Wednesday, 2015-04-01

*** harlowja has quit IRC00:02
*** harlowja has joined #openstack-cinder00:03
*** vilobhmm11 has joined #openstack-cinder00:16
*** emagana has quit IRC00:19
*** sseverson has quit IRC00:19
*** vilobhmm1 has quit IRC00:19
*** annashen has quit IRC00:19
*** dflorea has joined #openstack-cinder00:30
*** cpg is now known as cpg|away00:32
*** vilobhmm11 has quit IRC00:32
openstackgerritLi Yingjun proposed openstack/cinder: Add missing qos_specs_id to volume type show
*** rushil has quit IRC00:48
*** dflorea has quit IRC00:49
*** dflorea has joined #openstack-cinder00:50
*** mwichmann has quit IRC00:51
*** _cjones_ has quit IRC00:54
*** dflorea has quit IRC00:54
*** leeantho has quit IRC01:00
*** vilobhmm1 has joined #openstack-cinder01:00
*** MentalRay has quit IRC01:01
*** MentalRay_ has joined #openstack-cinder01:01
*** Lee1092 has joined #openstack-cinder01:05
*** akerr has joined #openstack-cinder01:10
*** xyang1 has quit IRC01:16
*** annegentle has joined #openstack-cinder01:18
*** annashen has joined #openstack-cinder01:20
*** haigang has joined #openstack-cinder01:22
*** diegows has quit IRC01:25
*** annashen has quit IRC01:26
*** MentalRay_ has quit IRC01:26
*** casusbel_ has joined #openstack-cinder01:28
*** haomaiwang has quit IRC01:29
*** kaisers1 has joined #openstack-cinder01:31
*** casusbelli has quit IRC01:31
*** kaisers has quit IRC01:33
*** Apoorva has quit IRC01:36
*** ebalduf has joined #openstack-cinder01:39
*** julim has quit IRC01:39
*** annegentle has quit IRC01:43
*** ebalduf has quit IRC01:43
*** haigang has quit IRC01:48
*** haigang has joined #openstack-cinder01:58
*** haomaiwang has joined #openstack-cinder02:00
*** patrickeast has quit IRC02:09
*** vilobhmm1 has quit IRC02:10
*** mtanino has quit IRC02:10
*** annegentle has joined #openstack-cinder02:19
*** annegentle has quit IRC02:24
*** akerr is now known as akerr_away02:25
*** harlowja is now known as harlowja_away02:29
*** jdurgin1 has quit IRC02:35
*** _cjones_ has joined #openstack-cinder02:36
*** ebalduf has joined #openstack-cinder02:40
*** ebalduf has quit IRC02:44
*** akerr_away is now known as akerr02:47
*** akerr has quit IRC02:47
*** _cjones_ has quit IRC02:52
*** _cjones_ has joined #openstack-cinder02:53
*** dims_ has quit IRC02:55
*** zongliang has quit IRC02:56
*** annegentle has joined #openstack-cinder02:57
*** MentalRay has joined #openstack-cinder03:01
*** Longgeek_ has joined #openstack-cinder03:03
*** rhefner has quit IRC03:04
*** Longgeek has quit IRC03:05
*** david-lyle has joined #openstack-cinder03:13
*** vilobhmm1 has joined #openstack-cinder03:16
*** annashen has joined #openstack-cinder03:23
*** vilobhmm1 has quit IRC03:26
*** annashen has quit IRC03:27
*** changbl_ has joined #openstack-cinder03:36
*** changbl has quit IRC03:38
*** _cjones_ has quit IRC03:47
*** thingee has quit IRC03:51
*** rushiagr_away is now known as rushiagr03:54
*** avishay has joined #openstack-cinder03:57
*** heyun has joined #openstack-cinder03:59
*** MentalRay has quit IRC04:05
*** MentalRay has joined #openstack-cinder04:05
*** emagana has joined #openstack-cinder04:07
*** lpetrut has joined #openstack-cinder04:13
*** shilpa_ has joined #openstack-cinder04:14
*** rushiagr is now known as rushiagr_away04:22
*** annegentle has quit IRC04:26
*** dflorea has joined #openstack-cinder04:36
*** haigang has quit IRC04:38
*** Mandell has joined #openstack-cinder04:41
*** haigang has joined #openstack-cinder04:44
*** eharney has quit IRC04:47
*** krtaylor has quit IRC04:48
*** eharney has joined #openstack-cinder04:49
*** haigang has quit IRC04:50
*** krtaylor has joined #openstack-cinder04:50
*** deepakcs has joined #openstack-cinder04:50
*** BharatK has joined #openstack-cinder04:58
*** emagana has quit IRC05:04
*** emagana has joined #openstack-cinder05:05
*** emagana has quit IRC05:10
*** dflorea has quit IRC05:13
*** dflorea has joined #openstack-cinder05:14
*** rushiagr_away is now known as rushiagr05:14
*** shilpa_ has quit IRC05:15
*** dflorea has quit IRC05:19
*** rushiagr is now known as rushiagr_away05:23
*** reed has quit IRC05:29
*** lpetrut has quit IRC05:29
*** haigang has joined #openstack-cinder05:35
*** Longgeek_ has quit IRC05:38
*** Longgeek has joined #openstack-cinder05:39
*** tshefi has joined #openstack-cinder05:39
*** vilobhmm1 has joined #openstack-cinder05:40
*** haigang has joined #openstack-cinder05:45
*** Maike has joined #openstack-cinder05:48
*** cpg|away is now known as cpg05:49
*** Maike_ has joined #openstack-cinder05:50
*** Maike has quit IRC05:53
*** spradeepv has joined #openstack-cinder06:03
*** pradipta has joined #openstack-cinder06:05
*** haigang has quit IRC06:07
spradeepvhemnafk: I am from brocade. Can you provide me the logs for this bug -
openstackLaunchpad bug 1438722 in Cinder "Brocade Zone Manager does not wait for zoning transactions to complete" [Medium,Triaged] - Assigned to Angela Smith (aallen-m)06:08
*** haigang has joined #openstack-cinder06:09
*** haigang has quit IRC06:10
*** nlevinki has joined #openstack-cinder06:12
*** haigang has joined #openstack-cinder06:13
*** spradeepv is now known as spradeepv_06:20
*** afazekas has joined #openstack-cinder06:21
*** spradeepv_ is now known as spradeepv06:23
*** haigang has quit IRC06:23
*** Liu has joined #openstack-cinder06:23
*** annashen has joined #openstack-cinder06:25
*** haigang has joined #openstack-cinder06:25
*** ankit_ag has joined #openstack-cinder06:25
*** annashen has quit IRC06:29
*** abehl has joined #openstack-cinder06:29
*** ronis has joined #openstack-cinder06:33
openstackgerritgit-harry proposed openstack/python-cinderclient: Fix bypass_url option, required by session object
*** shilpa_ has joined #openstack-cinder06:40
*** ishant has joined #openstack-cinder06:41
openstackgerritTushar Patil proposed openstack/python-cinderclient: Allow cinderclient to handle exception response
*** Raj_ has joined #openstack-cinder06:43
*** MentalRay has quit IRC06:44
*** dulek has joined #openstack-cinder06:45
*** vilobhmm1 has quit IRC06:45
*** TobiasE has joined #openstack-cinder06:48
openstackgerritYAMADA Hideki proposed openstack/cinder: Fix missing clone_image API support for sheepdog driver.
openstackgerritgit-harry proposed openstack/python-cinderclient: Allow manually specified auth token
openstackgerritJulien Danjou proposed openstack/cinder: tests: replace mox by mox3, clean out mox usage
openstackgerritJulien Danjou proposed openstack/cinder: Leverage timeutils, drop strtime() usage
*** markvoelker has quit IRC07:09
*** jordanP has joined #openstack-cinder07:09
*** geguileo has joined #openstack-cinder07:13
*** jistr has joined #openstack-cinder07:14
*** heyun has quit IRC07:19
*** casusbel_ has quit IRC07:25
*** casusbelli has joined #openstack-cinder07:26
*** rushiagr_away is now known as rushiagr07:30
*** sgotliv has quit IRC07:34
*** sgotliv has joined #openstack-cinder07:34
*** rushiagr is now known as rushiagr_away07:37
*** markus_z has joined #openstack-cinder07:44
*** markus_z has quit IRC07:47
*** abhijeetm has joined #openstack-cinder07:48
*** dims_ has joined #openstack-cinder07:50
*** dims_ has quit IRC07:55
openstackgerritVincent Hou proposed openstack/cinder: Set volume_attachment to [] for the temporary volume creation
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
*** ndipanov has joined #openstack-cinder08:07
*** Longgeek has quit IRC08:13
*** Longgeek has joined #openstack-cinder08:13
*** e0ne has joined #openstack-cinder08:16
*** lpetrut has joined #openstack-cinder08:18
openstackgerritVincent Hou proposed openstack/cinder: Delete the temporary volume if migration fails
openstackgerritVincent Hou proposed openstack/cinder: Delete the temporary volume if migration fails
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
*** e0ne is now known as e0ne_08:29
*** kashyap has joined #openstack-cinder08:30
*** alecv_ has joined #openstack-cinder08:40
*** ho has quit IRC08:43
*** e0ne_ is now known as e0ne08:43
*** yuriy_n17 has joined #openstack-cinder08:55
*** deepakcs has quit IRC08:58
*** Longgeek has quit IRC08:59
*** Miouge has quit IRC09:00
*** rushiagr_away is now known as rushiagr09:01
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
anishheh, gerrit has clippy today09:09
*** kashyap has left #openstack-cinder09:13
*** markus_z has joined #openstack-cinder09:15
*** Longgeek has joined #openstack-cinder09:27
*** annashen has joined #openstack-cinder09:27
*** ociuhandu has quit IRC09:28
*** annashen has quit IRC09:33
*** lpetrut has quit IRC09:39
*** Miouge has joined #openstack-cinder09:41
*** Raj_ has quit IRC09:51
*** haomaiwang has quit IRC09:51
*** heyun has joined #openstack-cinder09:54
*** dims_ has joined #openstack-cinder10:00
*** dims__ has joined #openstack-cinder10:01
*** alexpilotti has joined #openstack-cinder10:03
*** dims_ has quit IRC10:04
*** Miouge has quit IRC10:07
openstackgerritRick Chen proposed openstack/cinder: Add patch for consistency group update in ProphetStor driver
dulekDuncanT: hi!10:14
openstackgerritJeegn Chen proposed openstack/cinder: More error handling on VNX migration failure.
