Thursday, 2015-02-05

anishwell, dropped but patchset is not updated yet00:00
jungleboyjanish: We can't directly edit openstack/common files.  They have to come via a sync from oslo_incubator.00:00
jungleboyjI need to finish up the process of making the other namespace changes before I we can clean up oslo_incubator.00:00
jungleboyjHoping to work on that next week.00:00
jungleboyj*Crossing fingers*00:00
thingeevilobhmm: we still have some things that need to get done in k-2, so might not be able to today00:00
anishjungleboyj: right, I'm only talking about setup.cfg and the two files outside of openstack/common00:01
*** takedakn has joined #openstack-cinder00:01
vilobhmmthingee : thats fine00:02
*** jnrao has quit IRC00:03
jungleboyjanish: Eventually we need to get everything moved to using oslo_ instead of oslo.00:03
jungleboyjI just haven't finished the merges yet.00:03
*** dkingshott has quit IRC00:05
*** takedakn has quit IRC00:05
*** ebalduf has joined #openstack-cinder00:06
*** ebalduf has quit IRC00:16
*** ho has joined #openstack-cinder00:17
*** foobloo has left #openstack-cinder00:18
*** scottda_ has joined #openstack-cinder00:20
*** annegent_ has quit IRC00:20
*** scottda_ has quit IRC00:21
openstackgerritAnish Bhatt proposed openstack/cinder: Change oslo.* to oslo_*
anishsmcginnis: jungleboyj significantly smaller patch now ^00:24
openstackgerritKurt Martin proposed openstack/cinder: Lefthand driver fails to attach a cloned volume
*** gtt116_ has joined #openstack-cinder00:27
smcginnisanish: Maybe jungleboyj can clarify, but I think you might just want to abandon that patch and let him complete the migration.00:27
anishI hit more automated stuff didn't I ?00:28
anishbut sure, just give the word00:28
*** gtt116__ has quit IRC00:30
*** mriedem has joined #openstack-cinder00:32
*** annegent_ has joined #openstack-cinder00:33
smcginnisYeah, just hold on until jungleboyj can give details. He's the Captain. :)00:38
*** annegent_ has quit IRC00:38
*** diegows has quit IRC00:40
*** ebalduf has joined #openstack-cinder00:41
openstackgerritlvdongbing proposed openstack/python-cinderclient: Make query params of url in fixed order when list volumes
*** lcurtis has quit IRC00:44
*** annegent_ has joined #openstack-cinder00:52
*** annegent_ has quit IRC00:52
*** _cjones_ has quit IRC00:53
*** takedakn has joined #openstack-cinder00:53
*** ebalduf has quit IRC00:53
*** madskier has joined #openstack-cinder00:57
*** takedakn has quit IRC00:57
*** lcurtis has joined #openstack-cinder00:57
*** nellysmitt has joined #openstack-cinder00:58
*** oomichi has joined #openstack-cinder00:58
*** madskier has quit IRC01:02
*** nellysmitt has quit IRC01:02
*** chlong has joined #openstack-cinder01:05
smcginnisHate to ask, but if any cores have a few minutes to take a look:01:11
jdgsmcginnis: ask and ye shall recieve01:12
jdgoh... sorry; *receive*01:12
smcginnisjdg: Thank you sir. :)01:12
smcginnisjdg: Not sure if you were around earlier, but you might be interested in
jdgsmcginnis: oh.. yeah!  That would be awesome!01:13
smcginnisjdg: Could be cool.01:14
jdgsmcginnis: yeah, I think it would be a good one :)01:14
*** alop has quit IRC01:16
*** david-lyle is now known as david-lyle_afk01:18
*** rushil has quit IRC01:19
*** tsekiyam_ has joined #openstack-cinder01:19
*** jamielennox|away is now known as jamielennox01:19
*** chlong has quit IRC01:22
*** mtanino has quit IRC01:22
*** tsekiyama has quit IRC01:22
*** annegent_ has joined #openstack-cinder01:23
*** tsekiyam_ has quit IRC01:23
*** DaveChen has joined #openstack-cinder01:28
jungleboyjanish: smcginnis Responded to the review.01:29
smcginnisAye aye captain. :)01:32
*** annegent_ has quit IRC01:35
openstackgerritAnish Bhatt proposed openstack/cinder: Change oslo.* to oslo_*
anishjungleboyj: oh captain my captain, your wish is done01:37
anishoh nice, the bug gave me HIGH importance for like 3 lines of change01:38
*** dannywilson has quit IRC01:41
smcginnisanish: ;)01:41
*** mwichmann has quit IRC01:42
*** jcru has quit IRC01:42
*** rwsu is now known as rwsu-afk01:43
*** MentalRay has joined #openstack-cinder01:43
anishthingee: jdg your cutoff deadline is like 15 min away, do you need anything from me on ? or is the cutoff just for uploading code for review ?01:44
jdganish: should just be upload, but you'll have to defer to thingee01:47
jdganish: or *I* will :)01:47
anishshould be fine, last I heard from him, he was okay with it provided the tests were moved to mock, and that part is done01:47
anishjdg: also, I'm gonna abandon  (we talked about this a long time ago) unless you have a suggestion01:48
jdganish: yeah, I would if I were you01:50
*** vilobhmm has quit IRC01:51
jdganish: sometimes just not worth it :(01:52
*** markvoelker has joined #openstack-cinder01:54
*** vilobhmm has joined #openstack-cinder01:55
*** EmilienM is now known as EmilienM|afk01:55
*** ebalduf has joined #openstack-cinder01:58
*** Mandell has quit IRC01:59
*** vilobhmm has quit IRC02:02
*** leeantho has quit IRC02:02
*** ebalduf has quit IRC02:02
*** MentalRay has quit IRC02:04
*** casusbel_ has joined #openstack-cinder02:09
*** kaisers1 has joined #openstack-cinder02:09
*** mberlin has joined #openstack-cinder02:09
*** kaisers has quit IRC02:10
thingeeanish: I got some comments for it. We'll have to punt it.02:10
thingeesome are nit pick, but I have one biggie .02:11
*** mberlin1 has quit IRC02:11
*** casusbelli has quit IRC02:12
*** atan8 has joined #openstack-cinder02:14
*** haomaiwang has joined #openstack-cinder02:16
*** bill_az_ has quit IRC02:19
*** ho has quit IRC02:20
*** asmith_ has quit IRC02:20
*** chlong has joined #openstack-cinder02:27
*** zhithuang has joined #openstack-cinder02:28
*** zhithuang is now known as winston-d_02:28
*** tsekiyama has joined #openstack-cinder02:29
*** tellesnobrega_ has quit IRC02:31
*** tsekiyama has quit IRC02:33
*** Lee1092 has joined #openstack-cinder02:36
*** ho has joined #openstack-cinder02:37
*** rushiagr_away is now known as rushiagr02:40
*** chlong has quit IRC02:40
*** jcru has joined #openstack-cinder02:43
*** jcru has quit IRC02:47
*** tellesnobrega_ has joined #openstack-cinder02:53
*** tbarron has joined #openstack-cinder02:55
*** Longgeek has joined #openstack-cinder02:57
*** nellysmitt has joined #openstack-cinder02:58
*** ebalduf has joined #openstack-cinder02:59
*** tellesnobrega_ has quit IRC02:59
tbarronthingee: jenkins +1ed , but only after a recheck, so after the 02:00 UTC cutoff03:02
tbarrontoo bad we submitted upstream w/o checking that all unit tests passed the firs time03:03
*** ebalduf has quit IRC03:03
tbarronfirst time03:03
*** nellysmitt has quit IRC03:04
*** IanGovett has quit IRC03:04
*** tbarron has quit IRC03:10
*** rushil has joined #openstack-cinder03:11
*** markstur_ has joined #openstack-cinder03:11
*** rushil has quit IRC03:11
*** tellesnobrega_ has joined #openstack-cinder03:12
*** markstur_ has left #openstack-cinder03:12
*** MentalRay has joined #openstack-cinder03:14
*** rushiagr is now known as rushiagr_away03:18
openstackgerritTina Tang proposed openstack/cinder: Pool-aware scheduler support in EMC VNX Cinder driver
*** MentalRay has quit IRC03:27
openstackgerritBob proposed openstack/cinder: Remove LUN controller change
openstackgerritBob proposed openstack/cinder: Do a check before associate a LUN to a LUN group
openstackgerritBob proposed openstack/cinder: Fix problems under multipath
*** xyang has joined #openstack-cinder03:29
xyangwinston-d: ping03:29
xyangwinston-d_: ping03:29
*** chlong has joined #openstack-cinder03:30
*** bkopilov has quit IRC03:34
*** chlong has quit IRC03:37
*** winston-d_ has quit IRC03:39
*** mriedem has quit IRC03:39
*** theanalyst has quit IRC03:41
*** annashen has joined #openstack-cinder03:41
*** winston-d_ has joined #openstack-cinder03:42
winston-d_xyang: pong03:42
winston-d_xyang: hi03:42
*** rushiagr_away is now known as rushiagr03:43
*** theanalyst has joined #openstack-cinder03:44
*** hemna has quit IRC03:44
xyangwinston-d_: can you take a look of my comments on that patch?03:45
winston-d_xyang: sure03:45
winston-d_xyang: yes, when total in ['infinite', 'unknown'] and reserved = 0, that filter should return True.03:47
winston-d_xyang: no need to convert that to a number.03:47
xyangwinston-d_: ok, I'll make that change and re-submit then03:47
xyangwinston-d_: thanks!03:47
openstackgerritBob proposed openstack/cinder: Remove LUN controller change, etc
*** Longgeek has quit IRC03:53
*** harlowja is now known as harlowja_away03:56
*** ebalduf has joined #openstack-cinder04:00
*** patrickeast has quit IRC04:03
*** ebalduf has quit IRC04:04
*** BharatK has joined #openstack-cinder04:06
*** Longgeek has joined #openstack-cinder04:10
openstackgerritxing-yang proposed openstack/cinder: Support over subscription in thin provisioning
*** xyang has quit IRC04:23
*** harlowja_at_home has joined #openstack-cinder04:29
*** xyang1 has joined #openstack-cinder04:31
*** julim has joined #openstack-cinder04:32
xyang1winston-d_: Hi, I made the change and resubmitted.  Can you take a look again?04:33
*** lpetrut has joined #openstack-cinder04:35
*** chlong has joined #openstack-cinder04:36
*** dannywilson has joined #openstack-cinder04:38
openstackgerritxing-yang proposed openstack/cinder: Fix detach volume from host problem in VMAX driver
*** winston-d_ has quit IRC04:42
*** dannywilson has quit IRC04:42
anishthingee: regarding the temp dir ?04:46
anishprime reason was I followed the tgt driver test file, which does this as well04:49
anishbeyond that, there is no explicit cleanup as the cleanup is already happening04:49
anishi'm not sure who is cleaning up after, but importing tempfile is already removing the directory once tests are run04:50
