quit IRC12:20
*** e0ne_ is now known as e0ne12:26
flip214It's so nice to see that Openstack enforces good coding practices!! "H238  old style class declaration, use new style (inherit from `object`)12:32
flip214 /sarcasm off12:32
flip214the line is "from oslo_config import cfg12:33
*** jwcroppe has joined #openstack-cinder12:35
*** EmilienM is now known as EmilienM|afk12:41
*** lifeless has quit IRC12:43
*** chlong has joined #openstack-cinder12:44
*** rodrigods has joined #openstack-cinder12:44
openstackgerritPh. Marek proposed openstack/cinder: Re-add DRBD driver.
*** lpetrut has joined #openstack-cinder12:52
*** e0ne is now known as e0ne_12:54
*** jungleboyj has joined #openstack-cinder12:55
*** timcl has quit IRC12:55
*** timcl1 has joined #openstack-cinder12:55
*** zhipeng has joined #openstack-cinder12:59
*** sandywalsh has joined #openstack-cinder13:00
*** ir2ivps7 has quit IRC13:01
*** dulek has joined #openstack-cinder13:03
flip214DuncanT: john griffith asked for Cinder-Core feedback on, so if you have a minute please answer him. Thanks a lot!13:03
*** annegentle has joined #openstack-cinder13:03
*** julim has joined #openstack-cinder13:05
*** Yogi1 has joined #openstack-cinder13:05
*** e0ne_ is now known as e0ne13:06
*** julim has quit IRC13:06
*** mwichmann has joined #openstack-cinder13:06
*** julim has joined #openstack-cinder13:08
*** mwichmann has left #openstack-cinder13:11
*** ociuhandu has joined #openstack-cinder13:12
openstackgerritPranali Deore proposed openstack/cinder: cinder os-detach api returns 500
*** deepakcs has quit IRC13:14
*** julim has quit IRC13:18
*** dulek has quit IRC13:21
*** dulek has joined #openstack-cinder13:21
*** ir2ivps7 has joined #openstack-cinder13:31
*** jungleboyj has quit IRC13:34
*** EmilienM|afk is now known as EmilienM13:37
*** dustins has joined #openstack-cinder13:37
*** e0ne is now known as e0ne_13:38
*** lifeless has joined #openstack-cinder13:38
*** e0ne_ is now known as e0ne13:39
*** esker has quit IRC13:41
*** ebalduf has joined #openstack-cinder13:42
*** thangp has joined #openstack-cinder13:45
*** mriedem_away has quit IRC13:51
*** ganso_ has joined #openstack-cinder13:54
*** ganso has quit IRC13:54
*** marcusvrn1 has quit IRC13:54
*** mtreinish has joined #openstack-cinder13:54
*** marcusvrn has joined #openstack-cinder13:55
*** asselin has joined #openstack-cinder13:56
*** ebalduf has quit IRC13:59
*** Yogi11 has joined #openstack-cinder13:59
*** bswartz has joined #openstack-cinder14:00
*** dustins_ has joined #openstack-cinder14:00
*** rooneym has quit IRC14:02
*** Yogi1 has quit IRC14:02
*** dalgaaf has joined #openstack-cinder14:02
*** mriedem has joined #openstack-cinder14:03
*** xyang has joined #openstack-cinder14:03
*** dustins has quit IRC14:03
-openstackstatus- NOTICE: gerrit has been restarted to clear a stuck events queue. any change events between 13:29-14:05 utc should be rechecked or have their approval votes reapplied to trigger jobs14:05
*** rushiagr_away has quit IRC14:05
*** alexpilotti has joined #openstack-cinder14:05
*** mtreinish has quit IRC14:05
e0nei've got my Canada visa:). summit is closer now14:08
*** cbits has joined #openstack-cinder14:08
*** cbits has left #openstack-cinder14:09
*** haomaiwang has joined #openstack-cinder14:10
*** Lee1092 has quit IRC14:11
*** annegentle has quit IRC14:12
*** lpetrut has quit IRC14:15
*** rushiagr_away has joined #openstack-cinder14:16
duleke0ne: By luck I've noticed that I've needed to get a new passport to be able travel to Canada. Last preparation are in progress. ;)14:16
*** annegentle has joined #openstack-cinder14:17
*** zhenguo has quit IRC14:17
e0nedulek: i hope to see you during design sessions14:18
*** Yogi11 has quit IRC14:19
duleke0ne: I'll definitely participate. :)14:21
*** tshefi has quit IRC14:21
*** nkrinner has quit IRC14:24
*** TobiasE has quit IRC14:25
*** asrangne has joined #openstack-cinder14:26
*** esker has joined #openstack-cinder14:27
*** jungleboyj has joined #openstack-cinder14:28
*** aswadr has quit IRC14:28
*** julim has joined #openstack-cinder14:31
*** Yogi1 has joined #openstack-cinder14:37
*** lpetrut has joined #openstack-cinder14:38
*** zhipeng has quit IRC14:39
*** dims has quit IRC14:42
*** bkopilov has quit IRC14:42
*** dims has joined #openstack-cinder14:42
*** mtreinish has joined #openstack-cinder14:45
*** bkopilov has joined #openstack-cinder14:45
*** mriedem has quit IRC14:46
*** jwcroppe has quit IRC14:47
*** jwcroppe has joined #openstack-cinder14:47
*** rmesta has joined #openstack-cinder14:49
*** alexpilotti has quit IRC14:49
*** marcusvrn_ has joined #openstack-cinder14:49
*** mtreinish has quit IRC14:49
*** nlevinki has quit IRC14:50
*** BharatK_ has quit IRC14:52
*** mriedem has joined #openstack-cinder14:52
*** jwcroppe has quit IRC14:52
*** emagana has joined #openstack-cinder14:55
*** charlesr has joined #openstack-cinder14:56
*** ronis has quit IRC14:57
*** patrickeast has joined #openstack-cinder15:00
*** timcl has joined #openstack-cinder15:02
*** dulek has quit IRC15:02
*** BharatK_ has joined #openstack-cinder15:04
*** ganso__ has joined #openstack-cinder15:04
*** timcl1 has quit IRC15:06
*** Tross has joined #openstack-cinder15:11
*** alexpilotti has joined #openstack-cinder15:11
*** jdandrea has joined #openstack-cinder15:13
thingeeflip214: can you give me the link to the spec john garbutt is talking about? He linked to the same spec he left a comment on15:15
thingeee0ne: nice :)15:15
jdandreaLooking for secure_delete in cinder.conf documentation (Juno). Not seeing it. Does it go in [DEFAULT]? Maybe I can update the docs, unless that should be auto-generated (?).15:16
*** Adriano_ has quit IRC15:16
annegentlejdandrea: yeah that's strange since they are autogenerated15:17
annegentlethe config docs15:17
e0nejdandrea: it should be in driver's section for Kilo. let e check for J15:17
annegentlee0ne: I'm searching in there too and not finding it15:18
jdandreaIf it's something I can help fix it will give me more practice. ;)15:18
* jdandrea just noticed "The New Kids On the Block" ONLY NOW and cracked himself up.15:19
jdandreaOhhh, not secure_delete.15:22
jdandreaThat would explain it. :)15:22
*** eharney has quit IRC15:23
*** jwcroppe has joined #openstack-cinder15:23
*** Yogi1 has quit IRC15:24
annegentlejdandrea: it's odd, secure_delete is still in the devstack docs15:24
e0nejdandrea: it's a common mistake:(15:25
jdandreaannegentle: Yep. That's where I noticed. No harm no foul. :)15:25
annegentlejdandrea: well, fix it there if you can :)15:25
jdandreaannegentle: Certainly!15:25
e0nejdandrea: and fyi: - could be related to your case issues15:25
openstackLaunchpad bug 1255593 in Cinder "configuration options from the [DEFAULT] section are not applied to multiple backends" [Wishlist,Triaged]15:25
jdandreaYou're very welcome. The more practice I get the better.15:25
jdandreaEven with doc changes, it all counts.15:26