dulekDuncanT: Working on bug 1438744 I've run into a problem.10:15
openstackbug 1438744 in Cinder "AttributeError from _update_allocated_capacity upon create failure" [Undecided,Confirmed] - Assigned to Michal Dulko (michal-dulko-f)10:15
dulekDuncanT: Bug is there because in VolumeManager I'm catching CinderException10:15
dulekDuncanT: And comes out the from the create_volume flow also ProcessUtilException can be raised.10:16
dulekDuncanT: This is raised from the driver directly.10:16
dulekDuncanT: I wonder if I should repack the exception in create_volume flow or just catch also ProcessUtilException in the VolumeManager.10:17
dulekDuncanT: I also wonder if there aren't other exception types raised from the drivers.10:17
dulekDuncanT: Any opinion?10:17
DuncanTI suspect there are drivers that raise all sorts of things in corner cases10:18
*** Miouge has joined #openstack-cinder10:19
*** rushiagr is now known as rushiagr_away10:19
dulekDuncanT: So the question is - add try-except to places where create_volume flow is calling the driver and reraise VolumeDriverException10:20
dulekDuncanT: Or just catch Exception in VolumeManager10:20
*** lpetrut has joined #openstack-cinder10:21
DuncanTSorry, I'm working through code as fast as I can here, catching up. Can you point me at the catch block you mean in VolumeManager?10:21
*** rushiagr_away is now known as rushiagr10:22
*** Muntaner has joined #openstack-cinder10:22
Muntanerhi guys10:22
MuntanerI'm trying to add a NAS (buffalo) to my OpenStack all-in-one installation10:22
MuntanerI'd like to use it with iscsi10:22
*** jistr has quit IRC10:23
Muntanerit is in a LAN that I use for services,, and can ping it via the server @
Muntanerany advices?10:23
dulekDuncanT: And here's where exception is raised.10:23
DuncanTYeah, I think catching Exception there is fine10:23
*** haigang has quit IRC10:24
dulekDuncanT: Okay, thank you for your opinion. :)10:25
*** aarefiev has quit IRC10:25
openstackgerritBharat Kumar Kobagana proposed openstack/cinder: Using 'fallocate' instead of 'dd' to create raw volumes
DuncanTDrivers might raise all sorts of exceptions, and all you're doing is cleaning up then reraising10:25
DuncanTNow I've got the code in front of me, I'm intrigued by - we're going back to the db just to get the volume id, which we already have and we know is immutable10:25
DuncanTOh, wait, I see it, ignore me10:26
DuncanTBeing slow10:26
DuncanTThough really we only need to do that in the 'if not rescheduled:' case, right?10:27
DuncanT(It's a minor performance optimisation, but something we need to start looking at to improve out DB scaling. Don't feel you need to care now)10:28
dulekDuncanT: Yes, you're right I think.10:28
dulekI really hope this code is only nasty workaround we'll be able to get rid of in L.10:29
DuncanT:-) We've said that about lots of code10:29
dulekAww, I'm naive. ;)10:30
dulekI'll add that optimization to my TODO list. Thanks for pointing it out.10:30
openstackgerritTina Tang proposed openstack/cinder: Create consistgroup from cgsnapshot support in VNX driver
*** aarefiev has joined #openstack-cinder10:36
*** jistr has joined #openstack-cinder10:36
openstackgerritTina Tang proposed openstack/cinder: Attach/detach batch processing in VNX driver
Muntanernobody? I'm pretty lost right now10:40
*** Raj_ has joined #openstack-cinder10:41
*** IanGovett has joined #openstack-cinder10:41
openstackgerritGorka Eguileor proposed openstack/python-cinderclient: Remove duplicate find request in find_resource
openstackgerritGorka Eguileor proposed openstack/python-cinderclient: Add cache to Managers' find/findall methods
*** ociuhandu has joined #openstack-cinder10:44
*** ishant has quit IRC10:44
*** nellysmitt has joined #openstack-cinder10:50
*** e0ne is now known as e0ne_10:53
*** haigang has joined #openstack-cinder10:54
*** iwao has quit IRC10:56
*** e0ne_ is now known as e0ne10:57
*** heyun has quit IRC10:58
*** haigang has quit IRC11:10
*** jistr is now known as jistr|demo11:11
*** Muntaner has quit IRC11:22
*** rushiagr is now known as rushiagr_away11:28
*** rushiagr_away is now known as rushiagr11:28
openstackgerritMichal Dulko proposed openstack/cinder: Catch more general exception in manager's create_volume
*** annashen has joined #openstack-cinder11:29
*** annashen has quit IRC11:34
openstackgerritBharat Kumar Kobagana proposed openstack/cinder: Using 'fallocate' instead of 'dd' to create raw volumes
*** cpg is now known as cpg|away11:37
*** dulek has quit IRC11:40
*** pradipta has quit IRC11:45
*** lpetrut has quit IRC11:53
*** shilpa_ has quit IRC11:54
abhijeetmHi all, can someome please review patch :
abhijeetmI already got one +211:55
*** BharatK has quit IRC11:58
*** Lee1092 has quit IRC12:00
*** e0ne is now known as e0ne_12:01
openstackgerritZhiQiang Fan proposed openstack/cinder: Switch to oslo.policy 0.3.0
openstackgerritZhiQiang Fan proposed openstack/cinder: Switch to oslo.policy 0.3.0
*** bswartz has quit IRC12:09
*** BharatK has joined #openstack-cinder12:13
*** MentalRay has joined #openstack-cinder12:14
*** markvoelker has joined #openstack-cinder12:14
*** akerr has joined #openstack-cinder12:18
*** jungleboyj has joined #openstack-cinder12:29
*** Miouge has quit IRC12:34
*** dulek has joined #openstack-cinder12:39
*** MentalRay has quit IRC12:41
*** Miouge has joined #openstack-cinder12:42
*** e0ne_ is now known as e0ne12:43
*** diegows has joined #openstack-cinder12:46
*** timcl has joined #openstack-cinder12:48
*** BharatK has quit IRC12:48
*** diegows has quit IRC12:51
*** Yogi1 has joined #openstack-cinder12:53
*** ebalduf has joined #openstack-cinder12:55
*** lpabon has joined #openstack-cinder12:59
*** ebalduf has quit IRC13:00
*** timcl1 has joined #openstack-cinder13:02
*** BharatK has joined #openstack-cinder13:02
*** timcl has quit IRC13:05
*** annegentle has joined #openstack-cinder13:05
avishayabhijeetm: checking13:07
*** julim has joined #openstack-cinder13:07
abhijeetmavishay: ok, thanks13:08
*** mwichmann has joined #openstack-cinder13:09
*** rhefner has joined #openstack-cinder13:16
*** Raj_ has quit IRC13:23
*** mriedem_away is now known as mriedem13:24
avishayabhijeetm: the review site just went down for me :(13:26
avishayabhijeetm: looks OK, except the comment that you added in is very confusing IMO13:26
avishayabhijeetm: i would prefer it were removed13:26
avishayabhijeetm: oh, the site is back...13:27
-openstackstatus- NOTICE: gerrit has been restarted to restore event streaming. any change events missed by zuul (between 12:48 and 13:28 utc) will need to be rechecked or have new approval votes set13:28
*** haigang has joined #openstack-cinder13:31
abhijeetmavishay: While upload is in progress we can delete instance or detach volume from instance13:35
*** alexpilotti has quit IRC13:35
abhijeetmavishay: and in that case volume will remain in 'in-use' state13:35
avishayabhijeetm: i understand, but the code under the comment updates the state atomically, which is the correct thing to do in any case - it is not something specific to that issue13:36
avishayabhijeetm: if someone who isn't familiar with this bug looks at the code and they see the comment, they won't understand13:37
*** timcl1 has quit IRC13:37
abhijeetmavishay: ok, I will remove comment and upload new patch set13:37
avishayabhijeetm: thanks!13:38
*** Yogi1 has quit IRC13:38
*** Liu has quit IRC13:42
*** Liu has joined #openstack-cinder13:42
*** lpetrut has joined #openstack-cinder13:42
*** nlevinki has quit IRC13:44
*** diegows has joined #openstack-cinder13:46
*** abehl has quit IRC13:46
openstackgerritAbhijeet Malawade proposed openstack/cinder: Get volume from db again before updating it's status
*** anshul has joined #openstack-cinder13:46
scottdaIs it just me, or do the pages have an annoying Windows Office paperclip animation on them?13:51
*** anshul has quit IRC13:51
*** haigang has quit IRC13:52
*** timcl has joined #openstack-cinder13:53
avishayscottda: i assume it's for april fools13:56
jungleboyjscottda: I just noticed that too.  I am on a Linux box.  Was very confused for a minute.13:57
jungleboyjMan, he shows up on the review pages too.13:59