anisherr, that sentence was badly constructed04:51
thingeeanish: Ok cool, thanks for the explanation. I'll check myself before I wreck myself.04:51
thingeeanish: we'll get it in k-3, promise :)04:51
anishno worries04:52
anishI'll cleanup the tests file in the meantime04:52
anishcurrently it is written in a i-have-half-a-day-to-learn-mock format04:52
openstackgerritxing-yang proposed openstack/cinder: EMC VMAX driver Kilo update
*** chlong has quit IRC04:54
*** nellysmitt has joined #openstack-cinder04:59
*** ebalduf has joined #openstack-cinder05:01
*** bkopilov has joined #openstack-cinder05:01
*** haomaiwang has quit IRC05:02
*** annashen has quit IRC05:03
*** chlong has joined #openstack-cinder05:03
*** atan8 has quit IRC05:03
*** nellysmitt has quit IRC05:04
*** ebalduf has quit IRC05:05
*** jcru has joined #openstack-cinder05:13
*** nshaikh has joined #openstack-cinder05:19
*** harlowja_at_home has quit IRC05:21
*** jcru has quit IRC05:21
*** lcurtis has quit IRC05:33
*** lpetrut has quit IRC05:33
*** Mandell has joined #openstack-cinder05:36
*** coolsvap_ is now known as coolsvap05:49
*** Mandell has quit IRC05:55
*** tellesnobrega_ has quit IRC05:58
*** sgotliv has joined #openstack-cinder05:59
*** Mandell has joined #openstack-cinder06:01
*** ebalduf has joined #openstack-cinder06:01
*** sgotliv has quit IRC06:03
*** ebalduf has quit IRC06:06
*** casusbel_ has quit IRC06:09
*** casusbelli has joined #openstack-cinder06:09
*** abehl has joined #openstack-cinder06:13
*** Ilja has joined #openstack-cinder06:17
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Transifex
*** tshefi has joined #openstack-cinder06:29
*** ronenkat has joined #openstack-cinder06:29
*** vilobhmm has joined #openstack-cinder06:44
*** abhijeetm has joined #openstack-cinder06:49
*** Mandell has quit IRC06:55
*** dulek has quit IRC06:55
*** nkrinner has joined #openstack-cinder06:59
*** nellysmitt has joined #openstack-cinder07:00
*** Mandell has joined #openstack-cinder07:01
nikesh_vedams_thingee jdg : good evening07:01
thingeehey nikesh_vedams_ I got your message. we just finished the review stuff for k-207:01
thingeenikesh_vedams_: I'm beat07:01
thingeeso scst is moved over to k-307:02
*** ebalduf has joined #openstack-cinder07:02
openstackgerritwuyuting proposed openstack/python-cinderclient: Make cinderclient support snapshot-force-delete
nikesh_vedams_thingee : ok take some rest,any weekend plans :)07:05
*** nellysmitt has quit IRC07:05
thingeetraveling, but presenting on openstack stuff07:05
thingeeso not really a vacation07:05
*** zhithuang has joined #openstack-cinder07:05
*** markvoelker has quit IRC07:06
*** zhithuang is now known as winston-d_07:06
*** ebalduf has quit IRC07:06
*** triggerz has quit IRC07:09
*** Ilja has quit IRC07:21
*** harlowja_away has quit IRC07:23
*** ronenkat has quit IRC07:26
*** nlevinki has joined #openstack-cinder07:28
*** annashen has joined #openstack-cinder07:30
*** lpetrut has joined #openstack-cinder07:37
*** chlong has quit IRC07:39
*** annashen has quit IRC07:41
*** tshefi has quit IRC07:44
*** tshefi has joined #openstack-cinder07:47
*** sgotliv has joined #openstack-cinder07:48
*** e0ne has joined #openstack-cinder07:51
*** afazekas has joined #openstack-cinder07:55
*** ronis has joined #openstack-cinder07:56
*** e0ne has quit IRC07:58
*** ebalduf has joined #openstack-cinder08:03
*** vilobhmm has quit IRC08:07
*** ebalduf has quit IRC08:07
*** Mandell has quit IRC08:11
*** Ilja has joined #openstack-cinder08:12
*** Mandell has joined #openstack-cinder08:12
*** sgotliv has quit IRC08:15
*** Mandell has quit IRC08:17
*** nellysmitt has joined #openstack-cinder08:17
*** karimb has joined #openstack-cinder08:20
*** cinderella has joined #openstack-cinder08:28
*** ndipanov has quit IRC08:28
cinderellagood morning to everyone08:29
*** stefan_amann has joined #openstack-cinder08:33
DaveChengood morning :)08:33
*** amann has joined #openstack-cinder08:33
*** ronenkat has joined #openstack-cinder08:35
*** ronenkat_ has joined #openstack-cinder08:36
cinderellaanybody can help me with a volume attach problem?08:37
*** ronenkat has quit IRC08:39
cinderellaDuncanT: I'm back :D08:53
*** Ilja has quit IRC08:54
*** jpich has joined #openstack-cinder08:55
*** alecv has joined #openstack-cinder08:56
cinderellaDuncanT: having new problems with CInder, may you help me?08:58
DuncanTcinderella: I can try, though I've a few other things going on so I might be slow responding. Any progress from yesterday?08:59
cinderellaDuncanT: no, seems like I went backwards09:01
cinderellanow I'm not able to create volumes... driver instances09:01
cinderelladriver issues*09:01
*** ebalduf has joined #openstack-cinder09:04
*** Longgeek has quit IRC09:05
*** jordanP has joined #openstack-cinder09:05
*** jistr has joined #openstack-cinder09:07
*** ebalduf has quit IRC09:08
*** lpetrut has quit IRC09:13
cinderellanothing... I can't attach volumes. Can anyone help me?09:17
*** Ilja has joined #openstack-cinder09:23
*** winston-d_ has quit IRC09:29
openstackgerritPranali Deore proposed openstack/cinder: Snapshot of bootable volume goes in error state
*** dulek has joined #openstack-cinder09:30
*** primechuck has joined #openstack-cinder09:32
*** gtt116_ has quit IRC09:34
*** avishay has joined #openstack-cinder09:37
*** Ilja has quit IRC09:38
*** Ilja has joined #openstack-cinder09:38
cinderellaseems like I solved. I'd like to share my solution with other users, what should I do?09:38
cinderellaDuncanT: if you're interested, I solved :)09:39
DuncanTcinderella: What was the solution?09:40
cinderellafirstly, I had to modify the nova,conf and add the string "iscsi_helper=ietadm", which I had in cinder.conf BUT NOT in nova.conf09:41
openstackgerritShay Halsband proposed openstack/cinder: XtreamIO version 4.0 support
cinderellathen, I discovered I had both tgt daemon and iscsi process running together09:41
*** ndipanov has joined #openstack-cinder09:41
cinderellafollowing a workaround on askopenstack, I stopped the tgt one, rebooted iscsi and cinder-volume processes, created a volume and attached09:41
cinderellapoof! worked like a charm09:42
cinderellabtw, this question helped me a lot:
cinderellaas you can guess, the answer from briancline was very complete09:43
*** lpetrut has joined #openstack-cinder09:43
cinderellaDuncanT: should I write all of this on launchpad?09:43
*** eharney has joined #openstack-cinder09:44
DuncanTI honestly think you should raise a bug that cinder fails with a hard-to-debug mode if tgt and iscsitarget are both installed - it can check that at driver init and log a decent error09:44
*** sgotliv has joined #openstack-cinder09:46
cinderellaDuncanT: I'm pretty new to this - I've always been a launchpad-reader, may you help me?09:46
DuncanTIt's easy though - go to and hit 'report a bug' in the top right09:47
DuncanTOnce you enter a title, it will suggest a list of similar bugs. If you don't see one that looks like your problem, then enter a new one. Describe what you saw in the logs ('portal not found') and how you fixed it, and suggest that cinder should provide better warning/error level logging in this case09:48
*** obutenko has joined #openstack-cinder09:49
*** mdbooth has quit IRC09:57
*** mdbooth has joined #openstack-cinder09:58
openstackgerritShay Halsband proposed openstack/cinder: XtreamIO version 4.0 support
*** obutenko has quit IRC10:01
*** obutenko_ has joined #openstack-cinder10:01
*** eharney has quit IRC10:03
*** marcusvrn has joined #openstack-cinder10:04
cinderellaDuncanT: I need to plug in a Swift component to my OpenStack - can you suggest me the best way to do this?10:04
*** ebalduf has joined #openstack-cinder10:04
DuncanTcinderella: If you're still at the devstack stage, just enable swift in devstack. Maybe look at multi-node devstack if you want to run it on a different server10:05
cinderellaDuncanT: I'm not in a devstack stage, I installed the components one by one on my server10:06
DuncanTAh. I've never install swift by hand, I'm afraid. You'll need to find a tutorial10:06
DuncanTI either use devstack, or our in-house installer10:06
*** Ilja has quit IRC10:07
cinderellaok, thanks :)10:07
*** avishay has quit IRC10:07
*** ebalduf has quit IRC10:09
*** lpetrut has quit IRC10:11
DuncanTGood luck. If you find a good tutorial that works, can you send me a link, please? I'm bound to want to do it at some point in the future10:14
cinderellais that good?10:19
cinderella"Muntaner" is my world wide web nickname10:19
*** lpetrut has joined #openstack-cinder10:21
*** nkrinner has quit IRC10:24
*** nkrinner has joined #openstack-cinder10:27
*** tellesnobrega_ has joined #openstack-cinder10:28
*** coolsvap is now known as coolsvap_10:29
*** BharatK has quit IRC10:31
DuncanTThat's great, thanks10:34
cinderellaDuncanT: what if I need tgt in the futurE?10:41
*** tellesnobrega_ has quit IRC10:43
*** tellesnobrega_ has joined #openstack-cinder10:45
*** chlong has joined #openstack-cinder10:50
*** boris-42 has quit IRC10:52
*** alexpilotti has joined #openstack-cinder10:56
*** Ilja has joined #openstack-cinder10:58
*** Ilja has quit IRC11:01
*** ronenkat__ has joined #openstack-cinder11:04
*** aix has joined #openstack-cinder11:04
*** ebalduf has joined #openstack-cinder11:05
*** ronenkat_ has quit IRC11:07
*** ronenkat__ has quit IRC11:09
*** ebalduf has quit IRC11:10
*** tellesnobrega_ has quit IRC11:10
*** e0ne has joined #openstack-cinder11:12
*** e0ne is now known as e0ne_11:37
*** markvoelker has joined #openstack-cinder11:40
*** coolsvap_ is now known as coolsvap11:46
*** markvoelker has quit IRC11:47
*** e0ne_ is now known as e0ne11:55
DuncanTcinderella: I think if you change your helper config and uninstall iscsitarget, then tgt will work fine. Never actually tried though11:59