*** mtanino has joined #openstack-cinder15:38
*** bswartz has quit IRC15:39
*** dulek has joined #openstack-cinder15:41
*** julim has quit IRC15:43
*** mtreinish has joined #openstack-cinder15:44
*** rushiagr_away is now known as rushiagr15:46
jdandreae0ne: Thanks for the bug report. I haven't tried it just yet but we'll see. This particular cluster doesn't have multiple backends but this is great to know, thx.15:46
*** jwcroppe has quit IRC15:47
e0nejdandrea: it's not only about multiple backends15:49
jdandreae0ne: Ohhh.15:50
e0nejdandrea: it's issue is about options inheritence from DEFAULT to drivers' sections15:51
jdandreae0ne: ... and not just devstack, and we WILL be using multiple backends at some point.15:51
jdandreaGot it.15:51
jdandreaThis will be relevant then. <!>15:51
*** anshul has quit IRC15:51
*** _cjones_ has joined #openstack-cinder15:52
*** annegentle has quit IRC15:53
*** asrangne has quit IRC15:53
*** dannywilson has joined #openstack-cinder15:54
*** dannywilson has quit IRC15:54
*** bswartz has joined #openstack-cinder15:54
jdandreae0ne: Reality check Q. volume_clear goes on *every* node that runs cinder-volume, yes?15:54
*** dannywilson has joined #openstack-cinder15:55
e0nejdandrea: yes. it's a cinder-volume option15:55
jdandreaGood, tx. Just being extra sure. ;)15:55
*** scohen has joined #openstack-cinder15:55
jdandreaIt was only set on the controller node. Oops.15:55
openstackgerritAndrew Durbin proposed openstack/cinder-specs: Cinder Volume Performance and Health collection for cinder volume drivers.
*** adurbin_ has joined #openstack-cinder15:57
rhe00_erlon: ping15:58
flip214DuncanT: ?15:58
*** dulek has quit IRC15:59
DuncanTflip214: Probably worth having a cinder one too I think, it makes tracking the work easier - Mike does a lot of our prioritisation via blueprints. I think you can just copy & paste the text16:00
*** Yogi1 has joined #openstack-cinder16:01
mriedemthat's a lot of errors16:03
*** dulek_home has joined #openstack-cinder16:03
*** li_zhang has joined #openstack-cinder16:04
*** vilobhmm1 has joined #openstack-cinder16:06
*** dulek_home is now known as dulek16:06
flip214thingee: I'm not sure why he repeated the spec link.16:06
*** hemnafk is now known as hemna16:06
*** li_zhang has quit IRC16:08
*** mriedem is now known as mriedem_lunch16:09
*** eharney has joined #openstack-cinder16:09
*** akshai has joined #openstack-cinder16:09
*** jistr has quit IRC16:11
*** li_zhang has joined #openstack-cinder16:12
*** li_zhang has quit IRC16:16
*** julim has joined #openstack-cinder16:17
*** li_zhang has joined #openstack-cinder16:19
*** chlong has quit IRC16:19
*** rhe00_ is now known as rhedlind16:20
*** Apoorva has joined #openstack-cinder16:20
*** ociuhandu has quit IRC16:20
*** alexpilotti has quit IRC16:20
*** Apoorva has quit IRC16:21
*** pcaruana has quit IRC16:22
*** alecv has quit IRC16:25
openstackgerritTom Swanson proposed openstack/cinder: Dell SC driver honoring folder name after volume creation.
*** Apoorva has joined #openstack-cinder16:32
*** dulek_home_ has joined #openstack-cinder16:34
*** dulek has quit IRC16:35
*** dulek_home_ is now known as dulek16:35
*** leeantho has joined #openstack-cinder16:38
*** mriedem_lunch is now known as mriedem16:44
*** asselin has quit IRC16:47
*** ronis has joined #openstack-cinder16:52
*** ganso_ has quit IRC16:54
*** ganso__ is now known as ganso_16:54
*** annegentle has joined #openstack-cinder16:54
*** marcusvrn has quit IRC16:55
*** marcusvrn_ is now known as marcusvrn16:55
*** ndipanov has quit IRC16:56
*** ndipanov has joined #openstack-cinder16:56
*** marcusvrn1 has joined #openstack-cinder16:56
scottdaxyang, hemna: bad assumptions Nova makes about connector are in compute/ line 178316:58
scottdaI haven't filed a bug yet, but I'll be doing that...16:59
*** akshai has quit IRC16:59
vilobhmm1thingee : back from #openstack-meeting17:00
*** patrickeast has quit IRC17:00
dulekSo as use cases are involved - users would probably love to participate17:00
dulekThat would mean fishbowl17:00
jungleboyjthingee: I need to run to an appointment.  Back in about 45 minutes and will catch up then.17:01
xyangscottda: will catch up later.  thanks17:01
vilobhmm1i believe its an important feature…i would go through implementation details as well17:01
*** annegentle has quit IRC17:01
vilobhmm1and also race conditions involved…if at all any can be discussed …17:01
vilobhmm1thingee, DuncanT :
DuncanTA certain amount of this is definitely x-project17:01
hemnascottda, I haven't looked at that one yet, but that smells like it might be a fix in the area live migration needs.17:01
thingeefine with fish bowl to get feedback ...17:02
*** bswartz has quit IRC17:02
DuncanTThe behaviour of nested quotas should match in cinder/nova/swift17:02
thingeeimplementation/race conditions won't be as interesting to users though17:02
*** buster_604 has joined #openstack-cinder17:02
vilobhmm1thats why proposing for design session17:02
winston-dDuncanT: +117:02
vilobhmm1to get more inputs in quota areas17:02
vilobhmm1as we get bitten by quota related problems every now n then17:03
vilobhmm1and our exsiting quota driver is kinda broken17:03
thingeealso deprecated one for sometime still exists17:03
thingeekind of difficult to remove though17:03
DuncanTvilobhmm1: Until there's cross-project contentious on the high level behaviour, we should implement anything17:03
*** annegentle has joined #openstack-cinder17:03
vilobhmm1so i feel should be a good session to talk and get inputs and make a better quota driver17:03
scottdahemna: thx17:03
vilobhmm1i think nova has already procedded with changes as CERN people mentioned to me17:04
DuncanTvilobhmm1: I've some thoughts on making reservations richer to make debugging easier17:04
vilobhmm1so there is some agreement17:04
DuncanTvilobhmm1: Can we get links to that in the etherpad, please?17:04
vilobhmm1sure DuncanT : i think like such valueable feedback can be received in design session that fish bowl17:04
vilobhmm1sure…let me find them17:04
DuncanTvilobhmm1: If there's agreement to the high level behaviour already, great17:05
vilobhmm1yes there is one17:05
dulekJust the current status - we have one place for fish bowl and we have 2 topics too much for working sessions17:05
winston-dvilobhmm1: improving existing quota driver (hierarchy 1) and implmenting nested quota driver can be two separate tasks.17:05
*** bswartz has joined #openstack-cinder17:06
vilobhmm1winston-d : agree but we have to start somewhere17:06
*** ganso__ has joined #openstack-cinder17:06
vilobhmm1and i wish this can be that17:06
thingeemarked as potential fishbowl17:06
thingeeeverything does through scheduler17:06
thingeewinston-d: hi17:07
DuncanTwinston-d: I put a question on this on the etherpad.... what is the win?17:07
thingeefirst off agree this should not be in the fish bowl17:07
winston-dthis is basically revisit of the same topic that i discussed in Paris, with more use cases17:07
thingeewinston-d: :)17:07
winston-dDuncanT: reserved_percentage has no affect of vol clones? being one17:08
DuncanTwinston-d: we can't clone across backends though, right?17:08
DuncanTwinston-d: So you can just put the clone in error in the manager, no scheduler change required17:09