*** bswartz has joined #openstack-cinder13:59
*** dustins has joined #openstack-cinder14:00
*** crose has joined #openstack-cinder14:00
*** MentalRay has joined #openstack-cinder14:00
*** Yogi11 has joined #openstack-cinder14:00
*** anshul has joined #openstack-cinder14:02
*** Lee1092 has joined #openstack-cinder14:04
scottdaI'm on linux as well...14:05
avishayapril fools.....14:05
scottdaavishay: I think you are right! Actually, that's quite funny!14:05
*** anshul has quit IRC14:07
*** Maike_ has quit IRC14:08
*** annegentle has quit IRC14:09
*** tshefi has quit IRC14:11
*** ebalduf has joined #openstack-cinder14:11
*** rushiagr is now known as rushiagr_away14:14
*** shilpa_ has joined #openstack-cinder14:15
*** ebalduf has quit IRC14:16
*** thangp has joined #openstack-cinder14:16
*** e0ne is now known as e0ne_14:17
*** mtanino has joined #openstack-cinder14:18
*** Muntaner has joined #openstack-cinder14:22
Muntanerhey guys14:22
MuntanerI know I'm repetitive14:22
MuntanerI need some help in connecting my NAS to my Openstack cinder module :)14:23
*** e0ne_ is now known as e0ne14:23
thangpdid someone put clippy on gerrit reviews as an april fool's joke?14:23
*** pradipta has joined #openstack-cinder14:24
*** Yogi11 has quit IRC14:26
scottdathangp: no, we are all required to use Windows for OpenStack dev now.14:27
jbernardsadly, it doesn't help me navigate the UI any better14:27
thangpbut clippy comes from aws :)14:27
*** bkopilov has quit IRC14:27
thangpat least that's what the html says14:28
*** bkopilov has joined #openstack-cinder14:28
thangpi like the pacman on google maps though14:28
*** annegentle has joined #openstack-cinder14:36
*** alexpilotti has joined #openstack-cinder14:38
avishayMuntaner: do you have a specific problem?  which driver?14:40
*** diegows has quit IRC14:43
openstackgerritAlex Meade proposed openstack/cinder: Revert "Removing Netapp FC drivers for no reported CI"
*** patrickeast has joined #openstack-cinder14:45
*** Yogi1 has joined #openstack-cinder14:52
*** Mandell has quit IRC14:52
*** rmesta has joined #openstack-cinder14:53
*** nlevinki has joined #openstack-cinder14:54
*** tsekiyama has joined #openstack-cinder14:54
Muntaneravishay, I simply dunno how to start the configuration14:56
Muntaneractually I have this NAS, can ping it, etc, but dunno what to do exactly14:56
avishayMuntaner: you need to specify the appropriate configuration options in cinder.conf , each backend has its own14:56
avishayMuntaner: which driver is it?  most have documentation14:57
Muntaneravishay, I'd like to use iscsi14:57
Muntanerand I think this nas is set up with it (it's the first time see it)14:57
Muntanerthe problem is I'm not expert in this14:57
avishayMuntaner: which vendor, model?14:57
Muntanerit is a buffalo terastation14:57
*** emagana has joined #openstack-cinder14:58
Muntaneravishay, do you need the exact model?14:58
*** Bsony has joined #openstack-cinder15:00
*** guitarzan has quit IRC15:00
avishayMuntaner: is that based on Windows Server 2012?15:00
*** guitarzan has joined #openstack-cinder15:00
Muntaneravishay, dunno :(15:01
Muntanerbut it has ssh, so I think there's some linux inside15:01
*** AlkaD has joined #openstack-cinder15:02
*** asselin has joined #openstack-cinder15:02
*** kaisers has joined #openstack-cinder15:02
*** annegent_ has joined #openstack-cinder15:02
avishayMuntaner: I'm not familiar with it...there is a driver for Windows Server 2012 based boxes, and many others, but not for buffalo specifically as far as i know15:02
avishayMuntaner: i would suggest contacting the vendor15:02
*** nellysmi_ has joined #openstack-cinder15:03
*** nellysmitt has quit IRC15:04
*** annegentle has quit IRC15:04
*** nellysmitt has joined #openstack-cinder15:04
*** nellysmitt has quit IRC15:05
Muntaneravishay, the problem is: to say what?15:06
Muntanerhow do I tell cinder to use this NAS? these are my problems15:06
*** kaisers has quit IRC15:06
avishayMuntaner: Ask if the storage is supported by Cinder, and if so, do they have any documentation on configuring Cinder to work with it15:06
avishayMuntaner: If they don't know, ask if it is based on Windows Server 2012, because that has a driver15:07
Muntaneravishay, with NFS or iscsi shouldn't I be able to use it without knowing the vendor or model?15:07
jgriffithNo meeting today?15:07
Muntanerif it runs as a NFS server, for example?15:08
Muntanerjust asking, I'm not expert15:08
*** nellysmi_ has quit IRC15:08
*** kaisers has joined #openstack-cinder15:08
rhe00_jgriffith: 10am mountain time. :)15:08
jgriffithhehe :)15:08
jgriffithDST strikes15:08
*** ankit_ag has quit IRC15:08
rhe00_yes, and my outlook is not handling that well.15:08
avishayjgriffith: yea i checked google for "time now in UTC" :)15:09
Muntaneravishay, I look noobish, I know :)15:10
*** rushiagr_away is now known as rushiagr15:11
*** kaisers has quit IRC15:11
*** rhe00_ is now known as rhe0015:11
*** Yogi1 has quit IRC15:15
*** Muntaner has quit IRC15:16
mtreinishjgriffith: btw, not sure you saw this: but I'm going for #4 :)15:17
*** sks has joined #openstack-cinder15:17
*** abhijeetm has quit IRC15:17
mtreinishprobably worth testing locally before you merge though...15:17
*** kaisers has joined #openstack-cinder15:17
*** shyama has quit IRC15:20
*** kaisers has quit IRC15:22
jgriffithmtreinish: hehe... no I didn't see it.  Looking...15:24
jgriffithmtreinish: oh yay!!15:24
*** thingee has joined #openstack-cinder15:24
jgriffithmtreinish: I'll load it up in my test env then accept the PR15:24
jgriffithmtreinish: the shell.CONF.set_override is a nice touch as well BTW15:25
mtreinishjgriffith: heh, I really need to kill the oslo.db usage in there and drop using oslo.config15:26
jgriffithmtreinish: and this leads in to the next thing I wanted to add...15:26
jgriffithmtreinish: associating some meta with the run entry15:26
jgriffithmtreinish: keep track of patch-id, CI-name etc15:27
mtreinishoh, yeah I had no idea if the metadata storage I added there was useful/correct or not. I took a guess15:27
*** bkopilov has quit IRC15:27
jgriffithmtreinish: yeah, :)15:27
jgriffithmtreinish: although now you stoll my commit :)15:27
jgriffithmtreinish: if I'm not careful you'll be the top contributor LOL15:27
*** hemnafk is now known as hemna15:28
*** TobiasE has quit IRC15:28
jgriffithmtreinish: something else I'm considering depending on uptake...15:28
jgriffithmtreinish: just publishing on PyPi15:28
jgriffithmtreinish: it's already relatively "simple" to get up and running, but that would sort of make it a no brainer15:29
mtreinishso you can pip install sos-ci? Yeah, that would probably be useful for people15:29
mtreinishheh, although I'm probably the wrong person to ask because I've never run it :)15:29
jgriffithbeauty of open source collaboration15:30
*** jdurgin1 has joined #openstack-cinder15:31
*** kaisers has joined #openstack-cinder15:31
*** annashen has joined #openstack-cinder15:33
*** bswartz has quit IRC15:34
*** kaisers has quit IRC15:35
*** reed has joined #openstack-cinder15:35
*** hemna has quit IRC15:36
*** e0ne is now known as e0ne_15:37
*** annashen has quit IRC15:37
*** e0ne_ is now known as e0ne15:38
openstackgerritTom Barron proposed openstack/cinder: Only use operational LIFs for iscsi target details
*** kaisers has joined #openstack-cinder15:40
*** hemna has joined #openstack-cinder15:42
*** lpabon has quit IRC15:44
*** hemna has quit IRC15:45
*** rushil has joined #openstack-cinder15:45
*** bkopilov has joined #openstack-cinder15:46
*** bkopilov has quit IRC15:46
*** bswartz has joined #openstack-cinder15:47
*** _cjones_ has joined #openstack-cinder15:47
*** kaisers has quit IRC15:48
*** xyang1 has joined #openstack-cinder15:49
*** hemna has joined #openstack-cinder15:49
*** shilpa_ has quit IRC15:54
*** dannywilson has joined #openstack-cinder15:56
*** ebalduf has joined #openstack-cinder15:57
*** deepakcs has joined #openstack-cinder15:57
*** rmesta has quit IRC15:58
*** rmesta has joined #openstack-cinder15:58
*** rmesta has quit IRC15:59
*** rmesta has joined #openstack-cinder15:59
e0nelooks like we'll get a short meeting today15:59
*** Yogi1 has joined #openstack-cinder16:00
jgriffithe0ne: don't jynx it :)16:01
*** sks has quit IRC16:01
*** ebalduf has quit IRC16:02
jgriffithe0ne: so short we won't even have one :)16:02