*** e0ne has quit IRC12:00
*** ho has quit IRC12:00
cinderellaDuncanT: yes, that's fine. just wondering if there exists some scenario where I need both running - being constrained to turn off one of the two may be limitating12:01
*** coolsvap is now known as coolsvap_12:01
DuncanTI don't know of any such scenario - I don't think you can have both running at once easily, since they bind to the same port12:01
*** EmilienM|afk is now known as EmilienM12:02
cinderellaDuncanT: you're right, sorry for my noobness :)12:03
DuncanTcinderella: I'd rather deal with somebody like you, who tries things, can work google and comes back with feedback, than a lazy expect any day :-)12:04
*** ebalduf has joined #openstack-cinder12:06
*** e0ne has joined #openstack-cinder12:09
*** ebalduf has quit IRC12:10
*** avishay has joined #openstack-cinder12:12
*** Mandell has joined #openstack-cinder12:12
*** Mandell has quit IRC12:17
*** xyang1 has quit IRC12:25
*** cinderella is now known as cinderella_away12:28
*** cinderella_away has quit IRC12:31
*** chlong has quit IRC12:38
*** oomichi has quit IRC12:39
*** Lee1092 has quit IRC12:40
openstackgerritEdwin Wang proposed openstack/cinder: FlashSystem Code Cleanup
*** IanGovett has joined #openstack-cinder12:42
*** oomichi has joined #openstack-cinder12:42
*** markvoelker has joined #openstack-cinder12:43
openstackgerritEdwin Wang proposed openstack/cinder: FlashSystem Code Cleanup
*** eharney has joined #openstack-cinder12:44
*** oomichi has quit IRC12:44
*** markvoelker has quit IRC12:47
*** marcusvrn1 has joined #openstack-cinder12:47
*** marcusvrn1 has quit IRC12:48
*** marcusvrn1 has joined #openstack-cinder12:48
*** primechuck has quit IRC12:49
*** marcusvrn has quit IRC12:51
*** avishay has quit IRC12:54
*** avishay_ has joined #openstack-cinder12:54
openstackgerritwuyuting proposed openstack/python-cinderclient: Make cinderclient support snapshot-force-delete
*** markvoelker has joined #openstack-cinder13:01
openstackgerritChangBo Guo(gcb) proposed openstack/cinder: Add extral library oslo.concurrency to oslo.config.generator.rc
*** oomichi has joined #openstack-cinder13:04
*** afazekas has quit IRC13:06
*** ebalduf has joined #openstack-cinder13:07
*** oomichi has quit IRC13:08
*** akerr has joined #openstack-cinder13:09
*** ebalduf has quit IRC13:11
*** bswartz has quit IRC13:20
*** bill_az_ has joined #openstack-cinder13:23
*** sgotliv has quit IRC13:30
*** eharney has quit IRC13:30
*** sgotliv has joined #openstack-cinder13:35
*** bkopilov has quit IRC13:37
*** kromanenko has joined #openstack-cinder13:39
*** ndipanov has quit IRC13:40
*** mwichmann has joined #openstack-cinder13:51
*** afazekas has joined #openstack-cinder13:55
*** tbarron has joined #openstack-cinder14:03
*** Yogi1 has joined #openstack-cinder14:04
*** kaufer has joined #openstack-cinder14:04
*** andymaier has joined #openstack-cinder14:07
*** ebalduf has joined #openstack-cinder14:07
*** Lee1092 has joined #openstack-cinder14:10
*** ebalduf has quit IRC14:12
*** e0ne is now known as e0ne_14:14
*** nshaikh has quit IRC14:14
*** jungleboyj has quit IRC14:16
*** dustins has joined #openstack-cinder14:17
*** Yogi1 has quit IRC14:18
*** primechuck has joined #openstack-cinder14:19
openstackgerritShay Halsband proposed openstack/cinder: XtreamIO version 4.0 support
*** bswartz has joined #openstack-cinder14:23
*** e0ne_ has quit IRC14:24
*** boris-42 has joined #openstack-cinder14:25
*** mriedem has joined #openstack-cinder14:25
*** mtanino has joined #openstack-cinder14:26
*** e0ne has joined #openstack-cinder14:26
openstackgerritChangBo Guo(gcb) proposed openstack/cinder: Add extra library oslo.concurrency to oslo.config.generator.rc
*** ndipanov has joined #openstack-cinder14:36
*** rushil has joined #openstack-cinder14:40
*** mtanino has quit IRC14:40
*** diegows has joined #openstack-cinder14:44
*** abehl has quit IRC14:45
*** abehl has joined #openstack-cinder14:48
*** jcru has joined #openstack-cinder14:50
*** juzuluag has joined #openstack-cinder14:54
*** xyang has joined #openstack-cinder14:55
*** xyang1 has joined #openstack-cinder14:56
*** bswartz1 has joined #openstack-cinder14:59
*** bswartz has quit IRC14:59
*** bswartz1 is now known as bswartz15:00
*** markstur_ has joined #openstack-cinder15:00
*** markstur_ has left #openstack-cinder15:01
*** jecarey has joined #openstack-cinder15:01
*** dalgaaf has joined #openstack-cinder15:02
openstackgerritSean McGinnis proposed openstack/cinder: Update eqlx driver help text
*** abehl has quit IRC15:05
*** Yogi1 has joined #openstack-cinder15:05
*** ebalduf has joined #openstack-cinder15:08
openstackgerritEdwin Wang proposed openstack/cinder: FlashSystem Code Cleanup
*** Yogi11 has joined #openstack-cinder15:13
*** ebalduf has quit IRC15:13
*** Yogi1 has quit IRC15:14
*** dustins has quit IRC15:15
TobiasEHi, is there a way get the official cinder API  documentation from older releases?15:16
*** e0ne is now known as e0ne_15:17
*** akerr is now known as akerr_away15:17
*** tshefi has quit IRC15:17
*** e0ne_ is now known as e0ne15:17
openstackgerritEdwin Wang proposed openstack/cinder: FlashSystem Code Cleanup
*** jungleboyj has joined #openstack-cinder15:21
*** david-lyle_afk is now known as david-lyle15:23
*** nlevinki has quit IRC15:31
smcginnisTobiasE: This has v1 and v2 -
smcginnisTobiasE: Though not specific to its state for a given release.15:35
*** ebalduf has joined #openstack-cinder15:35
*** stefan_amann has quit IRC15:35
*** amann has quit IRC15:35
*** mtanino has joined #openstack-cinder15:37
*** dustins has joined #openstack-cinder15:41
*** jaypipes has joined #openstack-cinder15:43
*** theanalyst has quit IRC15:43
TobiasEsmcginnis: Thank you, I see15:45
*** theanalyst has joined #openstack-cinder15:47
*** annashen has joined #openstack-cinder15:52
*** akerr_away is now known as akerr15:52
*** Mandell has joined #openstack-cinder15:52
*** tsekiyama has joined #openstack-cinder15:53
*** annashen has quit IRC15:56
openstackgerritMikhail Khodos proposed openstack/cinder: Fix Nexenta NFS driver mounts
*** Ilja has joined #openstack-cinder16:00
*** rajinir_r has quit IRC16:00
*** rajinir_r has joined #openstack-cinder16:01
*** ronis has quit IRC16:01
*** abhijeetm has quit IRC16:04
*** rushil has quit IRC16:08
*** rmesta has joined #openstack-cinder16:10
*** vishy has quit IRC16:11
*** dustins has quit IRC16:12
*** dustins has joined #openstack-cinder16:12
*** vishy has joined #openstack-cinder16:13
openstackgerritSteven Kaufer proposed openstack/cinder: GET volumes API sorting REST/volume/DB updates
openstackgerritSteven Kaufer proposed openstack/cinder: GET volumes API sorting enhancements common utilities
openstackgerritMichal Dulko proposed openstack/python-cinderclient: Add set_management_url to cinderclient.client
*** ofek-baby has joined #openstack-cinder16:20
*** Mandell has quit IRC16:21
*** rushiagr is now known as rushiagr_away16:21
*** lpetrut has quit IRC16:21
*** kallebe has joined #openstack-cinder16:23
*** nellysmitt has quit IRC16:27
jdgxyang1: I'm almost finished with it16:27
xyang1jdg: Great. Thanks!16:28
*** rwsu-afk is now known as rwsu16:28
*** dannywilson has joined #openstack-cinder16:29
jdgthingee: hey... this is a good idea:
*** ndipanov has quit IRC16:32
*** tbarron has quit IRC16:34
*** tbarron has joined #openstack-cinder16:35
*** bkopilov has joined #openstack-cinder16:36
*** dulek has quit IRC16:37
*** lcurtis has joined #openstack-cinder16:40
*** kromanenko has quit IRC16:44
*** sgotliv has quit IRC16:46
*** vilobhmm has joined #openstack-cinder16:48
*** rushiagr_away is now known as rushiagr16:49
*** sgotliv has joined #openstack-cinder16:49
*** e0ne has quit IRC16:51
*** e0ne has joined #openstack-cinder16:51
*** vilobhmm has quit IRC16:53
*** obutenko_ has quit IRC16:54
*** nkrinner has quit IRC16:59
*** lcurtis has quit IRC16:59
*** lcurtis has joined #openstack-cinder17:01
*** alecv has quit IRC17:01
e0nejdg: are you talking about replacement for
*** Ilja has quit IRC17:03
*** lpetrut has joined #openstack-cinder17:05
*** khuss has joined #openstack-cinder17:07
*** kaufer has quit IRC17:08
khusshello.. i'm installing openstack with cinder and trying to figure out how to use a separate storage network.17:08
khussI want to provide a high bandwith data network for storage using cinder. Is it possible to do?17:09
*** karimb has quit IRC17:09
avishay_jdg: better idea - integration test for each feature, and auto-generate the support matrix based on what is passing :)17:09
avishay_khuss: storage network for I/O I assume?17:09
*** mriedem has quit IRC17:09
khussavishay_: yes. storage network for all block device access17:10
khussavishay_: how does cinder determine which interface to use for serving blocks to computes? Is it something configurable?17:11
*** sgotliv has quit IRC17:11
avishay_khuss: Cinder is not on the data path - only provides management capabilities17:11
guitarza1khuss: do you mean with the built in lvm driver?17:11
avishay_khuss: which protocol does your storage use?17:11
khussavishay_: who would be providing the datapath?17:11
jdge0ne: indeed17:11
jdge0ne: people won't let me kill that17:11
khussavishay_:  it uses lvm i believe. I'm installing cinder using ubuntu cinder charm17:12
jdge0ne: so at least put it under some sort of control17:12
avishay_khuss: the servers running Nova will connect via iSCSI directly to the iSCSI target created by the Cinder driver17:12
khussguitarza1: may be it is a question of configuing the lvm driver17:12
guitarza1khuss: the volume driver has an iscsi_ip_address config option17:13