winston-dDuncanT: right, the point of having it go through scheduler is not about letting scheduler to make alternative decision17:09
DuncanTWhat is the point then?17:10
winston-d it is about scheduler, who has the knwoledge of backend (is full, can't create more vols, etc) is able to stop those requests.17:10
DuncanTThe backend has all that info though, and is more up-to-date than the scheduler17:10
*** aix has quit IRC17:10
*** yuriy_n17_ has joined #openstack-cinder17:11
DuncanTWhere the request stops, unless it is in the API, really makes no difference17:11
winston-dDuncanT: correct, but at least for reserved_percentage, most drivers don't even do such check17:11
DuncanTDrivers shouldn't need to, manager should do it17:11
thingeeDuncanT: +117:11
*** buster_604 has quit IRC17:11
winston-dDuncanT: why? for now, reserved_percentage is simplied ignored17:11
winston-d for all drivers17:12
DuncanTwinston-d: ignoring it might be a bug, but then I thought it was there exactly to give room for extend, clone, snap, etc?17:12
DuncanTIf you reserve it from ever being used for anything, it is a bit pointless...17:12
scottdawinston-d, DuncanT: what about I/O requests going through the scheduler so that volumes are not tied to a single cinder-volume node?17:13
DuncanTscottda: rabbit can and does already do round-robin type stuff for that17:13
*** sgotliv has quit IRC17:13
winston-dDuncanT: not really, we have use cases where we would like to reserve (say 15%) capacity to makre sure backend is running OK17:13
DuncanTwinston-d: so get the driver to lie about available capacity17:14
DuncanTSame effect17:14
winston-ddeployer should implement their own in-house driver, you are saying?17:14
*** akshai has joined #openstack-cinder17:15
*** dustins_ has quit IRC17:15
DuncanTNo, the driver author should know if an array can't actually handle the capacity it says it is17:15
winston-dbtw, by vol clone, I mean create vol from source vol17:15
tbarronwinston-d: DuncanT: was just looking at reserved capacity as the complement of nfs_used_ratio17:16
DuncanTYes, that clone has to go on the same backend, right?17:16
DuncanTtbarron: I'm not familiar with that...17:16
*** li_zhang has quit IRC17:16
tbarronassumption in code seems to be that instead of lying about capacity or  free, report reserved and scheduler will adjust17:16
DuncanTRight, well the c-vol manager code can adjust too17:16
winston-dDuncanT: right, but the clone (new vol)'s capacity isn't from reserved, it is part of used_capcacity17:17
tbarronDuncanT: sure, should be in the manager on the way up17:17
vilobhmm1+1..also quota and scheduling decisions are tied17:17
vilobhmm1DuncanT : +117:17
thingeewinston-d: ok marking for potential in working sessions17:17
winston-dDuncanT: for snapshot, some drivers may not include that into used_cap, some may, not sure.17:17
DuncanTwinston-d: So make c-vol manager reject stuff that won't fit inside reserved.... we need to do that anyway since the scheduler's view might be out of date17:17
thingeeasync error reporting to the end user17:18
thingeeameade: hi17:18
thingeeso there is an agreement that we should probably start with a cross project17:18
DuncanTthingee: ++17:18
vilobhmm1thingee : +117:19
ameademy concern with that is it will end up being a lot more broad discussion, i rather go bottom up and then propose what we decide on the the ML17:19
ameadethat's my gut feeling anyway17:19
*** MentalRay has joined #openstack-cinder17:19
*** marcusvrn has quit IRC17:19
thingeeameade: have you reached out to this list to see if anyone else is looking at this from other projects?17:19
*** melwitt has joined #openstack-cinder17:19
*** MentalRay has quit IRC17:19
thingeeI fear like rolling upgrades there being a dozen individual sessions for each project17:20
ameadeyeah fair, I know other projects are interested but not enough to drive it to completion17:20
dulekthingee: Most projects haven't adopted versionedobjects yet.17:20
ameadeWhat i can do is the POC code in cinder but have it as a cross-project session17:20
dulekthingee: So their sessions would probably be based on that.17:20
dulekthingee: Cinder's one can go further.17:21
*** IanGovett has quit IRC17:21
DuncanTameade: PoC is fine as long as we don't merge it without some agreement17:21
ameadeDuncanT: yeah absolutely17:21
*** MentalRay has joined #openstack-cinder17:22
ameadeI definitely want a solution that will work all across openstack17:22
ameadei'll be focused on this for the next few weeks17:22
ameadethingee: should I fill out the cross-project session form?17:23
thingeeameade: please17:23
*** rushil has joined #openstack-cinder17:23
ameadesounds good17:24
thingeedoesn't look like vincent hou is here to discuss volume migration17:24
*** patrickeast has joined #openstack-cinder17:25
thingeejungleboyj: hi17:25
thingeeI don't think there is much to discuss since we already have this laid out
thingeemight just make sense to get an update in the sprints from vincent and jungleboyj17:26
thingeeOK, well I updated that item in the etherpad17:27
thingeetbarron: hi17:27
thingeescaling backup services17:27
thingeeDuncanT: ^17:27
*** jordanP has quit IRC17:27
tbarronI don't have a particular design proposal17:27
tbarronwanted to flush out known problem areas17:27
tbarronfind who is interested in the area17:27
tbarronSo this can be moved as appropriate.17:28
DuncanTWe tried multi-threading, it is really damn hard to get right17:28
tbarronDuncanT: right17:28
DuncanTWith ssl and compression, the process is verymuch CPU bound17:28
tbarronAlso e0ne knows of some potential threading issues17:28
DuncanTIn theory it is very trivial to parallelise17:28
tbarronwith NFS we potentially increase chunk size a lot17:28
DuncanTIn python, not so much17:28
*** buster_604 has joined #openstack-cinder17:28
tbarronand there is the whole issue of having to run on a single node today17:28
tbarronthere may be other areas17:29
vilobhmm1tbarron : anything locking, threading, concurrency related stuff please count me in17:29
tbarronso in some forum or another it would be good to find out about best practices and known problem areas17:29
DuncanTFixing up (removing) the coupling between v-col and c-backup should definitely happen17:29
e0neDuncanT: +117:29
tbarronAnd the summit seemed a good place to flush this stuff out.17:29
DuncanTvilobhmm1: I'd love to hear your thoughts17:29
tbarronthat's all I have :-)17:30
*** yuriy_n17_ has quit IRC17:30
vilobhmm1DuncanT : it would be nice that tbarron adds his findings here
vilobhmm1i am sure the underlying problem is the same17:30
vilobhmm1that we plan to discuss for the rest17:31
vilobhmm1tbarron : ^^17:31
tbarronvilobhmm1: I can, but DuncanT and e0ne know stuff I don't know enough to add17:31
*** buster_6_ has joined #openstack-cinder17:31
DuncanTvilobhmm1: backup in parallel is a totally different problem17:31
DuncanTvilobhmm1: There is very little overlap17:31
thingeetbarron: I would like there to be some plan on discussion and something to work off of if we keep this in working sessions17:31
