Friday, 2015-05-29

*** arif-ali has quit IRC00:00
*** vokt has quit IRC00:01
*** vokt has joined #openstack-cinder00:01
*** arif-ali has joined #openstack-cinder00:04
*** vokt has quit IRC00:06
*** yamada-h has joined #openstack-cinder00:07
*** yamada-h has quit IRC00:12
*** markvoelker has quit IRC00:12
asselin_patrickeast, my test now has 4 fc devices attached!00:15
*** setmason has joined #openstack-cinder00:16
patrickeastso much bandwidth!00:16
patrickeastwere you guys testing with multipath enabled before?00:17
asselin_patrickeast, yes, multipath enabled, but only one fc path00:19
*** ebalduf has joined #openstack-cinder00:19
asselin_I just switched to the common fc script and it picked up all 400:19
asselin_will see how to limit it to e.g. two00:19
asselin_or 'n'00:19
patrickeastasselin_: yea i was thinking about that, starts to get hard since we need a way to know which ones to try for, and what to do if we don’t get them00:20
asselin_my version would skip and keep going...i'll do a diff and see what else is missing00:20
patrickeastasselin_: like should it abort if it doesnt get n? or try for n and wait maybe up to some threshold and then give up00:20
asselin_patrickeast, oh I see....good question00:21
asselin_well...I'll start with the simple: get n or abort00:21
patrickeastsounds like a good plan to me00:21
*** ebalduf has quit IRC00:23
nikeshmpatriceast: i want to check recheck myvendor   is this correct in layout.yaml of zuul
nikeshmits not working for me in sandbox00:24
*** emagana has quit IRC00:25
*** heyun has joined #openstack-cinder00:25
*** IlyaG has quit IRC00:26
patrickeastnikeshm: i think you might need a space before DotHill00:30
patrickeastmines:  comment: (?i)^(Patch Set [0-9]+:)?( [\w\\+-]*)*(\n\n)?\s*recheck(( (?:bug|lp)[\s#:]*(\d+))|( Pure Storage CI))\s*$00:30
patrickeastwhich other than the space is the same as yours00:30
* asselin_ checking00:30
asselin_I use (?i)^(Patch Set [0-9]+:)?( [\w\\+-]*)*(\n\n)?\s*recheck\s*$00:32
patrickeastive got that one too00:32
asselin_but seems this is what upstream has
asselin_comment: (?i)^(Patch Set [0-9]+:)?( [\w\\+-]*)*(\n\n)?\s*(recheck|reverify)00:33
asselin_will probably change to that00:33
*** setmason_ has joined #openstack-cinder00:39
*** setmason has quit IRC00:41
*** ociuhandu has quit IRC00:41
*** setmason_ is now known as setmason00:41
*** dims_ has quit IRC00:42
*** annegentle has joined #openstack-cinder00:45
*** dims_ has joined #openstack-cinder00:47
*** Longgeek has joined #openstack-cinder00:49
*** tobe has joined #openstack-cinder00:52
*** IlyaG has joined #openstack-cinder00:54
*** IlyaG has quit IRC00:55
*** _cjones_ has quit IRC01:00
*** annegentle has quit IRC01:03
*** jwcroppe has quit IRC01:03
openstackgerritRick Chen proposed openstack/cinder: Use snapshot objects instead of snapshot dictionaries
*** jwcroppe has joined #openstack-cinder01:03
*** manishg has left #openstack-cinder01:04
*** jwcroppe has quit IRC01:08
*** pboldin has quit IRC01:09
*** garthb_ has quit IRC01:12
*** garthb has quit IRC01:12
*** markvoelker has joined #openstack-cinder01:13
openstackgerritDanny Wilson proposed openstack/cinder: Add Pure Storage FibreChannel driver
*** vilobhmm has quit IRC01:17
*** markvoelker has quit IRC01:18
*** tobe has quit IRC01:22
*** Lee1092 has joined #openstack-cinder01:23
*** pboldin has joined #openstack-cinder01:25
*** alexpilotti has joined #openstack-cinder01:28
*** emagana has joined #openstack-cinder01:32
*** dmz has quit IRC01:33
*** dims_ has quit IRC01:35
*** setmason_ has joined #openstack-cinder01:36
*** alexpilotti has quit IRC01:38
*** setmason has quit IRC01:39
*** setmason_ is now known as setmason01:39
*** dims_ has joined #openstack-cinder01:39
*** pboldin has quit IRC01:45
*** dims_ has quit IRC01:45
*** tobe has joined #openstack-cinder01:52
*** patrickeast has quit IRC01:54
*** emagana has quit IRC01:57
*** emagana has joined #openstack-cinder01:58
*** yamada-h has joined #openstack-cinder02:09
*** Apoorva has quit IRC02:13
*** yamada-h has quit IRC02:14
*** Longgeek has quit IRC02:16
*** Apoorva has joined #openstack-cinder02:20
*** mriedem has quit IRC02:20
nikeshmasselin: hi02:22
*** Apoorva has quit IRC02:24
*** rwsu has quit IRC02:25
*** earlephilhower has quit IRC02:26
*** setmason has quit IRC02:31
openstackgerritRick Chen proposed openstack/cinder: Use snapshot objects instead of snapshot dictionaries
asselin_nikeshm, hi02:38
*** jungleboyj has joined #openstack-cinder02:38
nikeshmasselin_: looks like my gearman is not working,so after recheck its not going to jenkins to run job02:40
asselin_nikeshm, double-check your regex & /var/log/zuul/debug.log file02:40
*** rajinir has joined #openstack-cinder02:43
*** Yogi1 has joined #openstack-cinder02:47
*** yamada-h has joined #openstack-cinder02:49
*** markvoelker has joined #openstack-cinder03:02
*** lixiaoy1 has joined #openstack-cinder03:03
openstackgerritTina Tang proposed openstack/cinder: Multiple pools support enhancement in VNX cinder driver
*** _cjones_ has joined #openstack-cinder03:04
*** _cjones_ has quit IRC03:05
*** rajinir has quit IRC03:05
*** _cjones_ has joined #openstack-cinder03:05
*** rajinir has joined #openstack-cinder03:06
*** markvoelker has quit IRC03:06
*** yamada-h has quit IRC03:07
openstackgerritJohn Griffith proposed openstack/cinder: Remove the no-parallel setting from
*** yamada-h has joined #openstack-cinder03:09
*** yamada-h has quit IRC03:09
*** yamada-h has joined #openstack-cinder03:10
*** yamada-h has quit IRC03:10
*** rajinir has quit IRC03:10
*** barra204 has quit IRC03:11
*** barra204 has joined #openstack-cinder03:13
*** dmz has joined #openstack-cinder03:15
*** vokt has joined #openstack-cinder03:18
*** vokt has quit IRC03:24
nikeshmasselin: working now03:26
*** harlowja_at_home has joined #openstack-cinder03:35
*** Yogi1 has quit IRC03:35
*** lixiaoy1 has left #openstack-cinder03:35
*** ks_ne has joined #openstack-cinder03:37
*** _cjones_ has quit IRC03:40
*** pradipta has joined #openstack-cinder03:42
*** yamada-h has joined #openstack-cinder03:42
*** yamada-h has quit IRC03:43
*** yamada-h has joined #openstack-cinder03:44
openstackgerritwanghao proposed openstack/cinder: Remove un-used import at
*** jwcroppe has joined #openstack-cinder03:46
*** yamada-h has quit IRC03:56
openstackgerritRick Chen proposed openstack/cinder: Prophetstor driver needs to return snapshot objects for create_cgsnapshot and delete_cgsnapshot.
openstackgerritRick Chen proposed openstack/cinder: Complete switch to snapshot objects
openstackgerritRick Chen proposed openstack/cinder: Switch get_all_snapshots to use objects
*** setmason has joined #openstack-cinder04:00
*** markvoelker has joined #openstack-cinder04:02
*** markvoelker has quit IRC04:07
openstackgerritwanghao proposed openstack/cinder: Implement function to manage/unmanage snapshots
*** yamada-h has joined #openstack-cinder04:11
*** garthb_ has joined #openstack-cinder04:19
*** garthb has joined #openstack-cinder04:19
openstackgerritTina Tang proposed openstack/cinder: Multiple pools support enhancement in VNX cinder driver
*** dannywilson has quit IRC04:28
*** harlowja_at_home has quit IRC04:28
*** jdurgin has joined #openstack-cinder04:28
*** sks has joined #openstack-cinder04:29
openstackgerritVincent Hou proposed openstack/cinder: Volume status management during migration
*** Longgeek has joined #openstack-cinder04:36
*** _cjones_ has joined #openstack-cinder04:37
*** Longgeek has quit IRC04:45
*** Longgeek has joined #openstack-cinder04:45
*** setmason has quit IRC04:48
*** setmason has joined #openstack-cinder04:50
*** tobe has quit IRC04:52
*** _cjones_ has quit IRC04:53
*** _cjones_ has joined #openstack-cinder04:53
*** _cjones_ has quit IRC04:54
*** avishay has joined #openstack-cinder05:10
*** vilobhmm has joined #openstack-cinder05:10
*** tobe has joined #openstack-cinder05:14
*** avishay has quit IRC05:22
*** avishay_ has joined #openstack-cinder05:22
*** avishay_ has quit IRC05:27
*** avishay_ has joined #openstack-cinder05:28
*** yamada-h has quit IRC05:29
*** IanGovett has joined #openstack-cinder05:39
*** setmason_ has joined #openstack-cinder05:40
*** garthb_ has quit IRC05:42
*** garthb has quit IRC05:42
*** vilobhmm has quit IRC05:43
*** setmason has quit IRC05:43
*** nkrinner has joined #openstack-cinder05:44
*** rushiagr_away is now known as rushiagr05:44
*** lixiaoy1 has joined #openstack-cinder05:45
*** setmason_ has quit IRC05:46
*** tobe has quit IRC05:47
*** avishay_ has quit IRC05:47
*** markvoelker has joined #openstack-cinder05:52
*** markvoelker has quit IRC05:56
openstackgerritRick Chen proposed openstack/cinder: Prophetstor driver needs to return snapshot objects for create_cgsnapshot and delete_cgsnapshot.