*** xyang has quit IRC17:13
avishay_khuss: so as long as the target is set up using the proper interface, you should be OK...not sure how to configure that with cinder17:13
e0nejdg: afair, some time ago DuncanT has idea to generate it from code after implementactio ABC for drivers17:13
avishay_guitarza1: ah, you're good17:13
openstackgerritxing-yang proposed openstack/cinder: Fix detach volume from host problem in VMAX driver
khussguitarza1: where is this configuration to be? Is it in the cinder configuration?17:14
*** aix has quit IRC17:14
*** Ilja has joined #openstack-cinder17:15
*** guitarza1 is now known as guitarzan17:15
*** avishay_ is now known as avishay17:15
jdge0ne: yeah, he talked about that the other week.  I think that would be a great thing17:15
guitarzanavishay_: but I'm not sure it's used like I'd expect17:15
khussguitarzan: i've several machinbe running cinder volumes17:15
khussguitarzan: and i want to use a second interface for the data path17:16
thingeejdg: I agree with having the matrix in tree for review on changes. Maybe not the dev docs though since we want deployers to see it.17:16
jdgkhuss: #iscsi_ip_address=$my_ip17:16
jdgfor lvm17:16
guitarzankhuss: I don't want to send you down the wrong path, so I'm trying to see if it does what I'd expect17:17
guitarzanjdg would know though17:17
jdgguitarzan: know what?17:17
khussjdg: let me check if there is such parameter in the config file17:17
* jdg is 1/2 paying attention17:17
jdgkhuss: there is :)17:17
guitarzanjdg: if that conf flag is for the connection info17:17
jdgguitarzan: the iscsi_ip_address?17:17
guitarzanjdg: yeah17:17
jdgThat's the one you want for LVM17:18
khussjdg: let me check your link17:18
jdgthingee: yeah, that's a good point17:18
guitarzanjdg: ah, I see, the provider_location vs discovery stuff now17:18
*** Ilja has quit IRC17:18
jdgguitarzan: yup17:18
guitarzanwoo, I guessed right17:18
thingeejdg: I'll see the source though. Thanks! :)17:18
jdgit's appache licensed, not stealing ;)17:19
*** kaufer has joined #openstack-cinder17:19
openstackgerritxing-yang proposed openstack/cinder: EMC VMAX driver Kilo update
khussjdg: in my /etc/cinder/cinder.conf file, that entry doesn't exist. may be I can create one17:20
*** andymaier has quit IRC17:20
guitarzankhuss: you can certainly set one17:20
jdgkhuss: right, yours just has the default "my_ip"17:20
jdgkhuss: add it and set it to what you need and you should be good to go17:20
jdgkhuss: don't forget to restart :)17:21
jdgkhuss: note that devstack and some other packages don't list out all the defauult settings17:21
khussjdg: restart scsi process?17:21
jdgkhuss: they only show the items that are modified17:21
jdgkhuss: no, restart the volume-driver17:21
jdgkhuss: conf changes aren't dynamic17:21
jdgkhuss: they require service restart17:22
jdgkhuss: which service depends on which conf option :)17:22
jdgkhuss: this particular entry is c-volume service17:22
khussjdg i see a cinder-volume process17:22
khussjdg:  may be that is the one I need to restart?17:22
jdgkhuss: that's the one17:22
*** jpich has quit IRC17:23
*** Yogi11 has quit IRC17:24
khussjdg: just to clarify.. Let's have I've 6 cinder machines. On each, machine I set the iscsi_ip to 10.0.1.x  .. now the nova_compute should be able to reach the same subnet ..17:24
khussjdg: so the nova-compute should be in the same subnet as well.. am I making sense?17:24
jdgkhuss: umm... we don't handle your networking config for you17:24
jdgkhuss: it's up to you to make sure you have routes and everything17:24
khussjdg: sorry.. that's not what I meant. essentailly, the iscsi_ip that we set should be reachable from nova-compute.. right?17:25
jdgkhuss: correct17:25
khussjdg:  in essence, I can put that in a separate storage network17:25
jdgkhuss: that's the ip address that the iscsi target is going to be setup on17:25
*** ronis has joined #openstack-cinder17:26
*** sgotliv has joined #openstack-cinder17:26
khussjdg: thanks guys.. it really helps a lot17:26
*** Yogi1 has joined #openstack-cinder17:27
kallebeHey, does python-cinderclient have an exclusive irc channel or do you all gather here?17:27
openstackgerritDanny Al-Gaaf proposed openstack/cinder: fix typo in
kmartinkallebe, it's here17:29
*** rushil has joined #openstack-cinder17:31
*** jordanP has quit IRC17:33
*** kaufer has quit IRC17:33
*** lcurtis has quit IRC17:35
*** lcurtis has joined #openstack-cinder17:35
*** leeantho has joined #openstack-cinder17:37
hemna_TobiasE, ping17:37
*** annashen has joined #openstack-cinder17:38
*** e0ne is now known as e0ne_17:39
kallebekmartin ok, thanks17:40
*** e0ne_ is now known as e0ne17:41
*** lpetrut has quit IRC17:44
openstackgerritMitsuhiro Tanino proposed openstack/cinder: Refactoring for export functions in Target object
*** upenr has joined #openstack-cinder17:45
*** xyang has joined #openstack-cinder17:45
upenrUnable to attach cinder volume to nova instance... can anyone help?17:45
upenrHere's the question::
jdgupenr: add answer ask17:49
jdgupenr: so the problem you have is your wwpn's aren't getting set up in the provider info17:49
jdgupenr: you need to doulbe check your cinder conf settings17:49
avishayupenr: are you working with iSCSI or FC?17:50
avishayi suspect i know the issue...17:50
*** Yogi11 has joined #openstack-cinder17:56
*** Yogi1 has quit IRC17:59
*** dustins_ has joined #openstack-cinder18:00
*** bswartz has quit IRC18:01
hemna_upenr, so that error happens when the storewize driver doesn't see any initiator iqns or FC wwpns.18:01
hemna_upenr, which, shouldn't happen18:01
hemna_so if this is for FibreChannel, this is because the compute host doesn't have any Online HBAs  (run: sudo systool -c fc_host -v)18:02
upenrhemna thanks... can u please explain how to addFC wwpns on OpenStack18:02
hemna_upenr, you need to make sure that 1) you have the FC HBA kernel driver loaded and 2) systool installed18:02
*** dustins has quit IRC18:03
upenrI get an error:  sudo systool -c fc_host -v - Error opening class fc_host18:03
hemna_upenr, no FC kernel driver is loaded.18:03
upenrah I suspected this but didnt know how to check... or install..18:03
avishayyep, that's the problem i suspected :)18:04
hemna_upenr, what FC card do you have ?18:04
upenrweird because I specify the name of the backend storage driver in my cinder conf file...18:05
hemna_if it's Emulex, try the lpfc driver18:05
upenroh no wait... this KVM is running on ESXi and my Qlogic card has been zoned to an FC switch which is attached to block storage18:05
upenrESXi is attached toa volume on this storage, on which KVM (VM) is running OpenStack18:06
hemna_I'm ESXi challenged, so I can't help you with that one.18:07
upenrif i dont have this installed, how am I able to create an empty volume on the storage?18:08
hemna_creating a volume has almost nothing to do with attaching18:08
upenrhmm, pl. explain to a noob18:09
*** jistr has quit IRC18:09
*** jistr has joined #openstack-cinder18:10
hemna_most backends/arrays don't need the connector (initiator/wwpns) at creation time, only at attach time)18:11
*** patrickeast has joined #openstack-cinder18:11
upenrIf u get my setup, i am just looking at an overview of what to do..... conceptually.. ESXi has been zoned and rhel-kvm is running on it. How do I make this attach to a block storage?18:11
hemna_jaypipes, ping18:11
*** Mandell has joined #openstack-cinder18:13
*** kallebe has left #openstack-cinder18:15
upenrAll, I am confused... Does this mean it's a KVM issue or a FC card issue?    sudo systool -c fc_host -v         Error opening class fc_host18:16
hemna_you don't have a kernel FC driver loaded.18:16
hemna_on the host18:16
hemna_it's not a KVM issue18:16
nikesh_vedams_jdg : morning :)18:17
upenru mean hypervisor by host?18:17
hemna_that's guest18:17
upenrhost = server?18:17
nikesh_vedams_jdg : good to see you online after long time :)18:18
upenrso it is the FC card driver,?18:18
hemna_hemna_> you don't have a kernel FC driver loaded.18:19
upenrto be installed where? I still don't get it18:19
*** e0ne has quit IRC18:21
upenrhemna, how to install the FC kernel driver and where is it installed?18:22
upenr* hemna_18:22
*** jistr has quit IRC18:22
*** annegent_ has joined #openstack-cinder18:26
upenrhemna_ so it should be installed on Linux-KVM? Is that correct?18:28
*** nellysmitt has joined #openstack-cinder18:28
hemna_The host OS needs the kernel driver loaded.18:29
hemna_but if the host OS is ESXi, I can't really help.18:29
*** harlowja has joined #openstack-cinder18:30
jaypipeshemna_: pong18:32
hemna_jaypipes, hey, so it looks like Thierry just pushed multi-attach bp out to L.  :(18:32
*** nellysmitt has quit IRC18:33
hemna_jaypipes, can we get an exception, or are we done ?18:33
upenrI understand thanks... however, for my previous question that if it has to be installed on ESXi (the hypervisor), you had said NO and that it shd be installed on the physical server.. that's confusing to a new person18:33
*** kaufer has joined #openstack-cinder18:33
hemna_upenr, whoever/whatever is serving up the VM's needs to have the FC kernel driver module loaded, or they will never see FC volumes.18:34
gary-smithupenr: in fairness, that's an ESXi-specific question, which I beileve that hemna already said he was not an expert in18:34
upenrok thanks hemna_ : u see my nova instance is running on KVM, so i'll take it that I shd think of installing the FC kernel driver module on the RHEL65-KVM machine I have...18:35
upenrIf I miss it,  I'll look around if sth needs to be done on ESXi - i apologize for the inconvenience18:36
jdgnikesh_vedams_: hehe18:36
hemna_upenr, no worries.  good luck18:36
*** rushiagr is now known as rushiagr_away18:36
jaypipeshemna_: let me chat with a couple cores about it.18:37