DuncanTvilobhmm1: Probably better to start a new etherpad for backup issues17:31
tbarronthingee: agreed.  I can make an etherpad with a strawman agenda if that would be helpful17:32
vilobhmm1DuncanT : ok…sure..let me go through code17:32
*** buster_604 has quit IRC17:32
DuncanTthingee: I can work with tbarron to get some sort of plan17:32
tbarronthingee: or if you think this topic should be in a difft forum ...17:32
tbarronDuncanT: +117:32
DuncanTtbarron: I think it is well worth talking out in person17:32
thingeevolume replication v217:32
DuncanTvilobhmm1: You'll be at the summit, right?17:33
thingeejungleboyj, jgriffith hi17:33
vilobhmm1DuncanT : yes17:33
thingeeI think everyone agrees this should be part of the working sessions17:33
vilobhmm1I will be..17:33
DuncanTvilobhmm1: Great17:33
vilobhmm1thingee : +1 looks good to me17:33
thingeethere's already a spec to go with it as well17:33
dannywilsonI wonder if replication should be fishbowl?  Do we know how users want replication done?17:33
*** emagana has quit IRC17:33
*** buster_6_ has quit IRC17:34
dannywilsondo users want replication done with CG or not?17:34
*** charlesr has quit IRC17:34
thingeedannywilson: I don't think that's out of the question for CG. But I don't think we're quite there with what we already have. Not sure if you read the spec, but this aims to simplify how replication works today for drivers so others can actually implement the feature17:35
thingeewe're not quite there*17:36
dannywilsonI have but if we are doing the work anyway we should make sure users are actually going to use it?17:36
* thingee double checks if this has api impact17:36
dannywilsonor just get a simple one done and then go to users from there is what you are saying?17:36
thingeedannywilson: if it has api implact sure, user endpoint would be good. if not, end users don't need to know how this feature works under the hood.17:37
thingeewe've already discussed that in Juno summit17:37
dannywilsongotcha, okay17:37
thingeeuser's input*17:37
thingeecan't type today =/17:37
dannywilsonthanks, I am good with working session17:38
thingeedannywilson: ah it does have api impact of introducing new endpoints17:38
*** xyang has quit IRC17:38
DuncanTOne big thing to come out of the last attempt at replication is that backends have weird limitatations on what they can implement17:39
thingeereplication-enable, replication, disable, replication-failover17:39
DuncanTSo 'what users want' is somewhat of a moot point - start with 'what is technically possible'17:39
thingeeDuncanT: true17:39
*** xyang has joined #openstack-cinder17:39
DuncanTNetapp seems to have a distinct set of limitations that others didn't17:40
dannywilsonDuncanT: that makes sense17:40
thingeeDuncanT: I agree. I think users maybe frustrated with attending if all we did was argue about our backend limitations :)17:40
thingeeok seems fine to keep then17:41
thingeerolling upgrades17:41
thingeethangp: hi17:41
thangpso i wanted to discuss how to proceed with rolling upgrades17:41
openstackgerritYuriy Nesenenko proposed openstack/cinder: Fix unit tests spam output
thingeeso this topic, to be clear should be looking back on you did with schema upgrades and how we should proceed with rpc compatibility as the last step17:42
*** buster_604 has joined #openstack-cinder17:42
*** buster_604 has quit IRC17:42
thingeein the look back, we should have a list of what's left to be turned into objects17:42
thingeeI think a lot of these are already in review now17:42
thangpi got volumes converted and tempest pass for it...should have it done before summit17:43
thangpjust got to update the unit tests17:43
DuncanTWe have zero examples of how to do compatibility work-arounds in the object layer currently (i.e. new/removed fields)17:43
dulekthangp: Wasn't DuncanT interested in working on volume's object?17:43
DuncanTdulek: No, just in reviewing it17:43
thangpdulek: not sure, but i probably left holes to be filled :)17:44
DuncanTdulek: If you wait for me to code, you'll be waiting a long while at the moment  :-(17:44
thingeethangp: can you start an etherpad of what's left?17:44
thangpso for rolling upgrades, i like to see what we do after the conversion is done17:44
thingeeand for the rpc compatibility work17:44
thangpthingee: there is an etherpad, let me check...17:44
thingeethangp: oh whoops. Can you add it to the session item here?
thangpthingee: ok, will do...once i find it17:45
thangpanyways, objects is only one part of rolling upgrades...there is also the rpc compat layer17:46
thangpthat, i was hoping you(thingee) could help explain17:46
thingeethangp: so nova is already doing this with their conductor.17:47
*** ganso_ has quit IRC17:47
* DuncanT tried to understand the nova RPC compat code and failed entirely17:47
thangpbut we dont have a conductor :(17:47
thangpare we planning one?17:47
vilobhmm1thangp : true17:48
thingeeI think we talked about sending over serialized objects?17:48
thangpthingee: yes, and it does today for snapshots17:48
thangpi have to fill in more details into that etherpad17:48
thangpof what is left to do17:48
thingeeDuncanT had some thoughts on this :)17:49
thingeeDuncanT: or not?17:49
thangpis that rpc version capping?17:49
*** buster_604 has joined #openstack-cinder17:50
DuncanTVersion capping is different to object version translation... both are needed17:50
*** buster_604 has quit IRC17:50
vilobhmm1thingee, thangp, DuncanT : if we are done with the objects discussion17:51
vilobhmm1thingee, DuncanT : nested quota spec is already accepted in nova… so I guess the nested quota proposal  has a cross-project backing…does it seem good to you for a design session ?17:51
thangpso for objects, we can transform a object of a certain version to a desired one17:51
DuncanTvilobhmm1: Seems reasonable, I'll need a bit of time to read. Can you add both links to the session etherpad, please?17:51
thingeevilobhmm1: it being accepted in nova isn't exactly a cross project agreement17:51
thingeethat's an agreement in nova17:51
*** buster_604 has joined #openstack-cinder17:52
*** jkremer has joined #openstack-cinder17:52
*** buster_604 has quit IRC17:52
vilobhmm1sure DuncanT…thingee: ok17:52
DuncanTthangp: are there any examples of that, even artificial ones?17:52
thangpyes, somewhere in nova, let me look17:52
DuncanTNova does it via conductor17:52
DuncanTWe should code up an example for cinder17:53
DuncanTIt would be nice if liberty never made a none-rolling friendly change, but that might slow down progress17:53
thangpDuncanT: I had thought it was done in obj_make_compatible17:53
thangpbut I can ask dansmith17:53
DuncanTthangp: I think that gets called in conductor context in nova17:54
* dansmith reads17:54
thangpDuncanT: ah ok17:54
* jungleboyj is back17:54
DuncanTthangp: A worked example in the cinder code would help clear things up, I think17:54
thangpDuncanT: ok, I can look into it17:54
DuncanTBack shortly, it is 9 pm and I need coffee17:55
thangpdansmith: we are wondering how to convert objects from one version to another in versionedobjects17:55