openstackgerritPeter Wang proposed openstack/cinder: Add 'source-id' and 'source-name' support in VNX driver
*** Maike has joined #openstack-cinder06:10
*** Longgeek has quit IRC06:11
*** tobe has joined #openstack-cinder06:15
*** Longgeek has joined #openstack-cinder06:16
*** nihilifer has joined #openstack-cinder06:18
*** Miouge_ has joined #openstack-cinder06:18
*** Miouge has quit IRC06:20
*** Miouge_ is now known as Miouge06:20
*** anshul has joined #openstack-cinder06:22
openstackgerrityogeshprasad proposed openstack/cinder: Add chap support to CloudByte cinder driver
*** IanGovett has quit IRC06:26
openstackgerritrakesh mishra proposed openstack/cinder: set-max-size-limit-for-per-volume
*** sks has quit IRC06:33
*** lpetrut has joined #openstack-cinder06:38
*** kaisers1 has joined #openstack-cinder06:39
*** e0ne has joined #openstack-cinder06:42
*** sks has joined #openstack-cinder06:48
*** jdurgin has quit IRC06:48
*** e0ne has quit IRC06:49
*** alecv has joined #openstack-cinder06:54
*** Maike_ has joined #openstack-cinder06:57
*** Maike_ has quit IRC06:59
*** Maike_ has joined #openstack-cinder06:59
*** yamada-h has joined #openstack-cinder07:00
*** Maike has quit IRC07:00
*** Miouge_ has joined #openstack-cinder07:02
*** kaisers2 has joined #openstack-cinder07:03
*** Miouge has quit IRC07:03
*** Miouge_ is now known as Miouge07:03
*** leopoldj has joined #openstack-cinder07:04
*** kaisers2 has quit IRC07:06
*** markus_z has joined #openstack-cinder07:08
openstackgerritVipin Balachandran proposed openstack/cinder: VMware: Create volume backing in specific clusters
*** dulek has joined #openstack-cinder07:18
*** ronis has joined #openstack-cinder07:23
*** yamada-h has quit IRC07:23
*** Miouge has quit IRC07:25
*** yamada-h has joined #openstack-cinder07:32
*** jwcroppe has quit IRC07:32
*** jwcroppe has joined #openstack-cinder07:33
*** Miouge has joined #openstack-cinder07:33
*** jwcroppe has quit IRC07:37
*** jistr has joined #openstack-cinder07:38
*** bkopilov has quit IRC07:39
*** markvoelker has joined #openstack-cinder07:40
*** markvoelker has quit IRC07:45
openstackgerritPranali Deore proposed openstack/cinder: Return 400 error instead of 500 for invalid body
*** Maike_ has quit IRC07:52
*** Longgeek has quit IRC07:54
*** Maike has joined #openstack-cinder08:02
openstackgerritVincent Hou proposed openstack/cinder: Volume status management during migration
*** Maike_ has joined #openstack-cinder08:10
*** dims_ has joined #openstack-cinder08:12
*** Maike has quit IRC08:14
*** Maike has joined #openstack-cinder08:14
*** kaisers1 has quit IRC08:15
*** Maike_ has quit IRC08:15
*** dims_ has quit IRC08:18
*** yamada-h has quit IRC08:18
*** tobe has quit IRC08:23
*** jordanP has joined #openstack-cinder08:27
*** tobe has joined #openstack-cinder08:29
*** alecv has quit IRC08:33
openstackgerritRick Chen proposed openstack/cinder: Prophetstor driver needs to return snapshot objects for create_cgsnapshot and delete_cgsnapshot.
*** lpetrut has quit IRC08:47
*** ndipanov has joined #openstack-cinder08:49
openstackgerritPradeep Sathasivam proposed openstack/cinder: Adds friendly zone name support
*** Miouge has quit IRC08:58
*** alecv has joined #openstack-cinder09:01
*** Miouge has joined #openstack-cinder09:01
*** Miouge has quit IRC09:01
openstackgerritPranali Deore proposed openstack/cinder: Made error message consistent across all apis
*** ankit_ag has joined #openstack-cinder09:05
openstackgerritVincent Hou proposed openstack/cinder: Test
openstackgerritwanghao proposed openstack/cinder: Notification with volume and snapshot metadata
*** Maike has quit IRC09:19
wanghaoDuncanT: ping09:28
wanghaoDuncanT:  I'm very very sad to have to rebase Notification with volume and snapshot metadata again after Mike has workflow +1 for it.09:29
*** markvoelker has joined #openstack-cinder09:29
wanghaoDuncanT: Cos the merge confilct.09:29
DuncanTwanghao: Sigh. This does happen often, unfortunately09:29
DuncanTwanghao: Once Jenkins has passed, I'll look again then maybe make it the first candidate for casual review friday09:30
wanghaoDuncaT: That's great. Thank you a lot.09:31
wanghaoDuncanT: For those days to bother you....09:31
DuncanTwanghao: You're welcome. Thanks for sticking with it09:32
wanghaoDuncanT: Got it.09:33
*** markvoelker has quit IRC09:34
*** alecv has quit IRC09:34
wanghaoDuncanT:  I'm done for the day. Go home for Friday night.  Have a good day:)09:36
*** lpetrut has joined #openstack-cinder09:36
DuncanTwanghao: And you, good luck!09:37
*** lpetrut1 has joined #openstack-cinder09:39
DuncanTAny netapp folks awake to take a look at please? The 'fix' has changed the semantics of the client, even if the original semantics were nonsense, so I'd like somebody to take a quick look, please09:40
*** lpetrut has quit IRC09:41
*** aix has joined #openstack-cinder09:41
*** pradipta has quit IRC09:45
*** lixiaoy1 has quit IRC09:45
openstackgerritJessy Lee proposed openstack/cinder: Add volume drivers for Infortrend Storage
openstackgerritTina Tang proposed openstack/cinder: Multiple pools support enhancement in VNX cinder driver
*** dims_ has joined #openstack-cinder09:51
*** bkopilov has joined #openstack-cinder09:53
*** pboldin has joined #openstack-cinder10:06
*** dulek_ has joined #openstack-cinder10:08
*** dulek has quit IRC10:10
*** afazekas has joined #openstack-cinder10:12
*** Maike has joined #openstack-cinder10:14
*** heyun has quit IRC10:16
*** anshul has quit IRC10:35
*** pboldin has quit IRC10:36
nikeshmtoday i changed my CI to work on cinder patches10:39
nikeshmDuncanT: i done recheck on my cinder patch too
nikeshminitially it failed then i figured out the reason and now it all tests should pass10:44
nikeshmtomorrow will ask walter boring to remove -2 and give =210:45
*** anshul has joined #openstack-cinder10:46
nikeshmif yu will be available tomorrow then had a look on patch and give some comments10:46
*** dulek_ is now known as dulek10:47
*** pboldin has joined #openstack-cinder10:52
*** dims_ has quit IRC10:52
*** aix has quit IRC10:54
openstackgerritGorka Eguileor proposed openstack/cinder: Fix backup metadata import missing fields
*** zhenguo has quit IRC10:58
*** dulek has quit IRC10:58
*** Miouge has joined #openstack-cinder10:59
DuncanTnikeshm: Looking at the patch now11:03
DuncanTnikeshm: Looks like it is still getting a -2 from your CI though?11:03
*** dulek has joined #openstack-cinder11:06
nikeshmDuncanT: in initial recheck it was,but that was due to some setup issue in my CI,i resolved that and after checking locally again rerun recheck11:07
nikeshmDuncanT: now every tests are passing11:08
nikeshmand sooner my CI will update on patch11:08
DuncanTAh, ok11:08
nikeshmDuncanT: talked to walter today morning,he said as soom my CI will come on patch,he will remove -211:09
nikeshmDuncanT: tempest passed in my system,it will take 20 minutes to upload logs,11:11
nikeshmand then it will come in that patch11:11
nikeshmmy log server is in INDIA11:11
nikeshmand this system running CI is in US11:11
DuncanTI'm looking at the patch, I'll leave comments or a +211:11
*** lpetrut1 has quit IRC11:12
DuncanTWe can sell you a nice log server in the cloud in the US if that helps ;-)11:12
*** mdbooth has quit IRC11:12
nikeshmDuncanT:ha ha, sooner i have to upload patches for HPMSA,so will ask HP to provide11:13
*** merooney has joined #openstack-cinder11:13
nikeshmi am in san jose,california for a short trip11:15
nikeshmbut works in hyderabad India11:15
nikeshmfor vedams11:15
*** mdbooth has joined #openstack-cinder11:17
*** markvoelker has joined #openstack-cinder11:18
nikeshmthanks for looking on the patch11:18
nikeshmgoing to sleep its 4:19 am of morning for me11:19
*** anshul has quit IRC11:19
nikeshmand also this setup is for iscsi and we are working on FC too11:21
nikeshmand we are skipping four tests like we discussed11:22
*** markvoelker has quit IRC11:22
*** merooney has quit IRC11:35
*** cdelatte has quit IRC11:38
*** tobe has quit IRC11:39
*** cdelatte has joined #openstack-cinder11:39
openstackgerritVipin Balachandran proposed openstack/cinder: VMware: Create volume backing in specific clusters