hemna_jaypipes, ok.   I'm trying to track down some tempest failures for the cinder side.18:37
hemna_anyone running devstack in a vm?   I can't seem to get the tempest encrypted volume tests to work18:38
hemna_off of master18:38
openstackgerritPatrick East proposed openstack/cinder: Add CHAP support to PureISCSIDriver
openstackgerritPatrick East proposed openstack/cinder: Add DB table for driver specific data
*** fandi has joined #openstack-cinder18:40
patrickeastjdg: hey, let me know if you have some time to chat about the driver-not-meta-data table… now that we are getting into k-3 i want to try and move forward with it18:40
*** afazekas has quit IRC18:40
smcginnisupenr: You also understand a virtualized KVM machine running on top of ESX will not see any FC adapters unless you configure the VM to pass the PCI device through to the guest?18:40
jdgpatrickeast: you be around in about 15 minutes?18:40
patrickeastjdg: yep18:40
nikesh_vedams_jdg : i got this message on patch 13     Jenkins has posted comments on this change.  Change subject: Add iSCSI SCST Target support to cinder ......................................................................   Patch Set 13: Verified-1  Merge Failed.  This change was unable to be automatically merged with the current state of the repository. Please rebase your change and upload a new patchset.18:41
smcginnisnikesh_vedams_: You need to rebase.18:41
patrickeastthingee: hey, wrt to your latest comment on you mean a spec for just the new db table (chap secrets maybe being a use case) and not for adding chap support for the driver, right?18:42
*** mriedem has joined #openstack-cinder18:43
thingeepatrickeast: correct18:43
openstackgerritPatrick East proposed openstack/cinder-specs: Add DB table for driver private data
patrickeastthingee: kk, thanks!18:44
upenrsmcginnis I see your point now, thanks... Any guides that show how to achieve this would be helpful...18:45
upenrVMDirectPath I/O passthrough ?18:47
hemna_smcginnis, so for libvirt VM's the FC HBA is never seen passed through to the guest OS.  The FC HBA needs to be there in the host OS and the volume is discovered in the host, and passed into the guest.18:47
hemna_ESXi might be different.  dunno18:47
thingeeanybody noticing issues with devstack with offline=true?18:47
thingee in that example is just empty. Guess I should double check if that's before running after setting offline=true18:48
hemna_thingee, I'm not seeing that currently.   I'm having other problems though against master, trying to get the tempest test_encrypted_cinder_volumes working.   it's borked18:50
thingeepatrickeast: you mind adding interested parties to that spec review. hemna_ was one of them18:52
smcginnishemna_: But that's just using FC based storage with libvirt as the compute, right?18:52
hemna_smcginnis, attaching FC cinder volumes to VM's via libvirt18:53
patrickeastthingee: yea, will do18:53
smcginnishemna_: If I understood right he is trying to run compute within a guest on top of ESX.18:53
smcginnisAnd get FC.18:53
hemna_smcginnis, goosh18:53
smcginnisMaybe I misunderstood though.18:53
thingeehemna_: is that actually a problem upstream with FC, without the separate out brick work you're doing?18:53
smcginnisAnother layer of inception. :)18:53
hemna_smcginnis, well the same is kinda true then, but the guest OS in this case, being the compute host, needs to get the PCI device via the hypervisor (ESXi in this case?)18:54
*** lpetrut has joined #openstack-cinder18:54
smcginnishemna_: Yep, that's what I was trying to point out. Maybe not clearly. :)18:54
hemna_thingee, this failure is happening with lvm18:54
hemna_trying a second vagrant devstack instance18:54
upenrsmcginnis u r right, i am running compute on a VM on top of ESX18:55
hemna_upenr, ok in that case you need to tell ESXi to give that VM the FC HBA device18:55
thingeehemna_: ok, but it's not consistent right? We'd have a report by now of it failing in gate.18:56
thingeethat's what I'm trying to understand18:56
smcginnisThis might help:
upenrhowever, that HBA is being used by ESXi to attach a SAN storage for itself (on which the compute is running)18:56
hemna_thingee,  it happens 100% of the time for me.  so I'm confused18:56
smcginnisupenr: Not going to work then.18:56
hemna_yah, just happened again on a clean vm18:56
thingeeok, is this with the separate out brick work you're doing, or not18:57
smcginnisupenr: You'll need to get an HBA dedicated to pass through.18:57
hemna_thingee, nothing to do w/ brick18:57
hemna_thingee, this is master18:57
smcginnisupenr: I don't even think multiport HBAs will work to break out ports.18:57
smcginnisupenr: It might allow you to configure it, but then I think it errors out.18:57
patrickeastthingee: maybe dumb question, which section should interested parties go in? the “Other contributers”? i did a quick search through the other specs and didn’t see a good example18:57
smcginnisupenr: So you need a full card to pass through to the guest in order for KVM to see a FC HBA.18:58
*** ronenkat has joined #openstack-cinder18:58
upenri have 2 FC ports, the 2nd one is not attached to an FC switch, the first port is zoned with storage already18:58
smcginnisupenr: Yep, not going to work.18:58
hemna_upenr, you still need to get "sudo systool -c fc_host -v" working in your compute host OS18:58
hemna_upenr, if that doesn't work, FC attaches will never work.18:58
smcginnisupenr: It can't be shared by host (ESX) and guest (KVM)18:58
upenryes sir, but i don't know how to install that module either18:58
smcginnisupenr, hemna_: But that's the second step.18:59
*** xyang has quit IRC18:59
*** xyang has joined #openstack-cinder18:59
smcginnisupenr: First you need an HBA getting passed through to the guest.18:59
hemna_upenr, it's a linux kernel driver module.  you'll have to find the right driver for your HBA and install it.18:59
openstackgerritxing-yang proposed openstack/cinder: Fix detach volume from host problem in VMAX driver
smcginnisupenr: Then your second problem is you need to make sure the kernel module is loaded and it recognizes it.18:59
hemna_smcginnis, yup18:59
upenryes i see18:59
smcginnisupenr: Time to get out the credit card. :]19:00
upenror wipe out ESXi and start with KVM?19:00
smcginnisupenr: Or you're stuck using iSCSI if you have to run this on top of ESX.19:00
smcginnisupenr: That would probably be better.19:00
upenrah, iSCSI may not be a bad ide19:00
smcginnisupenr: Assuming ESXi is not needed for other reasons of course.19:01
smcginnisupenr: That does work well.19:01
smcginnissmcginnis: Definitely simpler in this scenario.19:01
upenror, I can run this compute on a different ESXi and configure it to passthrough!19:01
openstackgerritDanny Wilson proposed openstack/cinder: Enabling volume replication on PureISCSIDriver
smcginnissimple == goodness19:01
upenri dont know the basics of KVM to even dare to start with KVM (instead of esxi) which is what scares me19:03
smcginnisupenr: It might be better in the long run to take VMware out of the picture if you can.19:03
upenrok, thanks19:03
smcginnisupenr: You can't get away from dealing with KVM.19:03
ronenkatjungleboyj: hi19:04
upenrso i'd install the OS controller on top of virt-manager?19:05
*** dustins_ has quit IRC19:05
smcginnisHow were you doing it within the virtualized instance. Same thing, you're just not virtualized.19:06
smcginnisupenr: If this isn't production I would recommend just installing Ubuntu or Red Hat and runing devstack.19:06
*** avishay has quit IRC19:07
jdgpatrickeast: yo19:07
smcginnisupenr: Otherwise, follow installation instructions or use deployment tool of your choice.19:07
patrickeastjdg: hey19:07
*** tbarron has quit IRC19:07
jdgpatrickeast: pulling the patch back up now19:07
upenrit isn't production, bt i need to test the Ibm packaging of OpenStack19:07
smcginnisupenr: The IBM cinder drivers, or IBM actually has their own packaged version of OpenStack?19:08
upenryes they do19:08
upenrthey also have cinder drivers19:08
jungleboyjsmcginnis: You are right.  I think taking VMWare out of the mix is the best solution.19:08
*** Yogi1 has joined #openstack-cinder19:09
jdgpatrickeast: yeah, so in Cinder host includes driver19:09
jungleboyjsmcginnis: We have a deployer, etc for installing.19:09
smcginnisupenr: If it's just the drivers then you need to just configure to use that driver.19:09
smcginnisjungleboyj: Cool, that should help.19:09
upenri have done it19:09
gary-smithhemna_, thingee: I'm seeing the same error as hemna on tempest test_encrypted_cinder_volumes19:09
gary-smith(when I run in a vm)19:09
patrickeastjdg: but if you have two cinder hosts using the same driver they would be different, right? (i might be totally missing something here)19:09
jungleboyjsmcginnis: Yes, but I have never set up Cinder running on a VM under VMWare.19:10
openstackgerritxing-yang proposed openstack/cinder: EMC VMAX driver Kilo update
hemna_cd /opt/stack/tempest && tools/ tempest.scenario.test_encrypted_cinder_volumes19:10
hemna_fails every time19:10
smcginnisjungleboyj: Yeah, it can be done, but if you don't have to I wouldn't do it.19:10
hemna_LVM as the cinder driver19:10
jungleboyjsmcginnis: +219:10
jdgpatrickeast: so cinder-volume will show on <machine-name>@d<driver-name>19:10
jdgpatrickeast: that's what I was getting at19:10
*** Yogi11 has quit IRC19:10
patrickeastjdg: gotcha19:10
patrickeastjdg: so my thinking is if i have two hosts devstack1@pure-driver-1 and devstack2@pure-driver-2 both pointing to the same backend i would want these keys to be shared between both of them, not specific to one or the other19:12
patrickeastthe scope is for the driver class, not really the instance19:12
jdgpatrickeast: well now you're just trying to make it hard :)19:12