*** buster_604 has joined #openstack-cinder17:55
thangpDuncanT says it's done in the conductor layer17:55
dansmiththangp: yeah, so we send whatever version we want to the destination,17:55
* e0ne bb in 2hours17:55
dansmithand if the destination is too old to handle it, it bounces the object to conductor, which does the backleveling17:55
vilobhmm1*bb in 1 hour need to attend a meeting*17:56
dansmiththat way, we can have two new computes talk to each other without needing to be pinned,17:56
thangpbut since we dont have a conductor, what's the best way to do it, or do we have to implement a conductor17:56
dansmithand computes don't have to know whether other computes are old or not (which doesn't always work, but it's how we do it)17:56
dansmiththangp: so if you don't want to do that,17:56
dansmithyou just have a set of object pinsm17:56
dansmithand have your code check the pin and obj_make_compatible() the object to the pin version before sending it17:57
dulekHeat guys are already planning to do that this way.17:57
dansmither, correct :)17:57
*** e0ne has quit IRC17:57
thangpdulek: do you have a sample?17:57
dulekI'll see if I can find a spec for that.17:57
thangpdulek: cool, thx17:57
dansmithnova may end up doing that as well for certain cases17:57
thangpthingee: so for rolling upgrades, we can have versionedobjects and object pins (instead of conductor)...17:58
thangpthingee: still needs to be tested17:59
thingeeDuncanT: think that's what you've been discussing already ^17:59
*** harlowja_away is now known as harlowja18:00
dulekthangp: But I don't think this is really helpful one.18:00
thangpdulek: ok, i'll take a look18:00
dulekthangp: My colleague is working on that, I'll ask him tommorow about progress and the ideas.18:00
thangpdulek: great, thx18:01
*** emagana has joined #openstack-cinder18:01
*** marcusvrn has joined #openstack-cinder18:01
dulekHm, so had we agreed if upgrades will make a working session?18:02
thingeethangp: I thought duncan was already looking into version pinning c-vol nodes18:02
dulekFor me it certainly should.18:02
*** Yogi1 has quit IRC18:02
*** buster_604 has quit IRC18:02
thingeethangp: so seems to align with what has been discussed, unless I'm missing something18:02
thingeethangp: is there anything else you need from me on this topic?18:03
thingeeadurbin_:  rhedlind around?18:03
openstackgerritJoshua Huber proposed openstack/cinder: Add cinder volume driver for Blockbridge EPS.
thangpthingee: i good with it18:03
thangpthingee: thx18:03
thingeeyour topic for collect and display backend status18:03
*** marcusvrn1 has quit IRC18:04
*** dannywilson has quit IRC18:05
rhedlindyes, adurbin is working on it. we hope to have some sort of POC at the summit18:05
*** dhellmann has quit IRC18:05
rhedlindit would be nice to get user feedback to know what kind of stats are most useful to users18:05
thingeeadurbin_: can you please move the spec to the liberty directory?18:06
adurbin_we're looking at adding the ability for cinder to periodically query cinder volume drivers for performance and health stats18:06
thingeerhedlind: it would be good to list the use case for exposing this information in the spec18:06
adurbin_thingee: yes18:06
*** Yogi1 has joined #openstack-cinder18:06
*** lpetrut has quit IRC18:06
*** angela-s has joined #openstack-cinder18:07
rhedlindthingee: ok, we'll come up with use cases18:07
thingeerhedlind: think we'll need to understand that more before we can decide on making this a summit session. So I'll wait around for that and bug people for some feedback18:08
thingeethangp: just to make sure I understand, we're doing pinning for rpc compat?18:08
adurbin_initial use case is for horizon to display the last known performance values18:08
thingeethangp: or just object versioning?18:08
thingeeadurbin_: I'm assuming this is admin viewable only?18:09
*** buster_604 has joined #openstack-cinder18:09
adurbin_thingee: yes18:10
*** buster_604 has quit IRC18:10
thingeeadurbin_: its been my experience that most vendors have some proprietary dashboard that's already good at giving stats18:11
thingeeI can see the case of wanting everything in horizon. just not sure if it's worth the effort though18:11
thingeethe Iops thing will be interesting. As you may or may not know, this is a controversial topic here, since everyone thinks of IOPS differently18:12
smcginnisvilobhmm1: Hey! Did you see my comment here:
smcginnisvilobhmm1: Let me know if I'm missing something.18:12
adurbin_thingee: ah, I was unaware of the controversial-ness18:13
*** rushil has quit IRC18:14
rhedlindthingee: so how would you like us to move forward? we'll add more details to the spec as feedback comes back. Wait until next week for a decision on this being a summit session or not?18:14
*** dustins has joined #openstack-cinder18:15
* thangp scrolling back to read18:15
nikeshmthingee: hi good morning,could u please approve and target this bp for kilo
thangpthingee: both, versionobjects and object pins18:16
*** harlowja has quit IRC18:16
*** harlowja has joined #openstack-cinder18:16
nikeshmreached in sanjose,california18:16
*** harlowja has quit IRC18:17
winston-dnikeshm: you mean liberty? :)18:17
thingeerhedlind: might require a change in direction. For example, using the stats that is already being reported by drivers ... I fear that this topic will get hijacked by vendors arguing about what some of these terms mean to them.18:18
jdandreaannegentle: Is this where you were looking for secure_delete earlier? Wondering if it should be kept or just dropped entirely by now.
thingeenikeshm: not sure if my outstanding questions were answered in the review?18:19
annegentlejdandrea: actually it was here
annegentlejdandrea: by doing a google search with secure_delete site:docs.openstack.org18:19
thingeenikeshm: CI, who is running that? who is maintaining the driver. is using HP fine in the name if they're not maintaining it. I'm just a little confused atm18:19
jdandreaannegentle: Ahh, I didn't catch that. That's why I didn't see it. I was searching devstack. :)18:19
thingeekmartin_, hemna ^18:19
*** harlowja has joined #openstack-cinder18:20
annegentlejdandrea: yeah you want
jdandreaannegentle: Also, the formatting of that page is ... odd.18:20
*** buster_604 has joined #openstack-cinder18:20
thingeerhedlind adurbin_ let me read the spec more closely though18:20
thingeeI saw iops and just started having fears :)18:20
thingeewould like this to be successful and no one to get frustrated :)18:21
rhedlindrhedlind: ok, :) the spec proposes adding some values to volume stats.18:21
annegentlejdandrea: it was once beautiful, but I can recall how it's built18:21
jdandreaannegentle: *nodnod*18:21
rhedlindI am apparently talking to myself. :)18:21
*** vilobhmm1 has quit IRC18:22
*** scohen has quit IRC18:23
rhedlindthingee: I know DuncanT had some thoughts on this topic at the last meetup. Hopefully we'll get some feedback from him once he have some coffee.18:23
thingeerhedlind: I feel like that all the time18:23
DuncanTSorry, which topic?18:24