*** aix has joined #openstack-cinder11:45
*** dims_ has joined #openstack-cinder11:50
*** rushiagr is now known as rushiagr_away11:50
*** Miouge has quit IRC11:52
*** sgotliv has quit IRC11:53
*** ganso_ has joined #openstack-cinder11:53
*** dulek has quit IRC11:54
*** dulek has joined #openstack-cinder11:55
*** markvoelker has joined #openstack-cinder11:59
*** markus_z has quit IRC12:00
*** nkrinner has quit IRC12:02
*** dulek has quit IRC12:03
*** markus_z has joined #openstack-cinder12:04
*** Yogi1 has joined #openstack-cinder12:05
*** dulek has joined #openstack-cinder12:05
*** nkrinner has joined #openstack-cinder12:05
*** aix has quit IRC12:06
*** aix has joined #openstack-cinder12:08
*** Yogi1 has quit IRC12:10
tbarronDuncanT: thanks for calling our attention to
openstackgerritVincent Hou proposed openstack/cinder: Test
*** tobe has joined #openstack-cinder12:13
*** tobe has quit IRC12:14
*** marcusvrn has joined #openstack-cinder12:14
tbarronDuncanT: we'll update this review today (I'm just running git blame now and seeing who to talk to & may run the candidate change through our CI pre-merge)12:15
DuncanTtbarron: you're welcome. Thanks for jumping on it12:18
openstackgerritPranali Deore proposed openstack/cinder: Made error message consistent across all apis
tbarronDuncanT: same author, no longer with us, for the nonsense dataontap/ and eseries/client.py12:23
tbarronI'm inclined just to endorse the change but since someone has asked cknight for his opinion in the review, let's get that as well12:23
*** anshul has joined #openstack-cinder12:26
DuncanTIt looks like the change is sane, but it's your driver12:29
openstackgerritVipin Balachandran proposed openstack/cinder: VMware: Create volume backing in specific clusters
openstackgerritPranali Deore proposed openstack/cinder: Validate name and description string
*** lpetrut has joined #openstack-cinder12:38
*** bswartz has quit IRC12:39
*** marcusvrn1 has joined #openstack-cinder12:40
*** fthiagogv has joined #openstack-cinder12:41
*** Yogi11 has joined #openstack-cinder12:42
*** rushiagr_away is now known as rushiagr12:43
*** marcusvrn has quit IRC12:44
*** marcusvrn1 has quit IRC12:48
*** marcusvrn has joined #openstack-cinder12:48
*** sks has quit IRC12:51
*** Yogi11 has quit IRC12:53
*** med_ has joined #openstack-cinder12:54
*** med_ has quit IRC12:54
*** med_ has joined #openstack-cinder12:54
*** marcusvrn has quit IRC12:55
*** merooney has joined #openstack-cinder13:01
*** marcusvrn has joined #openstack-cinder13:08
*** primechuck has joined #openstack-cinder13:09
*** nkrinner has quit IRC13:10
*** xyang1 has joined #openstack-cinder13:11
*** Tross1 has quit IRC13:13
*** mriedem has joined #openstack-cinder13:14
*** akshai has joined #openstack-cinder13:15
*** merooney has quit IRC13:16
*** merooney has joined #openstack-cinder13:23
*** mtanino has joined #openstack-cinder13:26
*** annegentle has joined #openstack-cinder13:29
*** Yogi1 has joined #openstack-cinder13:29
*** sgotliv has joined #openstack-cinder13:30
*** nihilifer has quit IRC13:32
openstackgerritMichal Dulko proposed openstack/cinder: Refactor scheduler's create_volume flow
*** rushiagr is now known as rushiagr_away13:35
*** jungleboyj has quit IRC13:35
*** nihilifer has joined #openstack-cinder13:37
*** marcusw has joined #openstack-cinder13:39
*** merooney has quit IRC13:39
*** pboldin has quit IRC13:43
*** bswartz has joined #openstack-cinder13:44
*** annegent_ has joined #openstack-cinder13:44
*** annegentle has quit IRC13:45
*** Tross has joined #openstack-cinder13:45
*** aix has quit IRC13:49
*** aix has joined #openstack-cinder13:49
*** timcl has joined #openstack-cinder13:56
*** annegent_ has quit IRC13:57
openstackgerritxing-yang proposed openstack/cinder: EMC VMAX Manage/Unmanage Volume
*** merooney has joined #openstack-cinder13:59
*** krtaylor has quit IRC14:02
*** ankit_ag has quit IRC14:02
*** Maike_ has joined #openstack-cinder14:03
openstackgerritrakesh mishra proposed openstack/cinder: set-max-size-limit-for-volume
*** Maike_ has quit IRC14:04
*** emagana has quit IRC14:04
*** Maike_ has joined #openstack-cinder14:04
*** mtanino has quit IRC14:05
*** rushil has joined #openstack-cinder14:05
*** anshul has quit IRC14:06
*** afazekas has quit IRC14:06
tbarronDuncanT: you can remove your -1 from now.  Thanks again for the caution.14:06
*** Maike has quit IRC14:07
*** annegentle has joined #openstack-cinder14:10
*** sgotliv has quit IRC14:11
openstackgerritwanghao proposed openstack/cinder: Notification with volume and snapshot metadata
*** crose has joined #openstack-cinder14:13
*** krtaylor has joined #openstack-cinder14:13
*** lpetrut1 has joined #openstack-cinder14:15
*** lpetrut has quit IRC14:16
*** dims_ has quit IRC14:16
*** leopoldj has quit IRC14:17
*** breitz has quit IRC14:20
*** breitz has joined #openstack-cinder14:21
*** rushil has quit IRC14:21
*** ronis has quit IRC14:22
*** jwcroppe has joined #openstack-cinder14:24
*** jistr has quit IRC14:26
*** dulek has quit IRC14:27
*** rushil has joined #openstack-cinder14:30
*** esker has joined #openstack-cinder14:35
*** Maike_ has quit IRC14:39
*** ebalduf has joined #openstack-cinder14:40
*** kvidvans has joined #openstack-cinder14:40
*** s0nea has quit IRC14:40
*** ebalduf has quit IRC14:40
*** ebalduf has joined #openstack-cinder14:43
*** jungleboyj has joined #openstack-cinder14:46
*** merooney has quit IRC14:47
*** jistr has joined #openstack-cinder14:47
*** cbader has quit IRC14:52
*** merooney has joined #openstack-cinder14:53
*** buster_604 has joined #openstack-cinder14:58
jgriffithmtreinish: any idea what this is about:
jgriffithmtreinish: I mean othe rthan "Length too long"14:58
jgriffithof course14:58
*** esker has quit IRC14:59
*** s0nea has joined #openstack-cinder14:59
*** annegentle has quit IRC14:59
*** esker has joined #openstack-cinder15:00
mtreinishjgriffith: my guess is that this attachment: is so long that it's causing an issue with the 1to2 filter15:01
mtreinishbut I've never seen that happen before15:01
*** buster_604 has quit IRC15:01
jgriffithmtreinish: I'm trying to figure out the coorelation between that and cranking up the number of processes in the test15:01
mtreinishit's probably a bug in 1to2, because I thought subunit should be able to chunk things properly for file attachments15:01
*** emagana has joined #openstack-cinder15:02
jgriffithmtreinish: I wonder...15:02
*** buster_604 has joined #openstack-cinder15:02
mtreinishit definitely seems like that test is loading all the migrations multiple times15:02
jgriffithmtreinish: yeah, I think that might be a problem15:03
mtreinishway more time then whatever the concurrency count is though15:03
*** buster_604 has quit IRC15:04
*** cbader has joined #openstack-cinder15:05
aarefievjgriffith: hi, have a min?15:06
mtreinishjgriffith: man, that log file is hanging up my browser15:06
mtreinishbut it didn't seem that big15:07
openstackgerritxing-yang proposed openstack/cinder: Not to deduct thin volume size in capacity filter
DuncanTtbarron: +2/A15:12
*** mtanino has joined #openstack-cinder15:12
*** ebalduf has quit IRC15:13
tbarronDuncanT: cool15:13
*** hemnafk is now known as hemna15:14
mtreinishjgriffith: oh, figured it out I had zuul.o.o open in another tab that was causing the hangs15:15
*** dims_ has joined #openstack-cinder15:16
*** dannywilson has joined #openstack-cinder15:18
*** jms has joined #openstack-cinder15:19
openstackgerritxing-yang proposed openstack/cinder: Fix capacity filter to allow oversubscription
*** dannywilson has quit IRC15:23
*** jordanP has quit IRC15:23
*** lpetrut1 has quit IRC15:24
*** jdurgin has joined #openstack-cinder15:24
*** dannywilson has joined #openstack-cinder15:25
*** agarciam has quit IRC15:25
*** nihilifer has quit IRC15:27
*** annegentle has joined #openstack-cinder15:30
*** tsekiyama has joined #openstack-cinder15:31
*** rmesta has joined #openstack-cinder15:36
jgriffithsorry... had a visitor15:38
jgriffithWTF?  "class TestSqliteMigrations...: pass"15:39
*** ebalduf has joined #openstack-cinder15:39
jgriffithnice.. the mixin bandit strikes again.  sigh15:40
jgriffithmtreinish: so I think that kinda screws up the parallelism15:41
mtaninoxyang1: Hi, I and Tomoki joined your session last night, it was very helpful to understand current and feature plan of Cinder :)15:41
xyang1mtanino: Were you there? I didn't see you.  Thanks for joining15:42
*** annegentle has quit IRC15:42
xyang1mtanino: I didn't know you are in Boston area15:42
mtaninoxyang1: we joined at little late15:43
*** annegentle has joined #openstack-cinder15:43
xyang1mtanino: Ok15:43