*** HoloIRCUser5 has joined #openstack-cinder16:04
*** timcl1 has joined #openstack-cinder16:05
*** MentalRay has quit IRC16:06
*** timcl has quit IRC16:07
*** Yogi1 has quit IRC16:07
*** vbala has joined #openstack-cinder16:08
*** Yogi1 has joined #openstack-cinder16:08
*** rongze has joined #openstack-cinder16:08
*** vbala has quit IRC16:08
*** jdurgin1 has quit IRC16:09
*** patrickeast has quit IRC16:09
*** avishay has quit IRC16:09
*** alecv has joined #openstack-cinder16:10
*** gary-smith_ has quit IRC16:10
*** gary-smith_ has joined #openstack-cinder16:11
*** dustins has quit IRC16:14
*** HoloIRCUser5 has quit IRC16:14
*** rushil has quit IRC16:15
*** alexpilotti has quit IRC16:16
*** enterprisedc has joined #openstack-cinder16:18
*** dulek has quit IRC16:19
jgriffithmtreinish: boooo
* e0ne bb in ~2hours16:19
*** annashen has joined #openstack-cinder16:20
*** dustins has joined #openstack-cinder16:20
*** e0ne has quit IRC16:20
mtreinishjgriffith: heh, oops16:21
*** annegent_ has quit IRC16:22
*** deepakcs has quit IRC16:23
*** MentalRay has joined #openstack-cinder16:23
*** kaisers has joined #openstack-cinder16:23
*** asselin has quit IRC16:24
*** rhagarty has joined #openstack-cinder16:24
*** bswartz has quit IRC16:24
*** rhagarty_ has joined #openstack-cinder16:25
*** AlkaD has quit IRC16:25
jgriffithboris-42: WTF.. haha, I just saw Clippy on my dashboard16:26
*** kaisers has quit IRC16:27
winston-djgriffith: yeah, saw that too16:27
hemnaman, I saw that earlier in a review, and thought...crap...what extension in chrome did I screw up now.....16:28
openstackgerritAnton Arefiev proposed openstack/cinder: Fix always false condition in glance wrapper
jgriffithboris-42: asked me about it yesterday and I told him I thought he was drunk :)16:28
thingeehemna: LOL16:28
hemnaI was getting pissed.16:29
winston-djgriffith: M$ revealed the secret behind Azure and Office 365, they are all on OpenStack16:29
hemnafriggin clippy did I do?16:29
jgriffithwinston-d: LMAO16:30
jgriffithwinston-d: That's funny!!!!16:30
boris-42jgriffith: LOL16:30
kmartinwinston-d, awesome16:30
*** leeantho has joined #openstack-cinder16:30
thingeehave you all looked at google maps today?16:30
hemnauh oh...no16:31
winston-dhemna: did you try left or right click the clippy? I tried many times...16:31
hemnaI tried right clicking on it, dragging it..16:31
hemnawth, this damn thing......16:31
thingeeI wish clippy would start nitpicking at code16:31
thingeelooks like that comment is missing a space and period16:31
hemnalol yah16:31
thingeelooks like you could use a mock decorator instead of as a nested context managers16:32
*** jogo has left #openstack-cinder16:32
jgriffiththingee: that would be AWESOME!16:33
winston-da gift from M$ office team, 108 yrs of professional nitpicking experience16:34
*** mdbooth has quit IRC16:35
*** jaypipes has quit IRC16:37
*** esker has joined #openstack-cinder16:37
*** shilpa_ has joined #openstack-cinder16:38
*** markus_z has quit IRC16:40
*** mdbooth has joined #openstack-cinder16:42
*** kaisers has joined #openstack-cinder16:42
mtreinishjgriffith: ok, I updated the pr16:44
*** kaisers has quit IRC16:45
*** jistr|demo has quit IRC16:47
*** winston-d is now known as winston-d_zZZ16:48
jgriffithmtreinish: there's an extra blank space in there16:49
jgriffithmtreinish: "cfg.DataBase. database_connection_string"16:49
*** yuriy_n17 has quit IRC16:50
*** jordanP has quit IRC16:51
*** timcl1 has quit IRC16:51
*** dims__ has quit IRC16:53
*** dims_ has joined #openstack-cinder16:53
*** Mandell has joined #openstack-cinder16:54
*** crose has quit IRC16:55
*** Mandell_ has joined #openstack-cinder16:57
*** rongze has quit IRC16:57
*** Yogi11 has joined #openstack-cinder16:58
*** lpetrut has quit IRC16:59
*** timcl has joined #openstack-cinder16:59
*** alexpilotti has joined #openstack-cinder17:00
*** Mandell has quit IRC17:00
*** Yogi1 has quit IRC17:01
*** kaisers has joined #openstack-cinder17:01
*** dustins has quit IRC17:02
*** harlowja_away is now known as harlowja17:02
*** dustins_ has joined #openstack-cinder17:02
*** emagana has quit IRC17:04
*** kaisers has quit IRC17:04
*** _cjones_ has quit IRC17:06
*** e0ne has joined #openstack-cinder17:07
*** dannywilson has quit IRC17:08
*** alecv_ has quit IRC17:11
*** annegentle has joined #openstack-cinder17:12
*** Apoorva has joined #openstack-cinder17:12
*** _cjones_ has joined #openstack-cinder17:13
*** patrickeast has joined #openstack-cinder17:14
*** Mandell_ has quit IRC17:18
mtreinishjgriffith: ugh, seriously17:20
mtreinishone sec I'll respin it right now17:20
*** e0ne is now known as e0ne_17:21
mtreinishjgriffith: ok pushed it again, now I hope it's typo free17:22
*** mwichmann has quit IRC17:22
*** timcl has quit IRC17:22
*** pradipta has quit IRC17:24
*** sks has joined #openstack-cinder17:26
*** shilpa_ has quit IRC17:26
*** e0ne_ has quit IRC17:27
*** timcl has joined #openstack-cinder17:27
jgriffithmtreinish: it pains me to EVER note a typo17:27
mtreinishjgriffith: heh, except when I make them things no work :)17:28
jgriffithmtreinish: I hosed ya.. can you rebase my bug fix on top or that?17:29
jgriffithmtreinish: never midn17:29
jgriffithmtreinish: github figured it out17:29
*** thingee has quit IRC17:29
*** Mandell has joined #openstack-cinder17:30
jgriffithmtreinish: thanks!!!17:30
kmartin or dial-in US: +1 800-451-8679 or +1 212 729 5016, IL: 809462557, IN: 02030291605, UK: 08445790676, CZ: 800-451-8679, AU: 0730870178,  Conf id: 781 929 3200#17:30
*** BharatK has quit IRC17:30
kmartinmy bad, wrong window17:31
mtreinishjgriffith: awesome, I guess we're just used to jgit screwing everything up in gerrit :)17:31
*** emagana has joined #openstack-cinder17:31
jgriffithkmartin: I'm sooo tempted to eavesdrop now :)17:33
*** e0ne has joined #openstack-cinder17:34
*** mwichmann has joined #openstack-cinder17:36
*** annegentle has quit IRC17:36
*** diegows has joined #openstack-cinder17:36
*** thingee has joined #openstack-cinder17:42
*** vilobhmm1 has joined #openstack-cinder17:43
*** vilobhmm1 has quit IRC17:44
*** vilobhmm1 has joined #openstack-cinder17:44
*** vilobhmm1 has quit IRC17:45
*** vilobhmm1 has joined #openstack-cinder17:45
sksI am getting the error: [demo] cinder list17:46
sks ERROR: Unable to establish connection to http://x.y.z.t:9856/v1/461cd6404b54428099d916291c7bf467/volumes/detail17:46
sks[stack@dhcp mytools]$ [demo] exit17:46
*** vilobhmm11 has joined #openstack-cinder17:46
sksDoes anybody know the issue?17:46
*** Yogi11 has quit IRC17:46
*** vilobhmm11 has quit IRC17:47
sksI could able to do nova list successfully17:47
*** vilobhmm11 has joined #openstack-cinder17:47
DuncanTsks: cinder-api service not running? Firewall issues?17:47
sksDuncanT, c-api was running17:48
*** ndipanov has quit IRC17:49
DuncanTsks: Firewall? Wrong port in the catalogue?17:49
sksDuncanT, If it is firewall issue then the nova list also should have been failed17:49
sksDuncanT, ok17:49
DuncanTsks: Nova and cinder are on different ports. Is this a devstack install? If so, unlikely to be firewall17:50
*** vilobhmm1 has quit IRC17:50
sksDuncanT, yes it is devstack install17:50
DuncanTsks: Do 'nova --debug list' and see if the connection string looks similar17:50
sksDuncanT, sure17:50
dansmithcan someone answer a question about volume creation where an image is involved (i.e. a BFV request)17:51
*** e0ne is now known as e0ne_17:52
DuncanTdansmith: Just ask, don't ask to ask. We'll answer if we know17:52
*** afazekas has quit IRC17:53
dansmithDuncanT: was finishing the -nova side of this, was planning to just ask :)17:53
dansmithwhen we create a volume and have the image laid down, what is the expected behavior with a uec image? does the rootfs get laid down and that's it? or is the kernel/ramdisk supposed to be injected and a bootloader put on somehow?17:54
*** dustins has joined #openstack-cinder17:55
DuncanTdansmith: I'm not certain, but I think only the rootfs gets laid down. Nomally I avoid uec images like the plague17:55
*** dustins_ has quit IRC17:55
dansmithDuncanT: hence why I don't know the answer to this :)17:55
DuncanTdansmith: Somebody might be able to tell you for sure17:55
DuncanTdansmith: But I remember I couldn't create working bootable volumes from them when I tried17:56
jbernarddansmith: with boot-from-volume, the image linking metadata is lost, no?17:56
dansmithwell, the volume boot scenario test was just re-enabled and it claims it should work :)17:56
dansmithjbernard: I don't know17:56