hemna_patrickeast, +119:12
jdgpatrickeast: You could still do that but you have to push the coordination back down into the drivers19:12
jdgseems like it should be doable19:12
*** dustins has joined #openstack-cinder19:13
patrickeastjdg: how can a driver know about other cinder hosts though?19:13
jdgpatrickeast: well, a driver "knows" what host it is19:13
hemna_patrickeast, drivers aren't19:13
patrickeastjdg: i guess we could maybe add an api that lets a driver view other hosts and their driver class type19:14
*** thangp has joined #openstack-cinder19:14
jdgpatrickeast: gimmie a second to check something19:14
*** vilobhmm has joined #openstack-cinder19:14
jdgI think you already have this info19:14
hemna_so that encypted volume test fails with LVM, but works against 3PAR19:14
jdgit's just embedded19:15
smcginnisjungleboyj: Is that the IBM Workload Deployer you were talking about?19:15
jdghemna_: it works in gate though19:15
hemna_yah :(19:15
hemna_trying to figure it out19:15
jdghemna_: means it's most likely your conf settings19:15
jdgpatrickeast: so the volume object already has the host in it19:16
*** lcurtis has quit IRC19:17
hemna_removed any custom settings in tempest.conf reset cinder.conf to lvm, bounced cinder and retesting19:17
jdgpatrickeast: the info is available there... so it's just a matter of using a common algo in the driver to make them match19:17
smcginnisthingee: Channel header still says Review-A-Thon Happening now.19:17
*** dustins has quit IRC19:17
*** lcurtis has joined #openstack-cinder19:17
xyanghemna_: hi, question for you19:17
*** mfink_ has joined #openstack-cinder19:18
hemna_xyang, shoot19:18
xyanghemna_: in the common driver class, we have 2.0 as default, FC and iSCSI driver override it with 2.1 now.  Do you think the common should stay as 2.0 as default19:18
thingeegary-smith, jdg, hemna_: this what I was trying to convey as well. Fires would be raised from infra otherwise.19:18
patrickeastjdg: hmm if we go that route we might run into the same issue with associating this info with volumes that blocks me from just using provider_auth in the first place, we would be depending on there being volumes in the db with the host available to look up the values19:18
thingeeand they got me on speed dial19:19
jdgthingee: :)19:19
thingeealthough jdg fixes them before I see them :)19:19
* jdg has been slacking past couple weeks19:19
thingeeor at least some helpful debug output to pin the issue down19:19
hemna_xyang, if the drivers override it, it doesn't really matter I guess.19:19
hemna_xyang, we just hard code it to '1.0' in our common, as our 2 drivers overwrite it19:19
jdgthingee: but yes, the gate is my litmus test19:20
jdgas it should be IMO19:20
xyanghemna_: so currently the common also changes it to 2.1.  do you want me to change it back to 2.0, or doesn't matter19:20
jdgpatrickeast: so I'm not saying you put it in the volume though19:20
smcginnisDid something change in common?19:20
jdgI'm just saying that that's at least one way to extract the info in the driver19:21
hemna_xyang, doesn't matter as long as the drivers overwrite it.  if anything set a VERSION in common and use that in common19:21
smcginnisI would think it should bump if something changes.19:21
*** sgotliv has quit IRC19:21
smcginnisOh, nevermind. I see.19:21
xyanghemna_: ok, thanks19:21
jdglemme see if it's innit as well... I think it is19:21
xyangsmcginnis: that's specific to our driver:)19:22
patrickeastjdg: maybe i misunderstood, wouldn’t we still need to look up the host on a volume in the database to enable us to find all the key-value pairs available for a driver?19:22
*** tbarron has joined #openstack-cinder19:22
smcginnisxyang: Yeah, thanks. I was confusing myself. Again.19:22
ronenkatjungleboyj: hello19:22
hemna_damn, same failure19:22
jungleboyjronenkat: Howdy sir.19:23
jdgpatrickeast: well in your driver you are only getting volumes that are for your host19:23
jdgpatrickeast: and that host info is included in self on driver init19:23
jdgpatrickeast: so what I'm saying is that you have your proposed table associated with each host19:23
ronenkatjungleboyj: you asked about how to test replication, did you find the answer?19:23
jdgpatrickeast: then when one needs to get info they do "not-meta-keys.for-host"19:23
jungleboyjronenkat: Not yet.19:23
jungleboyjMoved on to looking at the API section.19:24
jdgpatrickeast: see what I'm getting at?19:24
jdgpatrickeast: then you don't have the massive unstructured glob of keys19:24
ronenkatjungleboyj: its well hidden as part of create volume, there is a --source-replica option for the create API19:24
jdgpatrickeast: you still have globs, but they're at least associated with the specific driver that cares about them19:24
patrickeastjdg: yea i understand, but i’m not sure that solves the issue of two hosts i mentioned before, or maybe that scenario isn’t worth worrying about?19:25
jdgpatrickeast: IMHO there's no reason Pure's driver should know or care about HP's driver keys19:25
patrickeastjdg: i agree 100% with that19:25
jdgpatrickeast: so the two hosts thing I don't know if that's really an issue19:25
patrickeastjdg: but i should care about other pure keys from a different host19:25
jdgpatrickeast: unless your device can have "only one" value19:25
jungleboyjronenkat: Oh, that is what the --source-replica is for.19:25
jdgpatrickeast: in which case write a generation algorithm that makes it common19:25
ronenkatjungleboyj: if you specify --source-replica then the created volume would be of the secondary by running the "create_replica_test_volume" function19:25
jdgpatrickeast: even if it's duplicate, who cares19:26
patrickeastjdg: yea thats the problem, we need a single one that is randomly generated :*(19:26
jdgpatrickeast: so you end up right back where we started19:26
*** nellysmitt has joined #openstack-cinder19:26
jdgpatrickeast: so random is a "relative" term :)19:26
thingeepatrickeast: there's no crying in cinder19:26
patrickeastjdg: hehe19:26
jdgpatrickeast: use a seed an a prng to duplicate it19:26
patrickeastthingee: lol19:26
patrickeastthingee: only ascii tears19:26
ronenkatjungleboyj: yep, seems a better option than to duplicate the create volume code - it proved to be a minimal change in the create volume flow19:26
jdgpatrickeast: I've done similar things in the past; it's fairly trivial19:27
jdgpatrickeast: so take something from the device like it's UUID or FW version and use that as a seed19:27
*** ChanServ sets mode: +o thingee19:27
jdgpatrickeast: generate your "random" value that way19:27
patrickeastjdg: i want to say that is fine… but i have specific requirements to use pycrypto to generate them19:27
patrickeastjdg: i’m not sure if i can trick that into duplicating random values19:27
jdgpatrickeast: and voila... consistency no matter how many instances of it you have19:27
jdgpatrickeast: well now you're own your own :)19:28
jungleboyjronenkat: Ok, cool.  Will update that accrodingly in the spec.19:28
jdgpatrickeast: if you're using true crypto requirements you need to get the ability to read it from your device19:28
ronenkatjungleboyj: thanks19:28
jdgbecause you're going to have a hell of a time19:28
jdgalso it's silly to be strict crypto policies then dump the info into an unencrypted DB19:28
patrickeastjdg: well the whole thing is a bit of a sham anyway, i know...19:29
jungleboyjronenkat: For the REST API impact ... I need to remove the relationship section.19:29
patrickeastjdg: ok, so… do you see any way forward to make everyone happy with this?19:30
jdgpatrickeast: so actually I *believe* you can still do this with pycrypto even19:30
ronenkatjungleboyj: right, as Cinder is unware of the actual replication relationship, the API moved into the volume API - didn't make sense to create a new one19:30
jdgpatrickeast: *everyone* LOL19:30
jdggood luck with that!19:31
jungleboyjronenkat: Ok.  As I looked at the code again I saw that.19:31
jdgpatrickeast: honestly though I think my suggestion is pretty legit19:31
jdgand I don't see too many reasons for objections19:31
* patrickeast feels dirty intentionally gaming a random generator19:31
patrickeasti don’t want to be that guy the security audit finds and points out19:32
jdgpatrickeast: well, that's what prng's are for :)19:32
jdgpatrickeast: well then you need to make it solveable in the product19:32
jdgpatrickeast: and besides, security audit will point to the db anyway :)19:32
jdgpatrickeast: at least it would if I were doing the audit19:32
jdgpatrickeast: not to mention all the creds in the conf file, unencrypted as well19:33
jungleboyjronenkat: So we have just the promote and reenable that go in the REST API section.19:33
jdgpatrickeast: so maybe that's your answer to all of this...19:33
jdgpatrickeast: user randomly generates it... puts it in conf as an encyrpted key, your driver knows how to decrypt it19:33
jdgpatrickeast: skip all the silly tables/storing etc19:33
jdgpatrickeast: actually I like that idea even more :)19:34
ronenkatjungleboyj: yes, the create replica is just a create (enabled by extra spec)19:34
*** bswartz has joined #openstack-cinder19:34
*** mudassirlatif has joined #openstack-cinder19:34
ronenkatjungleboyj: but there is also the addition of --source-replica in the volume create API19:34
patrickeastjdg: yea but then we have to keep those files sync’d between all of the cinder instances19:34
*** asmith_ has joined #openstack-cinder19:34
jdgpatrickeast: yeah... well seurity is hard!19:35
patrickeastjdg: and we are back to my first chap support proposal that was shot down for not working in active/active19:35
patrickeastjdg: but, it doesnt have to19:35
jdgwell... that wasn't quite the same19:35