* DuncanT tries to read all the scrollback in three windows at once18:24
DuncanTVolume stats?18:24
rhedlindDuncanT: you're back. collecting more stats from the backend storage to give the admin a better picture of the current state.18:24
kmartin_thingee, what did you want me to look at?18:24
DuncanTAs I've (just) commented on the spec, I think the hypervisor is the right place to collect some of this18:25
DuncanTrhedlind: I'd suggest splitting the spec in two, and doing health status separate to stats since it is likely to be easier to agree on18:25
rhedlindDuncanT: ok18:25
thingeekmartin_: look at my message at 11:1918:25
thingeereply to nikeshm18:25
*** bswartz has quit IRC18:26
DuncanTrhedlind: A basic status of good/warning/bad plus a free-text field seems pretty universal18:26
DuncanTrhedlind: The default can be to assume all good unless the driver says otherwise18:26
*** jordanP has joined #openstack-cinder18:27
*** bswartz has joined #openstack-cinder18:28
DuncanTrhedlind: Stats are a bit more contentious. Depending on the design of your backend, collecting them at the backend can be hard. Nova already has a periodic task to emit the stats of any active volume(s) as events, I'd like to hear what is wrong with that aproach before we do something new18:28
thingeethangp: ok so we still need to figure out rpc compat?18:28
thingeethangp: confused :)18:28
*** markvoelker has quit IRC18:28
rhedlindDuncanT: Yes, I think we left out default behavior. Should add proposed default behavior to the spec.18:28
thangpthingee: so we will have some kind of object version pinning18:29
*** harlowja has quit IRC18:29
thingeeI get that. but how does that solve messaging?18:29
thangpthingee: we will look up what object version to send over, convert it, and then send over the serialized object18:29
*** harlowja has joined #openstack-cinder18:29
thangpthingee: not sure if it does18:29
DuncanTIt doesn't, though the pin management is identical18:30
thingeethangp: I don't think it does. What if the message contains field changes other than the message?18:30
thingeeother than the object*18:30
kmartin_thingee, what driver are they talking about the 3PAR, Lefthand, MSA?18:30
adurbin_DuncanT: latency at the compute level and storage array could be different18:30
thangpthingee: does oslo_messaging support something for it?18:30
*** IanGovett has joined #openstack-cinder18:31
*** buster_604 has quit IRC18:31
*** shyama has quit IRC18:31
DuncanTadurbin_: But is it different enough to be useful? In particular, who cares what it is at the array level since the VM level is what actually affects what tennants see18:31
*** e0ne has joined #openstack-cinder18:31
*** e0ne is now known as e0ne_18:31
hemnajungleboyj, I don't know.   there used to be a guy in there, but I think they bailed18:31
DuncanTthangp: I'm told it does, I haven't found it18:31
thangpDuncanT: :P18:31
thangpDuncanT: who maintains oslo_messaging?18:32
DuncanTthangp: I'm love to hear somebody else was working on it ;-)18:32
DuncanTthangp: Dunno, sorry18:32
*** e0ne_ is now known as e0ne18:32
*** e0ne is now known as e0ne_18:32
*** e0ne_ is now known as e0ne18:32
*** e0ne has quit IRC18:32
nikeshmthingee: sorry this bp is for liberty18:32
jungleboyjkmartin_: Do you know who the contact for Cisco Zone Manger is?18:33
DuncanTthangp: I've got a good breakdown of how to go about implementing RPC versioning (including testing etc)18:33
thingeenikeshm: I know18:33
thangpDuncanT: nice, can you help moderate that topic in the working session?18:33
thingeenikeshm: that's not what my question was though18:33
DuncanTthangp: But I ran into an early roadblock with the basic code and never dug into it18:33
DuncanTthangp: Yes, definitely18:33
thangpDuncanT: do you have it on gerrit?18:33
DuncanTthangp: I'll brain-dump onto an etherpad before hand too18:33
thingeeDuncanT, thangp can this happen before the working session? so there is something to go off of?18:34
DuncanTthangp: HP etherpad at the moment, I'll copy it over to a public one the minute I find my keyring for the VPN18:34
DuncanTthingee: Definitely18:34
thangpthingee: I can got the object pinning part18:34
kmartin_thingee, regarding BP this is a company by the named of Dot Hill, they will be CI'ing the driver18:34
adurbin_DuncanT: we can split the spec into perf and health and spend more time on performance18:34
*** thangp has quit IRC18:35
*** thangp has joined #openstack-cinder18:35
thingeethangp: please working with DuncanT on the rpc stuff so we can have something to go off of at the working session. I don't want DuncanT's brain dump to be happening at the working session18:35
thangpthingee: ok :)18:35
DuncanTthangp: Please feel free to unsubtly harass me, I don't get offended and I'm really easily distracted18:36
thangpDuncanT: sure18:36
DuncanTadurbin_: That sounds like a good way to make faster progress18:36
thangpi will get my part in order before i do18:36
*** annegentle has quit IRC18:36
*** emagana has quit IRC18:36
DuncanTthangp: :-)18:36
thangpmaybe by next week i get something for it18:36
kmartin_jungleboyj, looking I'll PM you the emails addresses18:36
jungleboyjkmartin_: You the man.  Thanks!18:37
adurbin_thingee: i'm interested to hear your thoughts on the 'fears' about iops18:38
jdandreaClue, please. Where is the repository that contains this?
thingeeadurbin_: sure some backends don't even expose that. iops for some are bronze/silver/gold. others it's a on/off. and then some have actual iops numbers18:41
*** rushil has joined #openstack-cinder18:41
nikeshmthingee: dothill is responsible for drivers and CI18:41
jdandreaGrabbed  but it isn't in there.18:42
*** buster_604 has joined #openstack-cinder18:42
thingeekmartin_: is the HP in the name still fine?18:42
*** emagana has joined #openstack-cinder18:43
kmartin_thingee, That's OK as long as dothill is listed as the maintainer and CI owner18:44
thingeekmartin_: thanks18:44
kmartin_nikeshm, you're from dothill correct?18:45
thingeenikeshm: ok done18:45
thingeewe'll require to see the CI working up front before merge. see
thingeenikeshm: ^18:46
thingeenikeshm: also review
thingeenikeshm: thank you!18:47
*** e0ne has joined #openstack-cinder18:48
*** ho has joined #openstack-cinder18:48
*** annegentle has joined #openstack-cinder18:49
jungleboyjthingee: Thanks for accepting the Replication V2 session.18:50
jungleboyjDon't see a comment on Volume Migration.18:50
jdandreaAh, found it.18:50
jungleboyjThoughts there?18:50
jungleboyjOh, looks like sprints now that I look closer.18:51
jungleboyjThat should be ok.  Think we should discuss it in some way.18:51
jungleboyjThanks for listening all while I talked to myself.  ;-)18:52
e0nejungleboyj: you're welcome!18:52
*** annegentle has quit IRC18:52
*** annegentle has joined #openstack-cinder18:53
*** annashen has joined #openstack-cinder18:55
nikeshmthingee: thanks,sure,18:56
nikeshmkmartin_: i am from vedams,working for dothill for hp drivers18:57
*** erlon has joined #openstack-cinder18:58
kmartin_nikeshm, ok cool...did you have a review posted for driver or just the BP?18:59
