Tuesday, 2015-11-17

*** emagana has quit IRC00:00
*** baoli has joined #openstack-nova00:00
*** haomaiwa_ has quit IRC00:01
*** burgerk_ has quit IRC00:01
*** haomaiwang has joined #openstack-nova00:01
melwittdansmith: looking at the cells tests, it looks like mox.IsA(objects.Instance) is expecting instance being passed to be an InstanceV1 but with the backport it's not?00:01
*** edtubill has joined #openstack-nova00:03
*** baoli has quit IRC00:05
*** xyang1 has quit IRC00:06
*** baoli has joined #openstack-nova00:06
*** edtubill has quit IRC00:08
*** pixelbeat has quit IRC00:10
*** baoli has quit IRC00:12
*** morgabra_ is now known as morgabra00:15
*** morgabra has quit IRC00:15
*** morgabra has joined #openstack-nova00:15
*** penick has quit IRC00:16
*** EinstCrazy has quit IRC00:19
*** annegentle has quit IRC00:22
*** annegentle has joined #openstack-nova00:24
*** penick has joined #openstack-nova00:26
dimsdansmith mriedem_away melwitt - o.vo break under control?00:26
dansmithdims: yes00:26
dimswhew thanks00:26
*** thorst_ has joined #openstack-nova00:29
*** yonglihe has quit IRC00:30
*** thorst_ has quit IRC00:30
*** thorst_ has joined #openstack-nova00:31
*** hemna is now known as hemnafk00:34
*** thorst_ has quit IRC00:35
*** pratikma_ has joined #openstack-nova00:35
*** pratikmallya has quit IRC00:35
*** agireud has quit IRC00:37
*** jerrygb has quit IRC00:40
*** jerrygb has joined #openstack-nova00:42
*** jerrygb has quit IRC00:42
*** otter768 has joined #openstack-nova00:44
*** annegentle has quit IRC00:47
*** penick has quit IRC00:47
*** annegentle has joined #openstack-nova00:47
*** otter768 has quit IRC00:48
*** mlavalle has quit IRC00:56
*** aginwala has joined #openstack-nova00:57
*** Yanyanhu has joined #openstack-nova00:59
*** haomaiwang has quit IRC01:01
*** haomaiwang has joined #openstack-nova01:01
*** baoli has joined #openstack-nova01:03
*** jyuso1 has joined #openstack-nova01:03
*** penick has joined #openstack-nova01:03
*** EinstCrazy has joined #openstack-nova01:04
*** davideagnello has quit IRC01:05
*** davideagnello has joined #openstack-nova01:06
*** salv-orlando has quit IRC01:12
*** penick has quit IRC01:13
*** smatzek has joined #openstack-nova01:13
*** rushil has joined #openstack-nova01:14
sdagueis there a pin up?01:15
dimssdague : a unit test fix is being merged01:16
openstackgerritSean Dague proposed openstack/nova: Removed extra spaces from double line strings  https://review.openstack.org/24298901:17
sdaguethat was the last nova patch ahead of it that wasn't already failing01:17
*** aginwala has quit IRC01:19
*** smatzek has quit IRC01:20
*** aginwala has joined #openstack-nova01:20
*** penick has joined #openstack-nova01:22
melwittdansmith: I left you a comment on the backport review that hopefully might be helpful01:25
*** aginwala has quit IRC01:25
*** sacharya has joined #openstack-nova01:25
*** penick has quit IRC01:27
openstackgerritZhenyu Zheng proposed openstack/nova: Split-network-plane-for-live-migration  https://review.openstack.org/24500501:28
*** thorst_ has joined #openstack-nova01:28
*** thorst_ has quit IRC01:30
*** rushil_ has joined #openstack-nova01:30
*** thorst_ has joined #openstack-nova01:30
*** ssurana1 has quit IRC01:32
*** rushil has quit IRC01:33
*** thorst_ has quit IRC01:35
*** aginwala has joined #openstack-nova01:36
*** zhenguo has joined #openstack-nova01:36
*** takashin has joined #openstack-nova01:36
*** pratikma_ has quit IRC01:36
*** markvoelker has joined #openstack-nova01:37
*** jdurgin1 has quit IRC01:41
*** annegentle has quit IRC01:41
*** jdurgin1 has joined #openstack-nova01:44
*** otter768 has joined #openstack-nova01:45
*** achanda has joined #openstack-nova01:46
*** suro-patz has quit IRC01:46
*** aix has quit IRC01:46
*** haomaiwang has quit IRC01:49
*** yangyapeng has joined #openstack-nova01:49
*** mriedem_away is now known as mriedem01:59
mriedemif i could travel back in time, i'd go back 10 years and thank myself for keeping copies of furniture receipts so i can use them when the cleaning service damages said furniture02:00