jungleboyjronenkat: Ok.19:35
patrickeastjdg: similar problem, different way the credentials were generated19:35
jdgpatrickeast: well, what I mean by "it's hard" is it requires some effort19:35
jungleboyjronenkat: Do you want me to take a stab at reworking that section or do you want me to push up the other changes I have now and you can fix it up?19:35
patrickeastjdg: i guess i’m not sure i understand why we *have* to make the table key to cinder host names19:36
patrickeastjdg: i see the benefit of scoping19:36
patrickeastjdg: but there isn’t any alternative?19:36
jdgpatrickeast: well, maybe I'm the only one who thinks that sucks19:36
openstackgerritTom Swanson proposed openstack/cinder: Expanded our unit test coverage:  This tests our api module. We've also added some minor driver fixes found by unit and other testing. Some logging statements have been updated. All changes are in the dell storage center driver and unit
jdgpatrickeast: the beauty of us not all agreeing on things is that maybe somebody else will say "griffith is stupid" and +2/A your patch :)19:37
ronenkatjungleboyj: if you can rework the API, it would be great - but I can do that as well19:37
*** Yogi1 has quit IRC19:38
patrickeastjdg: unfortunately right now you’ve been the strongest supporter of this… so id rather not take a chance on that ; )19:38
jungleboyjronenkat: I will take a stab and you can comment.19:38
jdgpatrickeast: well there's a switch :)19:38
jdgpatrickeast: I'm usually the grumpy ne-sayer19:38
jdgpatrickeast: so really I think it's a good thing to have, I just don't like the sharing19:39
jdgpatrickeast: I get the challenge here19:39
patrickeastjdg: so how would you feel about scoping the key based on driver class name?19:39
ronenkatjungleboyj: Thanks, time to quit, my day is at its ends...19:39
jdgpatrickeast: but not sure how to solve it, other than "Don't do multi-backend with the same pure device"19:39
jdgpatrickeast: which honestly seems perfectly reasonable19:40
jungleboyjronenkat: Sounds good.  Talk to you tomorrow.19:40
jdgpatrickeast: you can even make it configurable... "hey, you can use this for multiples, but you take a security hit because I'm using an algorithm to generate duplicate sec info"19:40
jdgpatrickeast: win/win19:40
patrickeastjdg: yea, i suppose thats not the worst outcome, we would require a lock in the driver when accessing this right now anyway19:40
patrickeastjdg: and i don’t see any good distributed lock solution for openstack right now19:41
jdgpatrickeast: honestly that's the route I'd probably go19:41
*** dulek has joined #openstack-cinder19:41
jdgpatrickeast: I believe barbican might solve that at some point19:41
patrickeastjdg: yea thats true19:41
jdgpatrickeast: if it doesn't have a foundation to do so already19:41
jdgpatrickeast: but I'm certainly not well versed on barbican19:42
jdgpatrickeast: hmmm... ya know19:42
jdgthat might actually be your answer19:42
patrickeastjdg: barbican?19:42
jdgwe have the interface (or parts of it)19:42
jdgpatrickeast: yeah19:42
jdgpatrickeast: so I believe that would allow an admin to upload a random key/value19:43
jdgpatrickeast: and you could get it from there19:43
jdgnot sure19:43
jdgpatrickeast: you should ping folks in #openstack-barbican19:43
jdgor on openstack-dev19:43
jdgor mail list19:43
patrickeastjdg: hmm is it something mature enough to jump in and use?19:44
jdgor I can shoot you Brianna's email address19:44
jdgpatrickeast: mature!!  Ha, surely you gest :)19:44
jdgThis is OpenStack19:44
*** r-daneel has joined #openstack-cinder19:44
jdgMature is a relative term19:44
patrickeastjdg: heh yea19:44
patrickeastjdg: relative to say.. cinder19:44
jdgseriously, that's at least something that I think is worth investigating19:45
jdgpatrickeast: I wouldn't spend weeks on it or anything but at least worth poking at for an hour or so19:45
patrickeastjdg: alright, i’ll take a look through the docs and any code we have in cinder already, i put a spec up for this db i’ll update it later for the hosts + key key19:46
jdgpatrickeast: be careful, I'm going to go back on first proposal and just make it a barbican problem for you to solve :)19:46
jdgpatrickeast: emphasis on *you* :)19:46
patrickeastjdg: yea, if it doesn’t look like barbican will solve this easily i’m going to push for a db table still19:47
patrickeastjdg: thanks again for your help on this19:47
jdgpatrickeast: you're very welcome, thanks for the patience19:48
*** scottda_ has joined #openstack-cinder19:56
*** thangp has quit IRC19:57
*** jnrao has joined #openstack-cinder19:57
*** thingee changes topic to "The OpenStack Block Storage Project Cinder | | The New Kids On The Block"20:00
*** ChanServ sets mode: -o thingee20:00
*** scottda_ has quit IRC20:00
*** dalgaaf has quit IRC20:00
thingeesmcginnis: fixed, thanks20:00
*** Mandell has quit IRC20:01
*** dustins has joined #openstack-cinder20:01
thingeenext summit we need to get a picture of the whole team.20:01
*** Yogi1 has joined #openstack-cinder20:02
*** Lee1092 has quit IRC20:10
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
hemna_thingee, lol20:13
hemna_nice edit20:14
thingeethe best you can do with gimp20:15
*** e0ne has joined #openstack-cinder20:17
*** e0ne is now known as e0ne_20:17
gary-smiththat new cinder core developer in the lower left looks an awful lot like hemna's wife20:18
openstackgerritOpenStack Proposal Bot proposed openstack/python-cinderclient: Updated from global requirements
*** e0ne_ is now known as e0ne20:19
thingeegary-smith: I consider significant others part of the team. They have to hear us rant at the end of the day.20:20
jdgthingee: LMAO20:20
jdgnice photo shop20:20
jdgor whoever put it in there20:20
jdgthat's funny!20:20
thingeejdg: you mean nice gimp20:20
hemna_thingee,  :)20:20
e0newhat did i miss?20:20
jdghaha... yeah, GIMP LOL20:20
thingeejdg: well this picture is missing a lot of people, but I felt weird posting this picture and you not being in it.20:21
thingeee0ne: see channel topic20:21
thingeeanyways, new picture at the summit.20:22
*** ronenkat has quit IRC20:22
jdgthingee: who you callin a *gimp*!20:24
openstackgerritTom Swanson proposed openstack/cinder: Dell Storage Center Unit Test Updates for Kilo
*** abehl has joined #openstack-cinder20:40
xyangthingee: nice job editing the picture :)20:40
jungleboyjthingee: I love it!20:43
*** Mandell has joined #openstack-cinder20:43
*** kaisers has joined #openstack-cinder20:44
*** Mandell has quit IRC20:44
jungleboyjWe will have to try to get a picture of the whole team in Vancouver.  Maybe at the 'Real Cinder Experts' meeting?20:45
*** IanGovett has quit IRC20:47
*** IanGovett has joined #openstack-cinder20:48
e0nejungleboyj: does it mean that you'll buy tickets for all of us? :)20:50
jungleboyje0ne: Hmmmm, you should be able to get there as Core now.  Right?  :-)20:50
e0nejungleboyj: looks like you said 'no'20:51
e0nei hope, i'll be at Summit20:51
jungleboyj:-)  Well, if I we win the Lottery, you are covered e0ne20:51
jungleboyjCome on PowerBall!  jungleboyj needs some spending money20:54
*** diegows has quit IRC20:55
*** annegent_ has quit IRC20:55
*** Mandell has joined #openstack-cinder20:56
*** xyang has quit IRC21:01
*** xyang has joined #openstack-cinder21:01
*** annegent_ has joined #openstack-cinder21:02
*** _cjones_ has joined #openstack-cinder21:05
*** Mandell has quit IRC21:10
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Wait for 60 seconds NoValidHost exception in sheduler
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Wait for 60 seconds on NoValidHost exception in sheduler
*** _cjones_ has quit IRC21:11
*** _cjones_ has joined #openstack-cinder21:12
e0nesorry for such workaround:(21:12
*** _cjones_ has quit IRC21:13
*** _cjones_ has joined #openstack-cinder21:13
*** Mandell has joined #openstack-cinder21:14
*** juzuluag has quit IRC21:15
*** _cjones_ has quit IRC21:18
*** abehl has quit IRC21:19
patrickeastjdg: hey, so after poking around barbican the good news is that it looks like it would work to store credentials and fixes the whole plain text secret in a database thing, the bad news is that we would need to keep track of the key uuid’s somewhere, or query barbican for all secrets, parse through them all, and search for one with a name we know21:20
patrickeastjdg: so we either still need a db table or we do a bunch of rest api calls every time we initalize a connection, not sure which is the lesser evil21:21
patrickeastjdg: i was thinking more about your idea of having it more like a model update, which lead me back to thinking maybe it should be a more specific model for initialize_connection that gets passed in and updated from the return value of the driver21:22
patrickeastjdg: limits the functionality quite a bit, but in a way thats nice as it solves the concerns about adding just a general meta data table21:23
*** jungleboyj has quit IRC21:23
patrickeastjdg: and from my point of view still meets all my needs ; )21:23
*** _cjones_ has joined #openstack-cinder21:23
jdgpatrickeast: oh, that all sounds promising21:24
hemna_I still can't get the tempest.scenario.test_encrypted_cinder_volumes to pass  (master)21:24
*** scottda_ has joined #openstack-cinder21:25
*** scottda_ has quit IRC21:25
jdgpatrickeast: I would def decline the "bunch'o rest calls"21:26
jdgpatrickeast: but using a key in barbican and store a reference to it by name in Cinder does sound like the most appropriate way to go here21:26
*** lpetrut has quit IRC21:26
jdgpatrickeast: I'm not sure how the table looks or what you have in mind21:27
patrickeastjdg: yea from a security standpoint it makes the most sense21:27
jdgpatrickeast: but you've had enough of my opinions on this to know what I like and don't like :)21:27