dansmithjbernard: if so, this should be failing for non-ceph cases too17:56
jbernarddansmith: i may retry with lvm and see if anything looks different17:57
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** ndipanov has joined #openstack-cinder17:58
dansmithDuncanT: does cinder import and store the image metadata with the volume?17:59
DuncanTdansmith: Yes. glance-metadata table17:59
*** e0ne has joined #openstack-cinder17:59
jgriffithdansmith: I believe the problem is that since Ceph typically *assumes* everything is on it's backend it doesn't do the convert18:00
jgriffithdansmith: if you look in the base drive you'll notice we download to a temp dir and run qemu-convert18:00
jgriffithdansmith: Ceph doesn't do that IIRC because it has "internal" knowledge of the image18:00
vilobhmm11jgriffith : qemu-img convert to a RAW format you mean18:01
jgriffithvilobhmm11: yes18:01
jgriffithamong other things...18:01
dansmithjgriffith: so in the normal case qemu-img convert should stuff the kernel and ramdisk into the image?18:01
jgriffithdansmith: I believe that's where the magic is yes18:02
dansmithjgriffith: okay, jbernard is going to try with lvm to see what a "normal" one looks like18:02
jgriffithdansmith: just a sec... looking in github18:02
*** diegows has quit IRC18:02
dansmithjgriffith: however, I'm not sure I get why this should really be different for ceph18:02
jgriffithdansmith: ok... so actually it does have the fetch_to_raw call18:02
jgriffithdansmith: yeah... agreed18:03
*** Yogi1 has joined #openstack-cinder18:03
jgriffithdansmith: in this case I don't think it should18:03
*** dannywilson has joined #openstack-cinder18:04
jbernardi am able to boot from volume on the command line, is something different about that?18:04
jgriffithjbernard: are you using a UEC?18:04
jbernardi did, yes18:05
*** e0ne is now known as e0ne_18:05
jbernarda day or so ago18:05
dansmithjbernard: the same image that tempest is configured with?18:05
jgriffithjbernard: hmm18:05
jbernardyep, as far as I know18:05
jgriffithjbernard: dansmith BTW, we should probably remove the duplicate code in the ceph driver for copy_image_to_volume and just use the base driver18:05
jbernardjgriffith: yeah, ill fix that up once i get this sorted18:06
jgriffithjbernard: dansmith haha... actually I lied :)18:06
jgriffithlooking at wrong tab18:06
jgriffithso I don't see any of the convert calls in the rbd driver18:07
dansmithokay, that would explain it18:07
dansmithjgriffith: so we have three options it sounds like:18:08
dansmith1. Make the driver do the conversion18:08
dansmith2. Skip this in tempest for ceph as "uec BFV not supported"18:08
*** nkrinner has joined #openstack-cinder18:08
dansmith3. Try to get the ceph job to use the disk-based cirros image18:08
dansmithsound right?18:08
*** nkrinner has quit IRC18:08
*** e0ne_ has quit IRC18:10
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** timcl has quit IRC18:12
dansmithsdague: is #3 above doable?18:14
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** xyang1 has quit IRC18:15
*** ndipanov has quit IRC18:16
*** lpetrut has joined #openstack-cinder18:17
*** MentalRay has quit IRC18:18
*** diegows has joined #openstack-cinder18:21
*** timcl has joined #openstack-cinder18:21
*** ndipanov has joined #openstack-cinder18:21
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** bkopilov has joined #openstack-cinder18:23
*** nkrinner has joined #openstack-cinder18:24
sksDuncanT, The connection string of nova and cinder seems to be different as the port for cinde ris diff18:27
*** Bsony_ has joined #openstack-cinder18:27
DuncanTsks: The port should be the only difference18:27
sksDuncanT, yes18:27
DuncanTsks: I'm not sure what else it can be then. You can try connecting to the port with nc or something to see if it is really open?18:27
*** Bsony has quit IRC18:27
sksDuncanT, I wonder I tried in a different system with f20 devstack and there was no issue18:28
*** e0ne has joined #openstack-cinder18:28
*** e0ne is now known as e0ne_18:29
sksDuncanT, I want to know the rootcause or its a bug18:29
*** e0ne_ is now known as e0ne18:29
sksDuncantT, connecting to port with nc?18:31
*** ndipanov has quit IRC18:31
sksahh not connected ?18:31
jgriffithdansmith: sorry... had to step away18:34
dansmithjgriffith: np18:34
jgriffithdansmith: I'd at least test that the conversion is the issue18:34
jgriffithdansmith: I'm curious about jbernard 's testing though18:34
*** timcl has quit IRC18:35
dansmithjgriffith: yep, we're doing that now18:35
*** erlon has joined #openstack-cinder18:35
dansmithjgriffith: the steps above are assuming that's the issue18:35
jgriffithdansmith: ahh... cool18:35
jgriffithgot ya18:35
dansmithjgriffith: is #1 an option? Assume it's an L thing if so18:35
DuncanTsks: nc = netcat, a simple tcp tool18:35
*** Bsony_ has quit IRC18:36
jgriffithdansmith: that where I'd start and see if it "works"18:36
DuncanTsks: run 'nc a.b.c.d <port>' and see if it connects18:36
jgriffithdansmith: I'd actually consider in L reworking that whole code path18:36
jgriffithdansmith: use the default method from the base driver if/when possible18:36
*** dflorea has joined #openstack-cinder18:36
dansmithjgriffith: what I mean is... is there any reason this is not done specifically for ceph (that you know of) or is this just an oversight?18:36
jgriffithand only do the backend Ceph magic when applicable (although that's likely to be 99% of the time in real world)18:37
sksDuncanT, ohh ok it is similar to ethereal18:37
dansmithcool, sounds like that'd be good18:37
jgriffithdansmith: yeah... to clarify I think it's NOT done now because typically Ceph is the Glance store and the Cinder backend18:37
jgriffithdansmith: so there's some 'magic' that it can do to make things more effecient and better18:37
dansmithjgriffith: sounds like it's that combined with "UECs are weird", but yeah18:38
*** bswartz has joined #openstack-cinder18:38
jgriffithdansmith: Yeah, that part is completely diff18:38
jgriffithdansmith: I don't know enough about the inner workings of Ceph to be of much help there18:38
jgriffiththingee: and jdurgin would know though probably18:38
jgriffithI think jdurgin wrote that code initially18:39
dansmithyep, I think jbernard is probably on that fix, if that's the fix18:39
*** timcl has joined #openstack-cinder18:47
*** timcl has quit IRC18:50
*** emagana has quit IRC18:50
*** ho has joined #openstack-cinder18:52
jgriffithmtreinish: :(  the introduction of oslo_config raises some problems18:55
*** dflorea has quit IRC18:56
*** bkopilov has quit IRC18:56
mtreinishit conflicts with argparse or ConfigParser?18:57
mtreinishI was afraid of that18:57
*** dflorea has joined #openstack-cinder18:57
mtreinishjgriffith: we can revert that patch and make it a todo, and we can look at ripping oslo.db out of subunit2sql18:58
jgriffithmtreinish: well... neither :)18:58
*** timcl has joined #openstack-cinder18:58
mtreinishoh, then what did it break>18:58
jgriffithmtreinish: double register:
mtreinishoh, because we're initing calling the register opts each time run is called18:59
jgriffithmtreinish: indeed18:59
mtreinishwe can move the parse call outside of run and that should work I think18:59
jgriffithmtreinish: those are threads18:59
jgriffithyeah, that's what I'm looking at right now19:00
jgriffithmtreinish: just wanted to keep you in the loop :)19:00
jgriffithmtreinish: and make sure I wasn't missing something19:00
mtreinishstill think I need to look at killing oslo.db there. There is no reason for me to be using it19:00
*** dflorea has quit IRC19:01
*** MentalRay has joined #openstack-cinder19:05
marcusvrnthingee: ping19:06
*** tellesnobrega_ has joined #openstack-cinder19:07
jgriffithmtreinish: looking at a quick hack to avoid revert19:08
*** tellesnobrega_ has quit IRC19:08
jgriffithmtreinish: but yeah.. maybe since I just got a ping "hey.. your sos-ci stuff doesn't work" LOL19:08
jgriffithmtreinish: I *think* could just set those config options when I init the threads19:09
*** emagana has joined #openstack-cinder19:09
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
mtreinishjgriffith: yeah that should work, since it should only be needed once per process space19:10
*** lifeless has quit IRC19:11
jgriffithmtreinish: yeah, going to try it now19:11