nikeshmnow just BP,sooner will have review18:59
thingeealright bbl19:00
kmartin_nikeshm, looking forward to it? thanks19:00
nikeshmwinston-d : liberty right :)19:00
*** timcl1 has joined #openstack-cinder19:01
*** emagana has quit IRC19:01
*** annashen has quit IRC19:01
*** vilobhmm1 has joined #openstack-cinder19:01
*** annashen has joined #openstack-cinder19:02
*** dustins has quit IRC19:03
*** timcl has quit IRC19:03
*** emagana has joined #openstack-cinder19:04
*** MentalRay has quit IRC19:05
*** MentalRay has joined #openstack-cinder19:05
*** dhellmann has joined #openstack-cinder19:06
*** annashen has quit IRC19:06
*** jaypipes has quit IRC19:06
*** Lee1092 has joined #openstack-cinder19:08
*** ronis has quit IRC19:11
*** vilobhmm1 has quit IRC19:13
*** MentalRay has quit IRC19:13
*** rushiagr is now known as rushiagr_away19:15
*** marcusvrn has quit IRC19:15
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** MRay has joined #openstack-cinder19:16
*** vilobhmm1 has joined #openstack-cinder19:17
*** marcusvrn has joined #openstack-cinder19:18
*** ganso__ has quit IRC19:19
*** vilobhmm1 has quit IRC19:20
*** timcl1 has quit IRC19:20
*** rushil has quit IRC19:20
openstackgerritOpenStack Proposal Bot proposed openstack/python-cinderclient: Updated from global requirements
*** timcl has joined #openstack-cinder19:23
*** vilobhmm1 has joined #openstack-cinder19:23
*** jaypipes has joined #openstack-cinder19:23
*** rushil has joined #openstack-cinder19:23
*** rushil has quit IRC19:24
*** _cjones_ has quit IRC19:25
*** Yogi1 has quit IRC19:25
*** dustins has joined #openstack-cinder19:25
*** vilobhmm1 has quit IRC19:26
*** luv_ has joined #openstack-cinder19:27
*** luv_ has quit IRC19:27
*** luv_ has joined #openstack-cinder19:28
*** jdandrea has quit IRC19:28
*** jdandrea has joined #openstack-cinder19:28
*** akshai has quit IRC19:28
*** nkrinner has joined #openstack-cinder19:29
*** luv_ is now known as luv19:29
*** _cjones_ has joined #openstack-cinder19:41
*** david-lyle has quit IRC19:44
*** david-lyle has joined #openstack-cinder19:44
*** anshul has joined #openstack-cinder19:48
*** rooneym has joined #openstack-cinder19:49
*** dannywilson has joined #openstack-cinder19:51
*** dannywilson has quit IRC19:52
*** dannywilson has joined #openstack-cinder19:52
*** ho has left #openstack-cinder19:53
*** arif-ali has quit IRC19:56
openstackgerritDiem Tran proposed openstack/cinder: Re-integrate Oracle iSCSI Cinder driver
*** dencaval has quit IRC19:58
*** jordanP has quit IRC20:01
*** uberjay has quit IRC20:01
*** ganso__ has joined #openstack-cinder20:04
*** uberjay has joined #openstack-cinder20:05
*** eharney has quit IRC20:07
*** esker has quit IRC20:08
*** akerr has quit IRC20:08
*** Rockyg has joined #openstack-cinder20:08
*** emagana has quit IRC20:11
*** jungleboyj has quit IRC20:11
*** emagana has joined #openstack-cinder20:14
*** annegentle has quit IRC20:15
*** scohen has joined #openstack-cinder20:19
*** scohen has quit IRC20:20
*** marcusvrn has quit IRC20:23
*** annegentle has joined #openstack-cinder20:23
*** jungleboyj has joined #openstack-cinder20:24
*** e0ne has quit IRC20:26
*** Apoorva has quit IRC20:35
*** jungleboyj has quit IRC20:35
*** e0ne has joined #openstack-cinder20:36
*** MRay has quit IRC20:39
*** charlesr has joined #openstack-cinder20:41
*** eharney has joined #openstack-cinder20:41
*** akerr has joined #openstack-cinder20:41
*** MRay has joined #openstack-cinder20:41
*** e0ne has quit IRC20:44
*** mtreinish_ has joined #openstack-cinder20:45
*** dustins has quit IRC20:48
thingeejgriffith: not sure if you saw this, but should take care of the warnings from cinderclient
thingeejgriffith: would like to do a release soon :)20:49
*** xyang has quit IRC20:50
jgriffiththingee: awesome20:50
jgriffith+2/A :)20:50
thingeeyay ninja!20:51
*** mtreinish has quit IRC20:51
*** mtreinish_ is now known as mtreinish20:51
* jgriffith just wants that to go away :)20:51
*** dulek has quit IRC20:52
*** Apoorva has joined #openstack-cinder20:53
*** Apoorva has quit IRC20:54
*** bswartz has quit IRC20:54
*** dannywilson has quit IRC20:57
*** dannywilson has joined #openstack-cinder20:58
*** akerr_ has joined #openstack-cinder20:59
*** ganso__ has quit IRC21:00
*** anshul has quit IRC21:00
*** akerr has quit IRC21:02
*** esker has joined #openstack-cinder21:04
*** lpabon has quit IRC21:05
*** Apoorva has joined #openstack-cinder21:07
*** MRay has quit IRC21:10
*** MRay has joined #openstack-cinder21:15
*** charlesr has quit IRC21:15
thingeejgriffith: other problems
thingeejgriffith: mtreinish tested for it being related to osprofiler, but no luck
*** mriedem has quit IRC21:17
mtreinishit did make the unit tests much cleaner though21:18
*** arif-ali has joined #openstack-cinder21:19
*** annashen has joined #openstack-cinder21:21
*** MRay has quit IRC21:24
*** nkrinner has quit IRC21:25
*** Longgeek has quit IRC21:29
*** Rockyg has quit IRC21:30
*** julim has quit IRC21:30
*** annashen has quit IRC21:33
*** vilobhmm1 has joined #openstack-cinder21:33
*** vilobhmm1 has quit IRC21:34
*** vilobhmm1 has joined #openstack-cinder21:34
*** IanGovett has quit IRC21:34
*** harlowja is now known as harlowja_away21:36
*** akerr_ has quit IRC21:37
openstackLaunchpad bug 1449980 in Cinder "Attaching the same multiattach volume to the same instance twice leaves the volume in 'Attaching' state" [Medium,Confirmed]21:37
*** harlowja_away is now known as harlowja21:37
hemnathingee, thanks I'll take a look21:38
thingeehemna: also
jgriffiththingee: what am I looking at?21:39
*** vilobhmm1 has quit IRC21:39
thingeejgriffith: your monitor?21:39
jgriffiththingee: holy crap!21:39
*** esker has quit IRC21:39
* thingee takes a screenshot21:39
*** MRay has joined #openstack-cinder21:40
*** esker has joined #openstack-cinder21:40
jgriffiththingee: wait... so what bug is this?21:40
thingeeuhh mtreinish is there one?21:40
thingeefor the token not found in header w/ cinder21:40
mtreinishthingee: I never bothered to file a bug for it21:41
* thingee makes one21:41
jgriffithmtreinish: how long has that been around?21:41
mtreinishjgriffith: about a week21:42
mtreinishat least according to logstash21:42
*** thangp has quit IRC21:42
mtreinishwhich has >3.3M hits for that warning :)21:42
thingeemaaaybe discovery related jgriffith21:42
thingeeI was talking to jamielennox about it yesterday21:42
thingeejgriffith: it is the recommended way to do discovery =/21:43
jgriffithperhaps we need a new recommendation :)21:44
thingeeperhaps we just need to move to the openstack client and having version/tenant in our endpoint21:45
thingeeaway from having*21:45
*** theanalyst has quit IRC21:46
*** theanalyst has joined #openstack-cinder21:49
*** annegentle has quit IRC21:51
*** bnemec has quit IRC21:52
*** vilobhmm1 has joined #openstack-cinder21:53
*** esker has quit IRC21:55
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: cinder os-detach api returns 500
jgriffiththingee: whatever works21:57