mtaninoxyang1: yes, we live in Boston area :) nice place15:43
mtaninoxyang1: Are you still in Boston?15:44
mtreinishjgriffith: hmm, how does nova and others do the migration testing in parallel?15:44
xyang1mtanino: Yes, I am in Boston area15:44
mtreinishI haven't looked at the cinder version of those tests, but they work fine in parallel for glance and subunit2sql (where I've added them)15:44
*** dansmith is now known as superdan15:44
* mtreinish pulls up the code15:44
mtaninoxyang1: I hope you will enjoy Boston sightseeing :)15:44
xyang1mtanino: I live here15:45
mtaninoxyang1: Oh, I didn't know it15:45
jgriffithmtreinish: yeah, doesn't seem that's the problem15:46
xyang1mtanino: my office is in Hopkinton15:46
jgriffithmtreinish: weird, and really weird I can't reproduce it locally either15:46
*** jdurgin has quit IRC15:46
xyang1mtanino: Where is your office15:46
mtaninoxyang1: our office is in Westford :)15:46
xyang1mtanino: Ok15:47
mtaninoxyang1: I know there is IBM's office along the route 495, but I didn't know also EMC's office is there15:49
*** leeantho has joined #openstack-cinder15:49
xyang1mtanino: EMC has several buildings on south street in Hopkinton15:49
*** dims_ is now known as dimsum__15:49
xyang1mtanino: That is the headquarter15:49
mtaninoxyang1: oh, I see15:50
xyang1mtanino: There is also an office in Cambridge15:50
mtaninoxyang1: good place :)15:50
*** daneyon has joined #openstack-cinder15:50
xyang1mtanino: Though I don't like to go there, can't find parking:(15:50
mtaninoxyang1: yup. I agree on it :(15:51
*** lpetrut has joined #openstack-cinder15:51
mtaninoxyang1: If you are OK, can you share the presentation slide yesterday?15:52
mtaninothat's very helpful :)15:52
xyang1mtanino: Sure, I'll ask the organizer where to publish slides15:52
mtaninoxyang1: thanks :)15:53
*** daneyon_ has joined #openstack-cinder15:53
xyang1mtanino: Although I believe you should already know everything:).  Nothing new there15:53
mtreinishjgriffith: oh if it's the db tests causing the issues, do you have the backends configured locally to reproduce?15:53
mtreinishIIRC it'll silently pass (may not even be a skip) if you don't have the backend db available15:54
mtaninoxyang1: I found my SPEC for liberty was in your slide :) I'm glad about it.15:54
jgriffithmtreinish: OH, crud... I forgot we changed all that for db mmigrations15:54
*** asselin_ has quit IRC15:54
jgriffithmtreinish: no, I'm just running the default sqlite15:54
xyang1mtanino: You should have anwsered questions for me:).  Did you see my disclaimer?15:55
jgriffithmtreinish: but one other thing that bothers me15:55
*** daneyon has quit IRC15:55
jgriffithmtreinish: grep out on LOG in the cinder migrate directory :(15:56
mtaninoxyang1: yes :)15:56
mtreinishjgriffith: I think there is a bug in the run script for unit tests, it doesn't need to use subunit-1to2 it can just do testr last --subunit (which is what we do on tempest jobs)15:56
mtaninoxyang1: These feares are not promised to get in Liberty :(15:56
jgriffithmtreinish: oh... let me have a look at that.  My run of subunit just hung it appears15:57
mtaninojust examples!15:57
*** timcl has quit IRC15:57
xyang1mtanino: I was just showing things that are already public, but   can't guarantee anything:)15:57
mtaninoxyang1: sure :)15:57
jgriffithmtreinish: well, that seems to work a bit better15:58
jgriffithOf course I'm not sure the output is whats wanted :)15:58
*** jistr has quit IRC15:58
mtaninoxyang1: Everyone can't guarantee what's happen in next 6 months.15:58
*** theanalyst has quit IRC15:58
xyang1mtanino: :)15:58
*** rushil has quit IRC15:58
*** garthb has joined #openstack-cinder15:58
*** linkedinyou has quit IRC15:59
jgriffithmtreinish: can we start a new repo?  Name it "infra-things-griffith-might-look-for"15:59
jgriffithmtreinish: and folks can put all the stuff I might be interested in there :)15:59
mtreinishjgriffith: :)16:01
*** annegentle has quit IRC16:01
jgriffithmtreinish: you did it again!!!16:01
mtreinishjgriffith: heh, fwiw I'd be perfectly fine renaming os-testr to things-griffith-might-look-for :)16:01
jgriffithmtreinish: one of these days I'm going to fix something for you!!!16:01
*** theanalyst has joined #openstack-cinder16:02
jgriffithmtreinish: that's next on my list... I promise you16:02
jgriffithmtreinish: and with the links you provided it shouldn't be hard to find the time to do it16:02
mtreinishheh, there isn't any rush, unless you want to beat me to being first to use it somewhere16:03
jgriffithmtreinish: Challenge Accepted!!!16:04
mtreinishha, well I'll have to start bribing people for reviews now :)16:05
*** jistr has joined #openstack-cinder16:06
*** krtaylor has quit IRC16:08
*** jistr has quit IRC16:09
jgriffithmtreinish: testing an ostestr patch now16:09
jgriffithmtreinish: I think I'm going to remove all that debug nonsense in the migration tests16:12
jgriffithIt's REALLY annoying16:12
mtreinishheh, yeah it is I think that's the default for running the migrate commands16:14
jgriffithmtreinish: so the kunundrum is... I have the ostestr change, but it opens the limiter, which leads me back to the stability issues16:14
*** rwsu has joined #openstack-cinder16:15
jgriffithmtreinish: I'm thinking maybe set a default in tox for 6 or nproc whichever is less16:15
jgriffithmtreinish: in .tox.ini that is16:15
jgriffithmtreinish: since 6 seems to be a magic number (without the test_gluster changes)16:15
mtreinishhmm, yeah if there is an issue with too much concurrency you'll have to set a default ceiling somewhere16:16
mtreinishI guess pretty_tox made that easy to do16:16
jgriffithmtreinish: hehe... if I just remove the I can run 12 processes reliably no issues :)16:18
jgriffithmtreinish: ok, I'm going to stop blabbering and combine the two patches16:18
mtreinishjgriffith: heh, seems like a viable solution to me :)16:18
*** markus_z has quit IRC16:22
openstackgerritDiem Tran proposed openstack/cinder: Re-integrate Oracle iSCSI Cinder driver
*** timcl has joined #openstack-cinder16:28
*** Yogi1 has quit IRC16:31
*** rmesta has quit IRC16:31
*** rmesta has joined #openstack-cinder16:31
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** _cjones_ has joined #openstack-cinder16:32
dhellmannthingee: I would like to cut a release of os-brick on Monday to unblock the work on dropping the oslo namespace packages. The changes that will be released are listed in -- Does that work for you?16:34
thingeedhellmann: I just did this for infra last week. :(16:36
dhellmannthingee: ?16:36
dhellmannthingee: I need "9247aa4 2015-04-28 22:25:57 +0000 Drop use of 'oslo' namespace package"16:36
thingeeI had to do a release because of some requirement blockage16:36
thingeeJust saying some coordination would be good so I don’t have to do this once a week.16:37
*** esker has quit IRC16:38
dhellmannthingee: well, I'm offering to do this one for you, I just want to make sure I'm not going to break anything by doing it16:38
dhellmannI don't know what the requirements thing was last week, did that have to do with pbr?16:39
thingeeNah it’s fine. I’m just raising this will be the second week in a row I’m doing this, just because of dependencies.16:40
*** lpetrut has quit IRC16:40
jgriffithFWIW it's a new project, it's going to be that way for a while likely16:42
*** jungleboyj has quit IRC16:43
*** ociuhandu has joined #openstack-cinder16:44
openstackgerritxing-yang proposed openstack/cinder: Fix capacity filter to allow oversubscription
*** Tross has quit IRC16:51
*** rushil has joined #openstack-cinder16:52
openstackgerritJon Bernard proposed openstack/os-brick: Add missing connectors to factory test
*** vilobhmm has joined #openstack-cinder16:53
*** patrickeast has joined #openstack-cinder16:56
nikeshmasselin: hi16:56
asselinnikeshm, hi16:56
nikeshmasselin: good morning,good comment regarding new lines,16:56
*** raildo is now known as needmoresummit16:59
*** Apoorva has joined #openstack-cinder16:59
nikeshmasselin:but not mentioned in,how to do this?17:01
asselinnikeshm, I don't know either...mine just worked.17:02
*** xyang1 has quit IRC17:02
asselinnikeshm, however when you find the answer, we can update that url17:02
nikeshmhemna: hi Ci is now showing on  thanks for you and rammy for all support17:04
*** daneyon_ has quit IRC17:05
*** dannywilson has quit IRC17:08
*** dannywilson has joined #openstack-cinder17:09
SwansonAnother day another devstack error.  (sigh)17:10