*** rushil has joined #openstack-cinder19:13
*** geguileo has quit IRC19:13
jgriffithmtreinish: I've always wanted to try passing a method to the thread constructor.. now's my chance :)19:13
jgriffithmtreinish: meh... I'll just make an init19:15
sdaguedansmith: so.... there was some reason why we are using euc images in devstack by default, but I can't tell you why now19:15
mtreinishjgriffith: heh, come on push the boundaries man19:15
*** Apoorva has quit IRC19:15
dansmithsdague: okay, we just confirmed that the problem is ceph+uec images19:15
jgriffithsdague: wasn't it because of size and compat for Xen at one ponit?19:15
sdagueI do agree using really img for boot from volume is probably more sensible19:15
dansmithsdague: okay, so should we convert devstack to that, or just make it optional and enable for the ceph job?19:16
sdaguejgriffith: maybe, I honestly dtroyer might know19:16
jgriffithdansmith: sdague so just my two cents... I'm not completely comfortable with the notion of changing how we do things because Ceph doesn't work19:16
jgriffithdansmith: not necessarily saying that's what you're proposing19:17
*** Rockyg has joined #openstack-cinder19:17
dansmithjgriffith: you mean the ceph cinder driver... :)19:17
sdaguedansmith: we need to get the img downloaded and installed ... probably as a second type. And I expect that tempest will need a config for 'use this other image'19:17
jgriffithdansmith: just kinda "icky"19:17
dansmithsdague: right, okay19:17
dansmithI'm not sure that using a uec image really mirrors how most people use openstack (without ec2 at least)19:18
dansmithso image-based seems more correct-er to me anyway, but adding it to glance as another option seems like a good plan19:18
jgriffithdansmith: maybe not.. BUT I would argue thatt given every other Cinder backend supports this Ceph shoudl as well19:18
dansmithjgriffith: no argument there19:18
jgriffithunless we're going to say OpenStack no longer supports UEC images?19:18
jgriffithdansmith: got ya19:18
sdaguejgriffith: any idea if it's ceph, or the cinder driver?19:19
jgriffithsdague: soryy.. I should clarify it's likely the ceph driver19:19
jgriffithsdague: in this case at least19:19
jgriffithsdague: Oh...wait19:19
jgriffithsdague: dansmith the Ceph devstack run uses Ceph as image store as well no?19:19
dansmithjgriffith: yeah19:19
jgriffiththat explains the differing results form jbernard maybe19:20
dansmithit really doesn't matter.. if the ceph cinder driver turns a UEC into what we're seeing, then it's broken19:20
jgriffithanyway... I don't want to confuse things.  I could be completely off base19:20
dansmithit either needs to fail that or do it right19:20
sdaguestill seems like there should be a bug on the cinder side though to fix the driver19:20
jgriffithdansmith: agree.. I'm just talking about how to figure out if it's *ceph* or a simple driver fix19:20
jgriffithsdague: +119:21
dansmithjgriffith: I don't think ceph knows anything about any of this19:21
*** rushiagr is now known as rushiagr_away19:21
dansmiththis is at least a layer above what ceph actually knows about, AFAIK19:21
dansmithit knows how to do the magic clone of the glance image to the thing cinder uses,19:21
jgriffithdansmith: perhaps, it's been a very long time since i've dug aroudn in there... but it used to be they had some internal knowledge for moving things from object to block to file etc19:22
dansmithbut it doesn't know about combining the UEC parts into a bootable linux image I don't think19:22
jgriffithbut I could be compeltely wrong19:22
dansmithsure, but that's not the problem here19:22
dansmithit's combining three things into one in the right way19:22
jgriffithand you're right, the UEC stuff in particular I wouldn't expect it to have any knowledge of at all19:22
jgriffithdansmith: yeah... I'm with ya.19:22
jgriffithdansmith: you want to file the bug or shall I?19:22
jgriffithdansmith: I def agree it's a driver bug19:23
dansmithjbernard might want the bug glory, I dunno :)19:23
jgriffithlol... I'll give him a chance to file it19:23
jgriffithif nothing there tonight I'll post some jibberish :)19:23
jgriffithwindow 4319:24
smcginnis43 windows? Dang.19:25
jgriffithsmcginnis: I need to close my session and start over at some point19:26
jgriffithsmcginnis: this byobu session has been up for almost a year19:26
smcginnisjgriffith: You know you can close some of those, right? :)19:26
jgriffithsmcginnis: I know... I'm lazy19:26
jgriffithsmcginnis: and irssi doesn't always "fix" up the numbers19:26
smcginnisjgriffith: Yeah, I've noticed that. Works good most of the time, but not always.19:27
jgriffithsmcginnis: so my count goes like 10,11,12,40,41,42,5019:27
*** dflorea has joined #openstack-cinder19:27
*** nkrinner has quit IRC19:28
smcginnisbyobu looks interesting. I've only ever used screen.19:28
jbernardjgriffith: ill file it, i have my own particular brand of jibberish :)19:29
jgriffithjbernard: :)19:30
jgriffithI find clippy being on my gerrit dashboard very disturbing19:31
jbernardi installed no-script specifically for those shenanigans19:31
jgriffithespecially when he/she is sleeping19:31
jgriffithjbernard: ahh.. nice!19:31
*** dflorea has quit IRC19:32
jgriffithjbernard: and the beauty is if I block something I don't want to, I'll never know :)19:32
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** lifeless has joined #openstack-cinder19:40
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** dflorea has joined #openstack-cinder19:40
*** reed has quit IRC19:40
*** dflorea has quit IRC19:42
*** dflorea has joined #openstack-cinder19:42
*** dflorea has quit IRC19:42
*** dflorea has joined #openstack-cinder19:43
*** sks has quit IRC19:47
*** cpg|away is now known as cpg19:48
*** nlevinki has quit IRC19:51
*** dflorea has quit IRC20:00
*** annashen has quit IRC20:00
openstackgerritAlessandro Pilotti proposed openstack/cinder: SMBFS: Lock on a per-volume basis
*** dflorea has joined #openstack-cinder20:01
*** lcurtis has joined #openstack-cinder20:04
*** dflorea has quit IRC20:05
*** annegent_ has joined #openstack-cinder20:06
openstackLaunchpad bug 1439371 in Cinder "Volume creation from image fails for UEC+Ceph" [Undecided,New] - Assigned to Jon Bernard (jbernard)20:09
jbernardjgriffith: milestone should be L, correct?20:09
thingeejbernard: bug fixes right now are ok20:16
*** vilobhmm11 has quit IRC20:16
*** vilobhmm1 has joined #openstack-cinder20:16
*** vilobhmm1 has quit IRC20:17
*** emagana has quit IRC20:17
jbernardthingee: ok, i still have a few gaps in my understanding, but ill try to get it sorted asap20:17
*** vilobhmm1 has joined #openstack-cinder20:17
*** vilobhmm1 has quit IRC20:18
casusbellijgriffith: Hi! Quick Q: Regarding the localconf.base in sos-ci, is the network config in there the config that you're really using? If not could you pls give me a paste or info on how you have that configured?20:18
*** vilobhmm1 has joined #openstack-cinder20:18
jbernardthingee: i didn't check the -initrd and -kernel options to qemu20:18
*** vilobhmm1 has quit IRC20:18
*** vilobhmm1 has joined #openstack-cinder20:18
jbernardthingee: presumably, if rbd only extracts the uec rootfs, libvirt could still boot the instance from a volume if it also fetched the remaining pieces, no?20:18
jbernardthingee: can you clarify what is /supposed/ to happen in that case?20:19
thingeemarcusvrn: pong20:22
jgriffithcasusbelli: it is20:22
casusbellijgriffith: thanks!20:23
jgriffithcasusbelli: np20:24
jgriffithcasusbelli: is the /20 throwing you off maybe?20:24
casusbellijgriffith: /20 ??20:24
marcusvrnthingee: does CI need to run tempest.thirdparty.boto tests, too?20:24
jgriffithcasusbelli: never mind then :)20:24
jgriffithbb in a hour20:24
casusbellijgriffith: :-D20:24
*** sseverson has joined #openstack-cinder20:24
*** vilobhmm1 has quit IRC20:27
casusbellijgriffith: ah, now got it, no different issue... :)20:28
*** Yogi1 has quit IRC20:30
*** emagana has joined #openstack-cinder20:30
*** annegent_ has quit IRC20:31
*** lpetrut has quit IRC20:32
thingeemarcusvrn: in my testing this is not skipped by default devstack install20:32
thingeeI think there is one skipped20:32
thingeemarcusvrn: setUpClass (tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest) ... SKIPPED: EC2 InstanceRunTest: requires ami/aki/ari manifest20:32
thingeemarcusvrn: but these aren't:20:33