*** bnemec has joined #openstack-cinder22:02
*** melwitt has quit IRC22:02
*** melwitt has joined #openstack-cinder22:02
*** patrickeast_ has joined #openstack-cinder22:04
*** patrickeast has quit IRC22:05
*** patrickeast_ is now known as patrickeast22:05
*** daneyon has joined #openstack-cinder22:06
daneyonI set iscsi_helper = lioadm in cinder.conf of my cinder-volume node, but I still see cinder using tgt when attaching a vol to an instance: tgt-admin --update execute22:08
daneyon^ Is this expected behavior?22:08
jgriffithdaneyon: nope, that's no good22:09
jgriffithdaneyon: however...22:09
jgriffithdaneyon: did you install/configure LIO?22:09
daneyoni installed targetcli22:09
jgriffithdaneyon: are you using devstack by chance?22:09
jgriffithdaneyon: so RHEL/CentOS I *thought* used LIO by default anyway22:10
jgriffithdaneyon: so it should be right22:10
daneyonnope, openstack-cinder-2014.2.2-1.el7.noarch22:10
jgriffithdaneyon: oh... confused me with "yum"22:10
daneyonCentOS Linux release 7.1.1503 (Core22:10
jgriffithyeah... CentOS... I know RHEL uses LIO by default, always thought CentOS would as well22:11
jgriffithshows what I know :(22:11
jgriffithhonestly I've never had much luck getting LIO to work correctly on anyhting other than RHEL22:12
jgriffithwhich just "did it for me"22:12
daneyoni'm a core for the kolla project: I am trying to containerize cinder and having problems with cinder-volume attaching the vol to an instance22:12
jgriffithdaneyon: oh, cool22:12
jgriffithdaneyon: That's become my favorite new stackforge project BTW :)22:13
jgriffithOk... so back to the task at hand22:13
jgriffithHonestly I'll have to spin up a system and poke at it22:13
jgriffithI don't have good answers for you off the top of my head22:13
daneyonjgriffith i have always used tgt, i was just trying to follow the juno docs. is their any issue running tgt with Centos 7? I ran across this that made it seems lioadm is required for centos7:
openstackgerritMike Perez proposed openstack/cinder: Change MODULEPATH env variable in config generator
jgriffithLike I say, I've had nothing but problems with LIO on anything but RHEL22:14
jgriffithdaneyon: not that I'm aware of (no issues that is)22:14
daneyonjgriffith let we switch back to tgt22:14
jgriffithdaneyon: let me know if that works22:14
thingeejgriffith: last we spoke, you figured out ubuntu/lio22:14
daneyonwill do22:14
jgriffiththingee: ummm... no, not really.  I got it "kinda" working22:15
thingeeah bummer22:15
*** jkremer has quit IRC22:15
thingeedaneyon: you will have problems if you use rtslib-fb22:15
jgriffiththingee: but it had a shit ton of manual steps; kernel update, missing things, and broken conf usage22:15
jgriffiththere we go with that whole thing again :(22:15
thingeedaneyon: it's a mess, but a rhel fork (rtslib-fb) exists and has issues on debian based systems22:15
daneyonthingee you are referring to lioadm, correct?22:16
thingeejgriffith: again, works fine for me if you just build from source.
thingeejgriffith: also if you want to talk about debian based packages, there's a lot of drama with the previous maintainer of the package.22:17
thingeejgriffith: would take that discussion offline though22:17
thingeebut he's no longer part of debian, lets put it that way22:17
thingeedaneyon: yup22:18
thingeedaneyon: I work for the company that originally wrote rtslib.22:18
jgriffiththingee: got it22:18
jgriffith"It worked on thingee 's machine" :)22:19
jgriffithThe next T-shirt for the summit ;)22:19
thingeejgriffith: there's a lot of drama. Really I just want to fork it and call it something else so I don't have to deal with rhel drama of *-fb confusing people22:19
jgriffiththingee: yeah, I don't know much about that stuff unfortunately (or maybe fortunately) :)22:20
thingeealso there is a lot of legacy HBA crap that I don't think needs to exist in the interface. it's a mess22:20
jgriffithThat's a shame22:20
thingeeand the switch from everything to be a generator returned instead of lists22:20
thingeethat's coming in the upcoming version. whether it's rtslib-fb or not22:21
thingeeupcoming major version22:21
thingeeto be specific22:21
daneyonin my setup i'm using a loopback device for testing. are their any gotchya's or should the loopback work like a regular disk from a cinder-vol config perspective?22:21
*** akerr has joined #openstack-cinder22:22
jgriffithdaneyon: loopback is great for test/dev work22:22
jgriffithdaneyon: no real gotchas22:22
jgriffithdaneyon: you'll want to disable secure_delete though22:23
daneyonok, i didn't think so but wanted to double-check. i'm able to create the pv/vg and cinder creates the lv's just fine using it.22:24
thingeejgriffith: ha that's two machines it works on22:24
daneyonwill do22:25
vilobhmm1smcginnis : updated
*** bswartz has joined #openstack-cinder22:27
*** esker has joined #openstack-cinder22:27
thingeejgriffith: lvm snapshots
thingeetwo words22:32
thingeemore like acronym + word22:32
*** MRay has quit IRC22:33
jgriffiththingee: can you give me any context here?22:34
*** ikbb has joined #openstack-cinder22:35
openstackLaunchpad bug 1450110 in Cinder "boto tests randomly fail in resource_cleanup with "Failed to delete volume"" [High,Confirmed]22:37
*** melwitt has quit IRC22:41
*** melwitt has joined #openstack-cinder22:41
*** buster_604 has quit IRC22:42
*** MRay has joined #openstack-cinder22:52
*** ndipanov has quit IRC22:54
*** angela-s has quit IRC22:57
thingeejgriffith: lol this thread
openstackDebian bug 659762 in lvm2 "lvm2 - Failed snapshot removal produces suspended devices" [Important,Fixed]22:57
thingeejgriffith: I know this is snapshot removal, but just run the command twice will fix your problem22:57
thingeejgriffith: what a mess22:58
thingeejgriffith: also looks like something you previously tried to fix?
openstackLaunchpad bug 1335905 in Cinder "Volume group for uuid not found" [High,Fix released] - Assigned to John Griffith (john-griffith)23:02
*** patrickeast has quit IRC23:03
*** dalgaaf has quit IRC23:07
*** dannywilson has quit IRC23:08
*** patrickeast has joined #openstack-cinder23:09
*** bkopilov has quit IRC23:10
*** dannywilson has joined #openstack-cinder23:11
*** buster_604 has joined #openstack-cinder23:13
*** MRay has quit IRC23:15
*** kmartin_ has quit IRC23:17
*** kmartin has joined #openstack-cinder23:17
*** Rockyg has joined #openstack-cinder23:20
*** MRay has joined #openstack-cinder23:22
*** rmesta has quit IRC23:24
*** bkopilov has joined #openstack-cinder23:25
*** MRay has quit IRC23:26
openstackgerritJohn Griffith proposed openstack/cinder: Add retry to lvm delete
*** melwitt has quit IRC23:33
*** akerr is now known as akerr_away23:36
*** akerr_away is now known as akerr23:40
*** emagana has quit IRC23:40
*** dims has quit IRC23:50
*** chlong has joined #openstack-cinder23:50
*** annashen has joined #openstack-cinder23:52
*** dannywil_ has joined #openstack-cinder23:52
*** dannywilson has quit IRC23:55
*** dannywil_ has quit IRC23:57
*** Tross has quit IRC23:57
*** yamada-h has joined #openstack-cinder23:58
*** buster_604 has quit IRC23:59

Generated by 2.14.0 by Marius Gedminas - find it at!