jgriffithSwanson: what problem are you seeing?17:11
*** patrickeast has quit IRC17:11
*** patrickeast has joined #openstack-cinder17:11
jgriffithSwanson: odd, I've been going through logs the past few days you've been talking about devstack issues and I'm not seeing anything?17:11
*** jdurgin has joined #openstack-cinder17:11
jgriffithSwanson: deploying devstack on average 15x per day17:12
jgriffithSwanson: and then there's the gate, which is always a pretty good indicator17:12
jgriffithSwanson: git timeouts?17:12
Swansonjgriffith: this time it appears to be an oslo_config duplicate opt error today.  That usually requires extra ajax to clean up but is usually my machine.17:12
jgriffithSwanson: are you "reusing" machines?17:13
jgriffithSwanson: yeah... don't do that :)17:13
openstackgerritJon Bernard proposed openstack/os-brick: Fix local connector test case inheritance
jgriffithSwanson: sersiously... your life would be soooo much easier17:13
jgriffithSwanson: get on board with the whole "cloud" philosophy17:13
jgriffithSwanson: make that crap disposable after use17:13
*** rushil has quit IRC17:13
SwansonI think that is about to happen.17:14
jgriffithSwanson: :)17:14
SwansonTypically hasn't been an issue, tho.17:14
*** crose has quit IRC17:15
*** crose has joined #openstack-cinder17:15
jgriffithSwanson: so I run things in OpenStack for "EVER" (one DB Instance has been running for a year now)17:15
jgriffithSwanson: but when it comes to things like deploying/redeploying devstack, it's not really meant to work that way17:15
jgriffithSwanson: unless you want to contribute some uber code to the devstack cleanup.sh17:16
hemnaI restack frequently17:16
hemnait's usually worked for me17:16
jgriffithhemna: and you frequently ask in channel "anybody else seeing problems with devstack today" :)17:16
hemnaunless I go long periods w/o updating devstack itself17:16
jgriffithjust saying17:16
hemnahehe yah17:16
jgriffithhemna: Swanson ok, I don't know what I'm talking about, do whatever you like17:17
Swansonjgriffith: Fair enough.  I think I'll keep restacking and then the second it fails me blow it away.17:17
jgriffithjust trying to offer a suggestion17:17
hemnait's super easy for me to just spin up a new vagrant with the latest17:17
hemnaand toss the old one17:17
*** krtaylor has joined #openstack-cinder17:17
jgriffithNo matter what I say though I think someone here will contradict it17:17
hemnaI do all my dev in vagrant VMs now anyway to combat those exact kinds of problems.17:17
* jgriffith says "the sky is blue"17:17
SwansonShould never have abdicated the throne.17:17
jgriffithsomeone will say "nope, it's azure"17:18
jgriffithSwanson: ?17:18
hemnaI'll typically try to fix it for a few minutes, then put and vagrant up a new vm17:18
dhellmannthingee: ah, well, in this case it's actually a fix inside your lib that I need in order to be able to make some releases in oslo, so it's not quite the same thing17:18
Swansonjgriffith: like a joke only different at the end.17:19
SwansonIt was17:19
jgriffithSwanson: oh... sorry, I didn't understand17:19
*** rushil has joined #openstack-cinder17:19
jgriffithSwanson: but I'll laugh anyway :)17:20
asselinpatrickeast, fyi, i'm using fc passthrough from upstream now!17:20
Swansonjgriffith:  Typically it is just best to laugh and then sidle away from me.  That's what smcginnis does.17:20
jgriffithSwanson: LOL.. now that joke I got :)17:20
jgriffithSwanson: that also means you and I have something in common :)17:21
*** buster_604 has joined #openstack-cinder17:22
*** buster_604 has quit IRC17:23
*** buster_604 has joined #openstack-cinder17:23
*** vivek_nandavanam has quit IRC17:23
*** barra204 has quit IRC17:26
*** amoturi has joined #openstack-cinder17:27
*** Tross has joined #openstack-cinder17:28
thingeealright, so looks like no one opposed to smcginnis joining core. welcome smcginnis :)17:30
hemnasmcginnis, +2!17:30
*** aix has quit IRC17:31
SwansonAre you sure he is allowed in?  I mean he took vacation somewhere without internet..17:32
* thingee needs to figure out how he's doing burning man this year with the release happening sep 317:32
*** madskier has joined #openstack-cinder17:32
*** Yogi1 has joined #openstack-cinder17:34
*** amoturi has quit IRC17:35
nikeshmhemna : morning17:35
*** ebalduf has quit IRC17:36
nikeshmhemna: now CI is showing with all tests paased on
*** madskier has quit IRC17:37
nikeshmhemna: thanks for all your and rammy support17:37
*** amoturi has joined #openstack-cinder17:37
hemnathingee, what's a vacation?17:38
hemnaI'm told an openstack summit is the same thing17:38
nikeshmhemna: can i expect to remove -2 :)17:40
*** ronis has joined #openstack-cinder17:40
hemnaso the CI logs are completely unreadable :(17:40
hemnano line breaks17:41
hemnathe logs are 1 long line :)17:41
hemnalooks like newlines are stripped ?17:42
*** sgotliv has joined #openstack-cinder17:42
*** ronis has quit IRC17:42
*** sgotliv has quit IRC17:42
*** sgotliv has joined #openstack-cinder17:42
nikeshmhemna: yes i will look into that,talk to rammy but he said he also donot know17:43
nikeshmhow to do this17:43
uberjaythingee: a campmate completed and submitted his med-school application in the first 3 days on playa (note: this did not look like a fun time)17:44
asselinnikeshm, how did you setup your log server?17:44
nikeshmasselin: i setup manually17:44
asselinnikeshm, try setting up a new one with my scripts. might fix it for you17:44
nikeshmasselin: sure17:45
*** annegentle has joined #openstack-cinder17:45
asselininstructions are int he readme17:45
asselinnikeshm, then you log server is guarenteed the same as mine & upstream17:45
*** simondodsley has joined #openstack-cinder17:45
*** jungleboyj has joined #openstack-cinder17:46
uberjayasselin: nikeshm that worked for me -- one gotcha were the assumptions about how your jenkins job is named (because of the directory name). this only controls what readme html gets attached to a particular directory, so it's not a huge deal, obviously.17:48
uberjaye.g.,   <Directory /srv/static/logs/*/*/*/*/*-tempest-dsvm*/*>17:49
uberjay     ReadmeName /help/tempest-overview.html17:49
uberjay  </Directory>17:49
*** xyang1 has joined #openstack-cinder17:50
asselinuberjay, what is your directory structure like?17:50
uberjaymy problem was the jenkins job name didn't match (*-tempest-dsvm*) originally, but I've since renamed it.17:52
uberjayi had a confusing moment wondering why my CI logs didn't look like other CI logs, is all. :)17:53
asselinuberjay, ok interesting...perhaps an upstream patch to customize that would be nice to have. thanks for mentioning it17:53
asselinuberjay, that just affects the readme right?17:53
uberjayasselin: all of the work for unzipping and annotating the log files themselves still worked, the readme footers were just not attached to each directory17:54
*** vokt has joined #openstack-cinder17:57
*** ebalduf has joined #openstack-cinder18:00
*** ebalduf has quit IRC18:00
*** ebalduf has joined #openstack-cinder18:00
*** ebalduf has quit IRC18:01
*** amoturi has quit IRC18:04
nikeshmasselin: thanks working on it,btw whats your work location :)18:05
asselinnikeshm, california18:05
nikeshmasselin: great, i am too in san jose,californina18:06
openstackgerritJohn Griffith proposed openstack/cinder: Remove pretty_tox and use ostestr
jgriffithmtreinish: ^^ I think I finally beat you on one!!!18:08
jgriffithmtreinish: going to celebrate by having lunch outside ;)18:08
*** annegentle has quit IRC18:08
*** barra204 has joined #openstack-cinder18:08
*** annegentle has joined #openstack-cinder18:09
*** lpetrut has joined #openstack-cinder18:12
uberjayhemna: thanks for taking a look at our driver. as for the question about requests -- yes, I would actually prefer to use it over httplib. when I first looked into it, i incorrectly determined requests wasn't necessarily available. I'd be happy to change it over if you think it's worth doing before merging.18:14
*** barra204 has quit IRC18:15
openstackgerritEarle F. Philhower, III proposed openstack/cinder: Implement Cinder Volume driver for HGST Solutions
uberjayhemna: the utilization question... I'm going to have to talk with some people to see if we can advertise a meaningful value. i was worried about depending on that for L1, and instead have it be an incremental change (hopefully for L1, or soon thereafter)18:17