thingeetempest.thirdparty.boto.test_ec2_volumes.EC2VolumesTest.test_create_get_delete [13.925216s] ... ok20:33
thingeetempest.thirdparty.boto.test_ec2_volumes.EC2VolumesTest.test_create_volume_from_snapshot [4.306667s] ... ok20:33
mtreinishthingee: heh, I forgot those boto tests would match a regex of 'volume'20:33
marcusvrnthingee: mine is not skipping and it is failing (not all time)20:34
marcusvrnthingee: xyang's CI is not skipping also (
*** annashen has joined #openstack-cinder20:35
thingeejbernard: I don't have experience of trying that with a UEC image, but worth a shot.20:36
thingeejbernard: I don't have any experience with uec really20:36
marcusvrnthingee: solidfire's CI is skipping it:
thingeemarcusvrn: just that one that I already listed here, yes20:37
thingeethe other two arent20:37
marcusvrnso....CIs are not running the same tests....20:37
thingeemarcusvrn: that was established on the list and pointed out20:37
thingeemarcusvrn: As I mentioned those two boto tests by default aren't skipped20:38
*** vilobhmm1 has joined #openstack-cinder20:39
thingeemarcusvrn: solidfire is doing the correct thing, only skipping the one because the aki manifest dne20:39
thingeemarcusvrn: so I'm not sure what you're a sking20:39
thingeemarcusvrn: the two I just pointed out in chat are running though and not being skipped20:39
marcusvrnthingee: how can I skip those boto tests while I'm trying to find out what is the problem with them?20:40
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Add volume:update rule to policy.json
thingeemtreinish: ^20:40
thingeemtreinish: not sure if we figured out how to allow ci's to skip stuff that's being fixed up20:41
mtreinishthingee: well the only method right now is to create a regex that excludes the tests you don't want to run20:42
mtreinishand use that instead of 'volume'20:42
mtreinishthingee: in the near term once I get os-testr released on pypi and tempest switched over to it there will be an exclude file people can use20:42
mtreinishso it'll make things a bit nicer20:43
thingeemtreinish: that sounds great long term, thanks20:43
*** melwitt has joined #openstack-cinder20:43
thingeemarcusvrn: so it's following
thingeemarcusvrn: but instead of volume, a regex to include volume tests, but exclude ones that contain boto20:43
mtreinishthingee: heh, well long term is as soon as I can get a +A on my project-config patch which already has 2 +2s :)20:44
thingeemarcusvrn: for now20:44
mtreinishthingee: I think something like: '(?!boto)(volume)' would work20:44
mtreinishbut I suck at regex so no guarantees20:44
marcusvrnthingee: I'm doing it, but it will be not acceptable after the deadline, will it?20:45
thingeemtreinish: rent a perl monk?20:45
thingeeI don't think these tests really have much to do with your driver itself failing, so I think it'll be fine...just get it fixed soon.20:46
thingeemarcusvrn: ^20:46
thingeemtreinish might correct me on that though20:47
mtreinishthingee: well, they might depending on which tests are failing they could just be tickling a bug in a feature20:47
mtreinishI'd have to see an example failure to really judge20:47
marcusvrngrep for tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest.test_compute_with_volumes[id-ab836c29-737b-4101-9fb9-87045eaf89e9]20:48
marcusvrnit seems like an SSH issue20:49
mtreinishmarcusvrn: well I found:
*** e0ne has quit IRC20:51
mtreinishyou might also want to switch to using config drive by default, because might mean that the metatadata service might be having issues20:51
*** e0ne has joined #openstack-cinder20:52
mtreinish(although it almost never works right on a good day)20:52
*** thingee has quit IRC20:53
mtreinishmarcusvrn: but yeah that particular failure might not be related to your driver, just a general ci config issue20:54
*** e0ne has quit IRC20:55
*** bkopilov has joined #openstack-cinder20:57
marcusvrnmtreinish: hmmm nice, so can I remove those tests for now and try to fix it up later, right? :)20:57
*** e0ne has joined #openstack-cinder20:58
*** rushil has quit IRC20:59
marcusvrnmtreinish: I have another issue, maybe you can help me...
mtreinishmarcusvrn: that's not my call. :) But, fwiw I think using the metadata service is probably the cause of your ssh failures and switching to config drive would help20:59
*** emagana has quit IRC21:00
*** emagana has joined #openstack-cinder21:00
marcusvrnmtreinish: that issue (minimum_basic_scenario) is failing on ssh authentication...It does all volume stuff, but fails on ssh authentication21:01
mtreinishmarcusvrn: in that case the test took too long, we kill any test which exceeds 500 secs21:01
*** dustins has quit IRC21:02
mtreinishthe traceback shows the fixture raised a timeout exception21:02
*** timcl has quit IRC21:03
mtreinishmarcusvrn: you can increase the timeout by setting: OS_TEST_TIMEOUT in the env that calls testr21:03
mtreinishmarcusvrn: if your backend or the ci environment is a bit slower than what we gate with, 500 secs might not be sufficient21:04
marcusvrnmtreinish: and about switching driver configurations,... I tried it out, but it didn't work... may be I didn't change the right config....21:04
*** emagana_ has joined #openstack-cinder21:04
*** emagana has quit IRC21:04
mtreinishmarcusvrn: you mean setting to default to config drive?21:05
*** e0ne has quit IRC21:05
mtreinishmarcusvrn: if so you should just need to set FORCE_CONFIG_DRIVE=True in your localrc/local.conf21:06
*** akerr has quit IRC21:06
marcusvrnmtreinish: ah... so that's the option that I need to set...21:06
marcusvrnthe OS_TEST_TIMEOUT, have I to set it in tempest.conf file?21:07
mtreinishmarcusvrn: no it's an env variable21:07
marcusvrnmtreinish: ok, I will export it21:08
*** thangp has quit IRC21:08
marcusvrnmtreinish:  the FORCE_CONFIG_DRIVE is already set as True...21:10
*** diegows has quit IRC21:10
*** Lee1092 has quit IRC21:10
smcginnismtreinish: Do you have a reference to FORCE_CONFIG_DRIVE that explains what that entails?21:12
*** thingee has joined #openstack-cinder21:13
smcginnismtreinish: nm, found it:
mtreinishmarcusvrn: hmm, yeah I see that in the logs now. I wonder what is querying the metadata server then?21:14
*** _cjones_ has quit IRC21:14
mtreinishbecause it should be attaching a config drive the guest and cloud-init uses that instead of asking the metadata server for anything21:14
jgriffithmarcusvrn: what?21:15
*** sgotliv has quit IRC21:15
mtreinishoh I wonder if the ec2 stuff does things differently21:16
*** julim has quit IRC21:16
marcusvrn mtreinish:hmmm21:16
marcusvrnjgriffith: what what?21:17
mtreinishyeah looking at a gate run, that's what it looks like, ec2 api calls are always querying the metadata server21:17
jgriffithmarcusvrn: what test is SolidFire skipping that others aren't?21:17
*** Liu has quit IRC21:18
marcusvrnjgriffith: ah, this one: tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest21:18
*** MentalRay has quit IRC21:18
smcginnisHow would that be run if you are filtering on 'volume'?21:18
*** MentalRay has joined #openstack-cinder21:19
jgriffithmarcusvrn: I'll check my enabled services, thanks21:19
jgriffithlet the stone throwing begin :)21:19
*** dims_ has quit IRC21:19
marcusvrnmtreinish: if I use the default metadata, Should it solve that problem?21:19
*** _cjones_ has joined #openstack-cinder21:20
*** dflorea has joined #openstack-cinder21:21
marcusvrnjgriffith: nooo, it's not my intention...I'm trying to fix this test in my driver and I saw that EMC CI is passing but yours is skipping I was trying to find why you guys are skipping it, because it can be the same problem that I have...21:22
marcusvrnjgriffith: I'm just trying to find out people with the same problem to try to solve it21:23
mtreinishmarcusvrn: what do you mean by default metadata? I think your error is caused by the nova metadata api service not working correctly which I thought default to always use config drive would fix21:23
*** Liu has joined #openstack-cinder21:24
mtreinishbut with ec2 requests it always queries the metadata api service so I'm not sure what to tell you next21:24
jgriffithmarcusvrn: so the secret lies in the local.conf21:24
jgriffithmarcusvrn: the different services that are enabled etc21:24
mtreinishjgriffith: heh, are you just disabling ec2?21:24
jgriffithmarcusvrn: and, not sure but I *think* EMC was running all tests, no filter21:24