patrickeastjdg: haha21:27
patrickeastjdg: i’ll update the spec this afternoon with some of the alternative approaches, maybe we can get some other feedback, my concern with the more limited approach is that it may not work for the other folks who mentioned they were interested in using this table21:28
patrickeastjdg: i want to avoid creating something just for pure… if i can21:28
jdgpatrickeast: that's the spirit! :)21:29
jdgpatrickeast: the only thing I'd say is that other folks that expressed some interest have solved the problem in other ways21:29
jdgso may not really be an issue21:29
patrickeastjdg: ah yea, good point21:30
jdgand chap creds seems to be a common stumbling block for people21:30
jdgpatrickeast: doing this would be nice because then we could push the existing entries out of the conf file21:30
jdgsince they shouldn't have ever been there in the first place IMHO21:30
patrickeastjdg: yea i always like the idea of a centralized solution so other drivers can work the same way… tends to help reduce bugs for everyone21:31
patrickeastjdg: also secrets in conf files is never a good thing21:31
jdgjdg: True statement21:31
jdgdefeats the whole purpose21:32
*** EduardBiceri_afk has quit IRC21:33
*** Yogi11 has joined #openstack-cinder21:39
*** ronis has quit IRC21:39
*** krtaylor has quit IRC21:40
*** casusbelli has quit IRC21:41
*** Yogi1 has quit IRC21:42
openstackgerritnikeshmahalka proposed openstack/cinder: Add iSCSI SCST Target support to cinder
*** dhellmann_ has joined #openstack-cinder21:51
*** sgotliv has joined #openstack-cinder21:52
*** krtaylor has joined #openstack-cinder21:53
*** Yogi11 has quit IRC21:54
*** nellysmitt has quit IRC21:54
*** dhellmann_ has quit IRC21:55
*** bswartz has quit IRC21:56
*** dhellmann_ has joined #openstack-cinder21:57
*** dhellmann_ has quit IRC21:58
*** e0ne has quit IRC22:00
*** dannywil_ has joined #openstack-cinder22:01
*** asmith__ has joined #openstack-cinder22:02
*** Yogi1 has joined #openstack-cinder22:02
*** patrickeast_ has joined #openstack-cinder22:03
*** patrickeast has quit IRC22:03
*** patrickeast_ is now known as patrickeast22:03
*** dhellmann_ has joined #openstack-cinder22:03
*** asmith_ has quit IRC22:04
*** dannywilson has quit IRC22:04
*** jaypipes has quit IRC22:04
*** mudassirlatif_ has joined #openstack-cinder22:05
*** jnrao has quit IRC22:05
*** vilobhmm has quit IRC22:05
*** mudassirlatif has quit IRC22:06
*** mudassirlatif_ is now known as mudassirlatif22:06
*** dhellmann_ has quit IRC22:08
*** Mandell has quit IRC22:08
*** khuss has quit IRC22:09
nikesh_vedams_rebased patch 1322:11
nikesh_vedams_and uploaded new patch22:12
nikesh_vedams_it was showing conflicts22:12
nikesh_vedams_merge conflicts22:12
nikesh_vedams_good link to understand how to reslove merge conflicts for new guys
*** Yogi1 has quit IRC22:13
*** Mandell has joined #openstack-cinder22:17
*** mudassirlatif has quit IRC22:17
*** upenr has quit IRC22:18
*** sgotliv has quit IRC22:19
*** sgotliv has joined #openstack-cinder22:19
*** dhellmann_ has joined #openstack-cinder22:20
*** kaufer has quit IRC22:21
jdgnikesh_vedams_: cool22:28
jdgnikesh_vedams_: By the way... you can also do "git review -d <review-number>"22:28
jdginstead of the checkout22:29
jdgnikesh_vedams_: then you can just fix things up and then do "git commit --amend -a" and "git review" to push it back up22:29
*** mriedem is now known as mriedem_zz22:34
openstackgerritThang Pham proposed openstack/cinder: Cinder objects base
*** jcru has quit IRC22:36
openstackgerritThang Pham proposed openstack/cinder: Snapshot object
gugl2Hi anyone know why the implemented blueprint got removed from blueprint list for kilo22:37
gugl2the old implemented in other release could be found...but for kilo the implmeneted are all gone22:38
gugl2thingee, hi ...the implemented blueprints in kilo are no where to find22:40
thingeethose are the implemented.22:40
nikesh_vedams_jdg: what to give in review number :)22:40
thingeethese are in progress
*** dannywil_ has quit IRC22:41
thingeegugl2: I talked about on the list about cutting the milestone 2
gugl2thingee, ok...let me take a look..thanks22:42
*** dannywilson has joined #openstack-cinder22:43
thingeethe list/validate extra spec work has a dependency. Unfortunately it's waiting on me
thingeegugl2: I'm sure hemna_ communicated this with you.22:43
*** tbarron has quit IRC22:45
gugl2thingee, I know the milestone list...but I am trying to find blueprint Anthony Lee added in kilo222:45
thingeewhat's the title of it?22:46
*** Mandell has quit IRC22:47
*** dannywilson has quit IRC22:47
*** dannywil_ has joined #openstack-cinder22:47
thingeeit's in
nikesh_vedams_jdg : by the way fingers crossed for positive review  comments this time since we removed db and sooner will try to upload new logs :)22:49
thingeexyang: let me know when we have an updated spec on over subscription. I need to understand the changes to communicate this to driver maintainers.22:50
gugl2thingee, is not under Anythony Lee name thanks. My 3par driver change need to add partial-implements: blueprint filtering-weighing-with-driver-supplied-functions22:50
xyangthingee: ok, I'll work on it tonight now that it is merged22:51
thingeegugl2: that's a separate blueprint22:51
thingeegugl2: I want to track it for the core impact in Cinder, then for drivers.22:51
thingeexyang: thanks22:51
nikesh_vedams_jdg: going to bed , bye :)22:52
gugl2thingee, that is part of implementation of that blueprint...I added a bug for adding the 3par driver piece, but was told I should not add a bug for that22:52
*** jungleboyj has joined #openstack-cinder22:52
thingeegugl2: right. it's not a bug that a driver doesn't implement an optional feature.22:52
gugl2thingee, I need my own blueprint for 3par driver?22:53
thingeegugl2: yes22:53
hemna_yup, I mentioned a BP should be filed for it22:53
thingeegugl2: no spec, just a blueprint22:53
gugl2thingee, good...only blueprint is easier...sure do. Thanks22:54
thingeegugl2: it's just for my tracking purposes and to annoy people. ;)22:54
gugl2thingee, hehe.22:54
gugl2hemna_, thank you too.22:55
*** annegent_ has quit IRC22:55
*** annashen has quit IRC22:59
*** annashen has joined #openstack-cinder22:59
*** primechuck has quit IRC23:01
*** annashen has quit IRC23:03
*** Mandell has joined #openstack-cinder23:05
anishthingee: quick question. do you still want the temp dir only created once even if it's being automatically cleaned up23:05
anishmy thinking was recreating it every time ensure every test gets a clean slate23:06
thingeeyou only use two out of the number of test cases in there.23:06
thingeeonly use it*23:06
*** dulek has quit IRC23:07
anishI can go either way, it is not affecting the test cases. just wanted to check beforehand23:07
thingeethe point I was making in the review was if it's just used in two test cases, just do it in those test cases instead of in setUp23:09
anishhrm, I already made the code changes but I moved it to __init__, so only one dir is ever made and cleanup is being handled automatically. is that acceptable ?23:10
thingeeI'll get on that review sir. Just catching up on emails23:11
*** sgotliv has quit IRC23:11
hemna_thingee, so, jogo (nova) gave a +1 to the idea of using brick in nova23:12
hemna_in the reply to the ML post I made23:12
thingeehemna_: I saw23:12
hemna_we need a new name for it23:12
hemna_as brick exists in pypi already23:12
*** rushil has quit IRC23:12
thingeehemna_: I think it's too late for nova though. They're already saying no to things in k-2 low priority23:12
hemna_thingee, totally.   the nova side is for L23:13
hemna_so I was just following up on what you had requested in the meetup to get feedback from nova so it makes sense to go forward or not.23:14
thingeehemna_: was just looking for ideas in name format23:15
hemna_os-cinder-brick ?23:16
hemna_or just os-brick23:16
*** annashen has joined #openstack-cinder23:16
thingeemy vote is just for os-brick. it's standalone from cinder. We'll of course still maintain it.23:17
anishhemna_: mortar :)23:17
thingeemaybe do a poll.23:18
hemna_ok, so the import will be os_brick23:18
thingeeto just a small group here23:18
hemna_I'll update my repo and then re-up the stackforge post to be openstack instead23:18
*** takedakn has joined #openstack-cinder23:22
*** chlong has joined #openstack-cinder23:23
*** takedakn has quit IRC23:23
*** takedakn has joined #openstack-cinder23:24
jdganish: +1 for mortar!23:24
*** alexpilotti has quit IRC23:25
*** rushil has joined #openstack-cinder23:27
*** dustins has quit IRC23:30
openstackgerritTomoki Sekiyama proposed openstack/cinder: Enhance iSCSI multipath support
*** rushil has quit IRC23:32
*** EmilienM is now known as EmilienM|afk23:33
*** annegent_ has joined #openstack-cinder23:35
*** jecarey has quit IRC23:35
*** akerr has quit IRC23:39
*** annegent_ has quit IRC23:41
*** bswartz has joined #openstack-cinder23:41
*** david-lyle is now known as david-lyle_afk23:41
openstackgerritGloria Gu proposed openstack/cinder: HP 3par driver filter and evaluator function
*** asmith__ has quit IRC23:49
*** mfink_ has quit IRC23:55
*** nellysmitt has joined #openstack-cinder23:55
*** IanGovett has quit IRC23:56
*** smoriya has joined #openstack-cinder23:56
*** openstack has joined #openstack-cinder23:57
hemna_whoa 0 bytes23:57
*** dannywil_ has quit IRC23:58
thingeeyeah so packer creates a vm with nova, installs devstack, set offline=true, creates an image...all that works fine.23:58
thingeebut when I create a vm off that image that has devstack installed with offline=true, it gives that error. digging into tempest_lib things are set to 0 bytes23:59

Generated by 2.14.0 by Marius Gedminas - find it at!