*** sgotliv has quit IRC18:18
*** barra204 has joined #openstack-cinder18:20
*** rushiagr_away is now known as rushiagr18:25
*** alexpilotti has joined #openstack-cinder18:26
*** barra204_ has joined #openstack-cinder18:28
dhellmannthingee: thinking about what you said, was 0.1.0 the first release of os-brick? if you needed to release something to fix requirements, it looks like there is only one release now, so I'm confused18:29
*** rasoto has quit IRC18:29
*** ebalduf has joined #openstack-cinder18:30
thingeedhellmann: yes, 0.1.0 was the first. we're on 0.1.118:30
dhellmannhmm, I don't see that tag in my repo18:30
*** barra204 has quit IRC18:31
dhellmannyeah, I see. I wonder why I don't have the tag18:31
dhellmannok, a fresh checkout brings it in18:32
*** patrickeast has quit IRC18:32
*** ebalduf has quit IRC18:34
*** annegentle has quit IRC18:41
*** patrickeast has joined #openstack-cinder18:42
*** patrickeast has quit IRC18:43
nikeshmasselin: tried to set up log server as your scripts
nikeshmasselin: i got this in the end18:51
asselinnikeshm, strange...seems network issue or the server's down?18:51
nikeshmbut able to ping that18:52
asselintry running it again18:52
*** barra204_ is now known as shakamunyi18:52
openstackgerritxing-yang proposed openstack/cinder: Fix capacity filter to allow oversubscription
*** fthiagogv has quit IRC18:55
*** vokt has quit IRC18:55
*** buster_604 has quit IRC18:58
*** rushiagr is now known as rushiagr_away18:58
nikeshmasselin: ok by mistake i gave private key now i changed it to public key
nikeshmin domain i gave public ip of my log server18:59
openstackgerritJohn Griffith proposed openstack/cinder: Add volume ID to fake volumes in Gluster tests
*** aix has joined #openstack-cinder19:00
asselinnikeshm, yeah...domain can be anything you like19:00
asselinnikeshm, I haven't seen it breaking anything19:01
hemnauberjay, it's fine to get this one in with 'unknown', I just wanted to make you aware of the ramifications of it19:03
mtreinishjgriffith: couldn't you just have rebased when you pushed up the gluster patch instead of abandoning?19:06
openstackgerritJohn Griffith proposed openstack/cinder: Remove pretty_tox and use ostestr
jgriffithmtreinish: yeah, but every time I do that I screw something up19:07
mtreinishheh, ok19:07
hemnauberjay, so requests is available fwiw.19:07
jgriffithmtreinish: I don't know why, I just can't get that workflow down19:08
jgriffithguess I don't do it enough19:08
mtreinishjgriffith: I normally just do a 'git review -x review_id' with the new patch as head19:08
*** crose has quit IRC19:08
mtreinishthat'll cherry-pick the review ontop19:08
jgriffithmtreinish: Oh... son of a19:08
jgriffithmtreinish: I was making it way harder than that19:08
jgriffithmtreinish: I always try and cherry-pick, merge etc19:09
uberjayhemna: ok, yeah, that makes sense. thanks! i think i'd rather get the base driver in, rather than hold it up for what will end up being a very small driver change when we sort out the backend details. i'm adding a task for the requests changeover, too.19:09
uberjayhemna: obviously, using requests would be nicer :)19:09
hemnayah you can switch to requests in a follow up19:09
hemnaas well as change the reporting stats19:09
hemnano big.19:09
*** ebalduf has joined #openstack-cinder19:10
uberjayok, great :)19:11
jgriffithmtreinish: the infra manual docs even spell it out for me :(19:11
jgriffithmtreinish: ok, filed that one away19:11
mtreinishjgriffith: heh, that's better than what I used to do when I first started using git. I'd use format patch and then git am to do a ghetto rebase19:11
jgriffithmtreinish: :)19:12
jgriffithsounds like my git-diff patch apply method :)19:12
jgriffithmtreinish: problem for me is "old habbits" die REALLY hard19:12
*** Yogi1 has quit IRC19:22
*** alexpilotti has quit IRC19:23
openstackgerritJohn Griffith proposed openstack/cinder: Remove pretty_tox and use ostestr
openstackgerritJohn Griffith proposed openstack/cinder: Add volume ID to fake volumes in Gluster tests
nikeshmasselin: done,but do we have to copy logs in /srv/static/logs/19:37
*** jungleboyj has quit IRC19:39
openstackgerritJohn Griffith proposed openstack/cinder: Remove pretty_tox and use ostestr
*** dannywilson has quit IRC19:44
*** dannywilson has joined #openstack-cinder19:48
*** patrickeast has joined #openstack-cinder19:48
nikeshmasselin:now on clicking on log,its giving "file not found"19:50
mtaninoDuncanT: Hi,19:54
*** IlyaG has joined #openstack-cinder19:57
openstackgerritxing-yang proposed openstack/cinder-specs: Non Disruptive Backup
*** dannywilson has quit IRC20:00
*** kvidvans has quit IRC20:01
*** timcl has quit IRC20:02
*** bkopilov has quit IRC20:04
*** openstack has joined #openstack-cinder20:06
uberjayhemna: is there anything else you think I should look at?20:08
openstackgerritTom Swanson proposed openstack/cinder: Dell: Added verify cert option for REST calls
*** merooney has quit IRC20:18
*** Lee1092 has quit IRC20:21
*** Apoorva has joined #openstack-cinder20:22
asselinnikeshm, need more details20:22
*** openstack has joined #openstack-cinder20:31
nikeshmaaselin: now its working20:34
nikeshmcan yu click on log link in and remove -1 :)20:34
*** needmoresummit is now known as raildo20:38
nikeshmhemna: done now please check20:39
asselinnikeshm, looks good, thank you20:39
hemnaok I'll review it now :)20:40
*** alexpilotti has joined #openstack-cinder20:41
asselinnikeshm, next is fc20:41
nikeshmasselin: sure20:42
*** annegentle has joined #openstack-cinder20:42
*** ndipanov has quit IRC20:43
*** rushil has quit IRC20:43
*** dannywilson has joined #openstack-cinder20:47
*** jungleboyj has joined #openstack-cinder20:47
nikeshmhemna: thanks,can we merge it?20:48
Swansonjgriffith: So, that starting over thing seems to, um, work.  Thanks!20:48
hemnanikeshm, so your patch contains 2 drivers20:51
hemnabut I think CI is only reporting 1 no ?20:51
hemnayou'll need to report both iscsi and FC results20:51
hemnasince there are 2 drivers20:51
asselinnikeshm, another possibility is to split your commit into two: one for each driver. But since you still have time to merge, just do the fc. You're almost there.20:52
nikeshmheman: sure20:52
nikeshmasselin: sure20:52
*** openstackgerrit has quit IRC20:59
*** openstackgerrit has joined #openstack-cinder21:00
*** raildo has quit IRC21:00
openstackgerritMike Perez proposed openstack/cinder-specs: Standard Capabilities
hemnaharlowja, ping21:02
harlowjahemna sup21:04
harlowjayo yo21:04
hemnajust curious why we do a local file lock around the create flow ?21:05
hemnajust trying to understand it21:05
hemnalooks like we are dynamically creating the lock name based upon some create params21:05
hemnato lock out the deletion of snapshots/volumes ?21:06
harlowjaseems like thats why it was added21:06
hemnaso to prevent someone from deleting a snapshot while we create a volume from it ?21:06
harlowjaseems like that one might explain why21:07
hemnaick ok21:07
*** julim has quit IRC21:07
hemnayet another thing we have to fix21:08
hemnaor HA won't ever work21:08
hemnait seems like we keep baking in HA problems21:08
harlowjalocks at any level do that :-/21:08
harlowjawell not any level, but most levels21:08
hemnawell these are local file locks21:08
hemnadistributed locking is bad, but doesn't preclude HA21:09
harlowjai've seen
harlowjawhich seems to try to switch those to distributed locking21:09
harlowjaseems to be actively being workedo n21:10
hemnaah yah ok21:10
harlowjaanother option/thought21:10
*** bswartz has quit IRC21:10
harlowjaironic has (which taskflow also has something like)21:11
harlowjathey i think use that before/after each operation on a volume21:11
harlowja*on a BM i mean21:11
hemnayou said tasflow21:11
hemnathat's still ok in here isn't it? :P21:11
harlowjau said tasflow21:11
harlowjaso i think that spelling is ok?21:11
harlowjathats a different secret project21:11
harlowjaso the other option is to do something like that, pass that state-machine around in cinder context (?) and transition on it21:12
*** simondodsley has quit IRC21:12
hemnaso I guess this locking is sort of special though21:12
hemnaas the state change needs to happen on the parent object21:12
harlowjathen u have a (and a nice picture to)21:12
hemnato prevent actions happening on them21:12
harlowjawell its on the volume right?21:12
harlowjajust like the ironic one i think is on the BM21:12