mtreinishjgriffith: I wish I could just do that in tempest21:25
jgriffithmtreinish: marcusvrn nahh... just enabling the minimum:
marcusvrnmtreinish: hmm I see21:25
jgriffithmtreinish: and using "tox -eall -- volume"21:25
*** dflorea has quit IRC21:25
jgriffithto filter only on Volume tagged tests21:25
*** dflorea has joined #openstack-cinder21:25
jgriffithmarcusvrn: what's your tempst command look like?21:25
marcusvrnjgriffith: I'm using devstack gate, so I'm using 'export DEVSTACK_GATE_TEMPEST_REGEX="volume"'21:26
jgriffithmarcusvrn: hmm... that should essentially be the same I think21:27
jgriffithmarcusvrn: well... other than the "other" config options21:27
jgriffithmarcusvrn: suppose I could modfiy my setup to install "everything" and duplicate what you've got going on21:27
*** boris-42 has quit IRC21:28
marcusvrnjgriffith: and I have the same services enabled as you have:
mtreinishmarcusvrn: do you have an example of jgriffith's solidfire runs that has the ec2 test your failing on skipped?21:29
*** cbader has quit IRC21:30
marcusvrnmtreinish: hm.... one sec21:30
jgriffithmarcusvrn: forgot the test in question already :)21:30
marcusvrnjgriffith: tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest21:30
mtreinishmarcusvrn: ah it's because he isn't setting up the ec2 images so it's getting skipped in setup class21:31
jgriffithmtreinish: correct good sir21:32
mtreinish{2} setUpClass (tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest) ... SKIPPED: EC2 InstanceRunTest: requires ami/aki/ari manifest21:32
*** dflorea has quit IRC21:32
*** dflorea has joined #openstack-cinder21:33
mtreinishman I hate the boto test code...21:33
jgriffithmtreinish: I hate boto21:33
mtreinishman is this clear:
marcusvrnmriedem: jgriffith: the question is...Is that test necessary? because I think what are been tested in that test, are tested in another tests, correct me if I'm wrong..21:36
marcusvrnops.. mtreinish**21:36
mriedemmtreinish: it's the decider, what's not clear about it?21:37
*** dflorea has quit IRC21:37
marcusvrnmriedem: sry, I tagged you wrongly21:39
jgriffithmarcusvrn: So I think this might be the same *problem* dansmith is working on with uec images on the Ceph driver21:40
mriedemmarcusvrn: i know :)21:40
mriedemhappens a lot21:40
jgriffithmriedem: :)21:40
mriedembut now i'm here and you're all going to have to deal with it21:40
mtreinishmarcusvrn: I can't really judge whether it's necessary in your use case (personally I want to nuke the ec2 tests in tempest)21:41
marcusvrnmriedem: lol :)21:41
mtreinishI think the test in question does overlap, if at least partially, with one of the volume scenario tests21:41
jgriffithmarcusvrn: any indication why it fails?21:41
mtreinishbut I'd have to compare the code to be sure21:41
mtreinishjgriffith: he's getting an error from nova metadata when ec2 tries to talk to it and he's never able to login because of that21:42
jgriffithnot what I thought21:42
mtreinishof course the nova metadata logging doesn't really say anything:
*** Rockyg has quit IRC21:43
mtreinishoh, wait that's the 3rd failure I wonder if it's just a cascade issue from the first one21:43
mtreinishthat's probably the real reason your stuff is failing21:44
jgriffithcan you have metadata service without config drive?21:44
mtreinishjgriffith: yeah, they're not mutual exclusive21:45
jgriffithk... I'm painfully unfamiliar :(21:45
jgriffithmarcusvrn: FWIW, I'll spin it up and see if I get the same problem21:45
mtreinishforce_config_drive is just a way to tell nova to boot guests assuming all incoming api requests have the use config drive api param set21:46
jgriffithmtreinish: Ahh... ok21:46
mtreinishwhich would bypass having cloud init query the metadata server for anything21:46
mtreinish(of course not with ec2 requests because it always goes to metadata for those)21:46
jgriffithOhhh.... I remember this21:47
*** xyang1 has joined #openstack-cinder21:47
marcusvrnjgriffith:  ok21:47
mtreinishjgriffith: we do it on purpose because using the metadata service (especially with neutron) goes wrong more often than not21:47
jgriffith^^ I'll be of no help to you here21:48
mtreinishbecause it's a terrible idea to hard code an ip for things to query to get the metadata21:48
marcusvrnthis IP is my local repository server21:50
*** mriedem is now known as mriedem_away21:51
marcusvrnIs this not the right way to point to the local repository?21:52
jgriffithUmmm... stupid question, but; why are you setting that anywa marcusvrn ?21:55
jgriffithlooking at logs form a recent dsv-full run by the gate:
*** vilobhmm1 has quit IRC21:57
jgriffithmarcusvrn: we don't even run it in the gate21:57
*** boris-42 has joined #openstack-cinder21:57
*** thingee has quit IRC21:59
marcusvrnjgriffith: to try to make builds faster, because, initially,  heat was enabled and it uses fedora image, which is a little bit huge to download every time....then as I had to setup local repo to fedora image, I put all images needed, locally21:59
marcusvrnjgriffith:'re talking about the test...lol22:00
*** emagana_ has quit IRC22:01
marcusvrnjgriffith: actually, I'm running this test because I'm  using the wiki cinder regex and it runs that test22:02
*** dflorea has joined #openstack-cinder22:04
marcusvrnI can exclude that test, ... but I don't know if it is acceptable, once I'm excluding a test manually...22:04
jgriffithmarcusvrn: it's the local.conf22:04
*** ekarlso has quit IRC22:04
*** ekarlso has joined #openstack-cinder22:04
jgriffithnot the regex22:04
*** vilobhmm1 has joined #openstack-cinder22:07
marcusvrnbut I have the same SKIP_EXERCISES ->
*** dflorea has quit IRC22:08
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Fix missing translations for log messages
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Logging not using oslo.i18n guidelines
smcginnishemna: Thanks!22:12
*** 77CAAKTHH has joined #openstack-cinder22:13
*** scottda_ has joined #openstack-cinder22:16
*** thingee has joined #openstack-cinder22:16
marcusvrnjgriffith: well....I have to... tomorrow morning I'll continue on it....22:17
marcusvrnhave to go*22:17
marcusvrnmtreinish: jgriffith thanks a lot :)22:18
*** asselin has joined #openstack-cinder22:20
*** scottda_ has quit IRC22:21
*** reed has joined #openstack-cinder22:21
jgriffithmarcusvrn: hehe.. what I'm saying is remove the uec image entries from your local.conf22:22
jgriffithmarcusvrn: we don't run those in dsvm-full in the gate anyway22:22
jgriffithmarcusvrn: unless you just want to be an over-achiever :)22:22
*** zhipeng has joined #openstack-cinder22:23
*** 77CAAKTHH has quit IRC22:23
*** asselin has quit IRC22:25
*** esker has quit IRC22:29
*** scottda_ has joined #openstack-cinder22:29
*** lcurtis has quit IRC22:30
*** emagana has joined #openstack-cinder22:35
*** thingee has quit IRC22:42
*** ebalduf has joined #openstack-cinder22:43
*** thingee has joined #openstack-cinder22:43
*** emagana has quit IRC22:43
*** mwichmann has quit IRC22:45
*** ebalduf has quit IRC22:47
*** annegentle has joined #openstack-cinder22:48
*** annegentle has quit IRC22:48
*** mwichmann has joined #openstack-cinder22:57
*** annegentle has joined #openstack-cinder22:58
*** hemna is now known as hemnafk22:59
*** mwichmann has quit IRC23:00
*** mwichmann has joined #openstack-cinder23:05
*** scottda_ has quit IRC23:06
*** chlong has joined #openstack-cinder23:07
*** zhipeng has quit IRC23:13
*** annashen has quit IRC23:14
*** annegentle has quit IRC23:15
*** iwao has joined #openstack-cinder23:16
*** dflorea has joined #openstack-cinder23:16
*** scottda_ has joined #openstack-cinder23:19
*** annegentle has joined #openstack-cinder23:29
*** harlowja has quit IRC23:31
*** harlowja has joined #openstack-cinder23:32
*** annegentle has quit IRC23:37
*** dflorea has quit IRC23:39
*** dflorea has joined #openstack-cinder23:40
*** mwichmann has quit IRC23:42
*** scottda_ has quit IRC23:43
*** dflorea has quit IRC23:44
*** harlowja has quit IRC23:49
*** harlowja has joined #openstack-cinder23:50
*** zongliang has joined #openstack-cinder23:54
openstackgerritHahyun proposed openstack/os-brick: Fix wrong command for _rescan_multipath The _run_multipath command is wrong, it should be 'multipath -r' not 'multipath - r'.

Generated by 2.14.0 by Marius Gedminas - find it at!