hemnathe locks are being placed on the parent snapshot21:13
hemnaand or parent volume21:13
hemnato prevent them from being deleted21:13
hemnawhile the new volume is being created from them.21:13
hemnathat's the tricky sticky part21:13
harlowjaright, gotcha21:13
hemnalife is complicated21:14
hemnaok thanks21:14
hemnasorry to bug ya21:14
harlowjaparent_volume.transition(SNAPPING) and then do stuff21:14
harlowja* (which i think is the ironic code to do similar stuffs)21:14
* hemna wonders if some of this magicness could be handled by objectifying everything21:14
harlowjameh, no magic, just requires work :-P21:14
hemnaturning the volume dicts into objects that can have their internal states and parents touched automatically21:15
harlowjamagic not usually solve much of anything, ha21:15
harlowjaunless its the good kind, ha21:15
harlowjait might be intersting to poke the ironic folks, i only know a little bit about there state machien and how they use it21:15
* harlowja be interested to know21:15
harlowjamaybe its applicable to u guys to21:15
hemnacould be21:16
hemnait's a big can to open....21:16
harlowjai know the code that is the state-machine (since i made it), but not exactly sure how they use it21:16
hemnaI'm not sure I'm the right guy to do that21:16
harlowjaengage can opener!21:16
hemnabut it's a big problem for Cinder that we need some sort of plan to finally address it21:16
*** annegentle has quit IRC21:17
harlowjaso there are a few avenues, imho the explict state-machine is the best21:17
harlowjadistributed locks will help, but will imho be hard to figure out when they fail21:17
harlowjaand they will fail, lol21:17
harlowjaand they fail differently than local file locks21:17
harlowja*local file locks don't have servers diseapper, or expiries...21:18
hemnaand those failures happen in very difficult scenarios to reproduce21:18
hemnamakes my brain hurt on a Friday21:18
harlowjathats why i was almost thinking, just having top-level state machine that is very strict at the API level might help, do big transitions and deny anything else21:20
hemnawe do a lot of this type of checking deep down in the manager21:21
hemnawe should be doing more checks in the API and denying there.21:21
harlowjavolume Z being created, do little stuff in init, transition parent_snapshot X, parent_volume Y to things that can not be acted on until that Z is done21:21
harlowjaor if X or Y can not be transitioned, stop21:22
harlowjaand as final final action of Z being created, move X, Y back to 'useable' or whatever21:22
harlowjasomething like that21:22
harlowjathen as time goes on, releax that21:22
hemnanova shits all over that now though21:22
harlowjasucks for nova, lol21:22
hemnanova expects stuff to just always work or timeout21:22
harlowjashould learn to put its diapears on21:22
hemnaheh yah21:23
*** jungleboyj has quit IRC21:23
hemnaI dunno21:23
hemnaI think it's ok for Cinder to just simply say....sorry I'm busy right now.21:23
harlowjaso u said timeout, ok, turn VolumeIsBusyException -> VolumeTimedOutWaitingtoBecomeUnBusyException21:23
harlowjaif u can make the exception name longer, that'd be cool21:23
hemnaif only we could put ! in a name21:24
hemnayah, really busy!21:24
harlowjahemna i'd agree, i feel its easier to be strict, and later relax that strictness as u want, vs starting the other way around21:24
hemnaok that's more like it on a Friday.21:25
harlowjaonce is finally all good to go, someone could think about that state-machine path to21:25
* harlowja is waiting on a few other reviews for that21:26
harlowjaits another possiblity...21:26
harlowjau can not escape me cinder!!!!21:26
harlowja* (waiting on)21:26
openstackgerritYucong Feng proposed openstack/cinder-specs: Introducing parallel zoning
*** shakamunyi has quit IRC21:26
harlowjaand we could figure out how its worked out for ironic21:27
harlowjadev does mention it in his emails, and stuff, and operators seem to like it, ha21:27
hemnanew lib ?21:28
hemnaoh sorry21:28
*** krtaylor has quit IRC21:28
*** lpetrut has left #openstack-cinder21:28
harlowjaya, ironic copied out the state-machine from taskflow that i wrote, into there tree21:28
harlowjaand then i was like, eck, not gonna work out long-term21:28
harlowjasooo, waited until they had stabalized a little, and were still using it21:29
harlowjaand make new oslo library (really general purpose lib)21:29
harlowjaand will eliminate both copies with that21:29
*** lcurtis has joined #openstack-cinder21:29
hemnaheh, I like your examples21:30
*** openstack has joined #openstack-cinder21:32
hemnato get those checks happening up front21:32
hemnaalmost like a  "can I do this thing I'm about to do on this object?"21:32
harlowjaya, should be possible to make them appear on the object itself21:32
hemnayes = go to volume manager and do it21:32
* harlowja was thinking about having tasks (in tasflow) also have that same method ('transition') or something21:33
*** primechuck has quit IRC21:33
harlowjaright now its an external function call to do it21:33
hemnano = ResourceIsBusy or something along those lines21:33
hemnaask permission to do it21:33
harlowjahemna well the manager needs to do stuff for async things though right?21:33
hemnaget approval, then go ahead21:33
harlowjaif all api's were blocking, then i guess u could21:33
hemnaif mom says no, then report no up the chain21:33
harlowjaright, ask all the mom's that are needed to ask21:34
harlowja*grandmom, mom...21:34
harlowjaand then transition, and then do stuff21:34
harlowjaand then transition mom and grandmom when done (async or other)21:34
hemnawhich asks the moms21:35
hemnaor reports back failures21:35
hemnaand then call manager to do the action21:35
*** jungleboyj has joined #openstack-cinder21:35
hemnaand then volume.end_action21:35
hemnaprobably too simplistic21:35
harlowjaya, something like that i think21:35
hemnaI do like the idea of encapsulating that logic into the volume object itself21:36
hemnathen anyone can ask those same transitions21:36
harlowjaor volume.transition('useable', 'clone')21:36
harlowjaor hiding that, idk21:36
hemnaI think the api into the volume object should be simple21:37
hemnaand sort of hide the transitions states21:37
harlowjaok dokie, thats fine21:37
hemnaso the caller on the volume doesn't have to know any smarts about it21:37
hemnabut simple asks, hey, can I clone you ?21:37
hemnaor, can I delete you ?21:37
harlowjasure, makes sense21:37
harlowjahide the complexity in start_action or whatever21:37
hemnainternals knows the current state, and locks it or whatnot21:38
hemnaat least it's all in 1 spot then21:38
hemnaright now we have so many checks all over the place21:38
harlowjaand just have each API call do volume.start_action() as its first thing, or die21:38
hemnait's damn near impossible to know what you can do on a volume now21:38
*** primechuck has joined #openstack-cinder21:38
harlowjaand a volume.end_action at the end21:38
harlowjaand thats all21:38
hemnaso reporting to Nova or any caller is easy21:39
hemnaand up front21:39
harlowjablock everything else imho, until u want to make it possible to run states at the same time (but meh)21:39
hemnaand for volume creation21:39
hemnathe start_action21:39
hemnacould put those distributed locks if needed on the parents21:39
*** merooney has joined #openstack-cinder21:40
hemnafor create from snap, or prevent deletion, etc21:40
hemnaand upon a failure21:40
harlowjajust transition the parents to?21:40
hemnawe might be able to get away with simple calling end_action on the same object to clean up21:40
hemnatransition and save inside of start_action21:40
harlowjaavoid locks, just move parents into states that can not be touched21:40
harlowjaleave grandma alone goddamnit21:41
harlowja*thats the message u can use21:41
hemnaand then the API calls into delete snapshot will puke because snapshot_start_action('delete') says no.21:41
hemnaon the parent(s)21:41
hemnamake it so21:41
harlowjahey, u make it so, lol21:42
harlowjavilobhmm make it so!21:42
vilobhmmstate machine would be a better choice  than distributed locks or any kind of locking isn't it21:42
hemnaI'll be drinking beer all weekend and fishing.21:42
*** primechuck has quit IRC21:42
harlowja  ( hemna u can revive that )21:42
harlowjaor combine it with automaton21:42
harlowjaorrrrr something, lol21:42
hemnathe trick is you still need some sort of locks to make the transitions atomic21:42
hemnaespecially in the case of cascading those transitions to parent(s)/children(s)21:42
hemnayah something....21:43
*** merooney has quit IRC21:43
harlowjafew options for that, select for, tooz...21:43
hemnaI'll forget all of this by Monday21:43
harlowjabut at least the lock stuff is in one place21:43
harlowja*and is around a tiny set of methods21:43
harlowjahemna good things there are loggs, lol21:43
vilobhmmstate machine validation at API level locking while doing compare and set at the db level….21:44
harlowjasomething like that21:44
harlowjaand profit!21:44
hemnaooh yah profit!21:44
vilobhmmstate machine validation at API level ; locking while doing compare and set at the db level….something of that sort was tried here disregard the micro-states part but the concept of "state machine validation at API level ; locking while doing compare and set at the db level"21:46
harlowjaso hemna ya, there u go :-P21:46
vilobhmmhemna : ^^21:46
hemnaanyway, I'm a big fan of encapsulation and concentrating code21:46
hemnaso it's not spread the hell around everywhere21:46
harlowjaget er' done21:46
*** annegentle has joined #openstack-cinder21:46
harlowjajust starting with the start_action and end_action and saying 'no' alot more than currently i'd think be a major win21:48
harlowjawould allow u to remove a bunch of locks21:48
harlowjasince the initial 'no' would make them not needed21:49
harlowjaand people that are trying to do things concurrently on the same volume right now, are probably messing themselves up right now anyway21:49
harlowja*even with locks21:49
*** IlyaG has quit IRC21:51
*** mriedem has quit IRC21:51
harlowjamy 2 cents :-P21:52
*** marcusvrn has quit IRC21:53
*** dims_ has joined #openstack-cinder21:53
*** akshai has quit IRC21:53
hemnaharlowja, yah I agree21:53
*** IlyaG_ has joined #openstack-cinder21:53
*** mdenny has quit IRC21:54
harlowjacools, let me know when thats finished and i'll review it, lol21:54
*** annashen has joined #openstack-cinder21:55
*** dimsum__ has quit IRC21:55
*** alexpilotti has quit IRC21:56
vilobhmmhemna : May be we can start here add the methods like start() end() and then invoke it whereever needed before doing a state validation ?21:57
vilobhmmsince you said you would prefer it doing in an object…and have the obj expose these methods21:57
hemnayah I'd prefer that21:58
hemnathat volume object is more of a DB abstraction object right?21:59
*** jungleboyj has quit IRC21:59
hemnahrmm yah I was thinking of something slightly higher I guess.22:00
*** jms has quit IRC22:00
harlowjaobjects <-> rpc <-> db orm <-> sql abstraction ?22:00
vilobhmmif you see
harlowjaits a layered sandwich of goodness, lol22:00
harlowjaobjects <-> rpc <-> db orm <-> sql abstraction <-> sql <-> some btrees ?22:00
hemnain my past life22:01
hemnaI had something similar22:01
hemnaI had a DB abstraction object22:01
harlowjaa sandwich of goodness?22:01
hemnathat knew how to do DB related things22:01
hemnaand another 'business logic' style of container object that knew how to take actions and do 'work' on the db abstracted objects22:01
harlowjabut it wasn't a seven layered burrito22:02
harlowjaand how did u survive without that22:02
vilobhmmburrito haha22:02
hemnano sour guacamole!22:02
harlowja oh yaaa22:02
hemnayah TB's guac is awful22:02
hemnamostly lemons22:02
hemnaand some green like liquid22:02
hemnastuffs of champs baby!22:03
harlowjaturns u into a man22:03
harlowja*and/or woman22:04
harlowjamight turn a man into a woman?22:04
hemnagirly man!22:04
vilobhmmok hemna….i think we have a clear plan (which I feel we do after a long time) I can help you on this effort…may be we can discuss this in cinder weekly on 06/03 ?22:04
harlowjaplan is to turn cinder into taco bell?22:04
hemnavilobhmm, well, it's an idea22:05
hemnaI'm not sure I'd go so far as calling it a plan :)22:05
hemnaneeds more beers22:05
vilobhmmok idea ....22:05
hemnait's a boatloads of change22:05
hemnato implement22:05
vilobhmmi know22:05
hemnaand affects nearly damn everything22:05
hemnalets boil the ocean and break the world.22:05
harlowjau can call it the pyramid sub-project22:05
hemnawho's ready?22:05
harlowja*pyramids built out of cinder bolks22:06
vilobhmmcount me in22:06
vilobhmmbut you have to do it someday …why not do it now :P22:06
harlowjawe got a lively one, he might need more 7-layer burritos22:06
hemnaharlowja, that might get us back into diapers though22:07
harlowjano might22:07
harlowjait will22:07
hemnakinda early in the weekend for that22:07
harlowjaok, on sunday then22:07
harlowjadiaper night22:07
harlowjasounds like a very awkward party :-/22:08
hemnayah, send invites to the managers and lets just go to the bar.22:08
harlowjaand someone deliver 100 7-layer burritos22:09
hemnayah!  party on!22:10
harlowjaparty will likely more be in those diapers22:11
*** annashen has quit IRC22:11
* harlowja runs away, lol22:11
*** cbader has quit IRC22:12
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** annegentle has quit IRC22:18
*** annegentle has joined #openstack-cinder22:20
*** annashen has joined #openstack-cinder22:27
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** IlyaG_ has quit IRC22:34
*** IlyaG has joined #openstack-cinder22:36
*** hemna is now known as hemnabeer22:36
*** krtaylor has joined #openstack-cinder22:36
*** aix has quit IRC22:39
*** annegentle has quit IRC22:40
*** madskier has joined #openstack-cinder22:46
*** boris-42 has quit IRC22:48
jgriffithSwanson: cool!22:49
jgriffithSwanson: your next hiccup will be clone errors :)22:49
*** ebalduf has quit IRC22:50
*** madskier has quit IRC22:50
jgriffithhemnabeer: harlowja That sounds like the simple object queue I drew up on the board in FTC last fall22:52
*** emagana has quit IRC22:55
*** emagana has joined #openstack-cinder22:55
nikeshmasselin: hi22:57
jgriffithvilobhmm: hey... did you see my response to your -1 on my patch for os-testr?22:58
jgriffithvilobhmm: seems your "concern" is completely unrelated to my change22:59
jgriffithvilobhmm: unless I'm missing something you're trying to point out22:59
vilobhmmjgriffith : the patch seems ok to me..just wanted to make sure we dont have any code duplication…23:05
vilobhmmi did reply now23:05
vilobhmmwas busy with meeting in afternoon sorry for late reply23:06
*** sgotliv has joined #openstack-cinder23:06
nikeshmasselin: trying to add in my job
vilobhmmjgriffith :
*** sgotliv has quit IRC23:07
asselinnikeshm, what's th eissue?23:08
*** annashen has quit IRC23:08
nikeshmasselin: after,manuall copying manually copied in /opt/nodepool-scripts/,but when running jon its saying script not found23:09
asselinnikeshm, are you using stackforge/third-party-ci-tools or something else?23:09
asselinI recommend you git clone that repo in your pre_test_hook and then you have the file23:10
asselinno need for /opt/nodepool-scripts/23:10
jgriffithvilobhmm: I get ya... but you gave it a -1 :(23:10
vilobhmmsorry :(23:11
vilobhmmmust have been a 0 or so…will remove it23:11
vilobhmmbcz the code changes look good to me apart fromt hat concern23:11
vilobhmmi mentioned that beforehand as weell23:12
jgriffithvilobhmm: right, so you should've given it a +1 :)23:12
jgriffithvilobhmm: if the change set looks good, then that seems like +1 :)23:12
vilobhmmcan't give a +1 like that :P had few concerns on it…23:12
*** ganso_ has quit IRC23:14
jgriffithvilobhmm: Hmm... ok23:15
jgriffithvilobhmm: well, whatever I guess23:15
jgriffithvilobhmm: thanks23:15
nikeshmasselin : good idea,and what about private key file of FC provider23:15
vilobhmmjgriffith : you have few min23:16
asselinnikeshm, get that on the vm any way you like23:16
*** severson_ has quit IRC23:24
*** rongze has joined #openstack-cinder23:26
*** rmesta has quit IRC23:30
*** lcurtis has quit IRC23:31
openstackgerritSean Chen proposed openstack/cinder: Tintri Cinder Volume driver
*** rongze has quit IRC23:36
*** rongze has joined #openstack-cinder23:37
*** jdurgin has quit IRC23:39
winston-djgriffith: ping23:39
*** rongze has quit IRC23:41
openstackgerritJohn Griffith proposed openstack/cinder: Use SolidFire snapshots for Cinder snapshots
jgriffithwinston-d: hey23:45
*** xyang1 has quit IRC23:52
jgriffithwinston-d: sorry I missed ya, I'm heading home.  Catch ya later?23:52
*** bswartz has joined #openstack-cinder23:53
*** vilobhmm has quit IRC23:53

Generated by 2.14.0 by Marius Gedminas - find it at!