*** yonglihe has joined #openstack-nova02:02
*** mwagner has joined #openstack-nova02:03
*** eglynn_ has quit IRC02:05
*** jmccrory has joined #openstack-nova02:05
*** fawadkhaliq has joined #openstack-nova02:05
*** mriedem has quit IRC02:07
*** alejandrito has joined #openstack-nova02:08
*** alejandrito has quit IRC02:08
*** READ10 has joined #openstack-nova02:08
*** alejandrito has joined #openstack-nova02:08
*** otter768 has quit IRC02:10
*** pece has quit IRC02:10
*** baoli has quit IRC02:11
*** baoli has joined #openstack-nova02:11
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/24524802:14
*** sacharya has quit IRC02:16
*** alejandrito has quit IRC02:21
*** pratikmallya has joined #openstack-nova02:21
*** armax has joined #openstack-nova02:24
*** jwcroppe has quit IRC02:26
openstackgerritZhenyu Zheng proposed openstack/nova: Add project-id and user-id when list server-groups  https://review.openstack.org/20901802:26
*** leseb has quit IRC02:28
*** klkumar has joined #openstack-nova02:28
*** mtanino has quit IRC02:29
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/24524802:29
*** haomaiwang has joined #openstack-nova02:31
*** leseb has joined #openstack-nova02:32
*** alejandrito has joined #openstack-nova02:33
*** haomaiwang has quit IRC02:36
*** haomaiwa_ has joined #openstack-nova02:36
*** leseb has quit IRC02:37
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/24524802:38
*** apoorvad has quit IRC02:38
*** unicell has quit IRC02:40
*** aginwala has quit IRC02:40
*** aginwala has joined #openstack-nova02:41
*** shz has joined #openstack-nova02:41
*** vladikr has joined #openstack-nova02:42
*** aginwala has quit IRC02:44
*** shihanzhang has quit IRC02:44
*** aginwala has joined #openstack-nova02:44
*** baoli has quit IRC02:45
*** baoli has joined #openstack-nova02:45
*** takashin has left #openstack-nova02:48
*** aginwala has quit IRC02:48
*** aginwala has joined #openstack-nova02:51
*** leseb has joined #openstack-nova02:51
*** vilobhmm has quit IRC02:52
*** agireud has joined #openstack-nova02:52
*** _RuiChen has joined #openstack-nova02:52
*** baoli has quit IRC02:52
Kevin_Zhengpython27 and python34 always fails02:56
*** leseb has quit IRC02:56
*** RuiChen has quit IRC02:56
*** leseb has joined #openstack-nova02:56
openstackgerritZhenyu Zheng proposed openstack/nova: Split-network-plane-for-live-migration  https://review.openstack.org/24500502:57
*** baoli has joined #openstack-nova02:58
*** venkat_p has joined #openstack-nova02:58
*** annegentle has joined #openstack-nova03:02
*** Sree has joined #openstack-nova03:05
*** bapalm has quit IRC03:06
*** rushil_ has quit IRC03:07
*** Sree has quit IRC03:07
*** Sree has joined #openstack-nova03:08
*** READ10 has quit IRC03:10
*** Yanyanhu has quit IRC03:11
*** EinstCrazy has quit IRC03:11
*** EinstCrazy has joined #openstack-nova03:12
*** EinstCra_ has joined #openstack-nova03:12
oomichimrodden: like "back to the future" :-)03:13
*** pratikmallya has quit IRC03:16
*** EinstCrazy has quit IRC03:16
*** salv-orlando has joined #openstack-nova03:16
*** davideagnello has quit IRC03:18
*** coreywright has quit IRC03:18
*** devananda has quit IRC03:18
*** coreywright has joined #openstack-nova03:21
*** davideagnello has joined #openstack-nova03:21
*** shuquan has joined #openstack-nova03:24
*** Marga__ has quit IRC03:24
*** annegentle has quit IRC03:25
*** jwcroppe has joined #openstack-nova03:26
*** shuquan has quit IRC03:26
*** eliqiao_ has joined #openstack-nova03:29
*** devananda has joined #openstack-nova03:31
*** aginwala has quit IRC03:31
*** ccard__ has joined #openstack-nova03:31
*** eliqiao has quit IRC03:31
*** aginwala has joined #openstack-nova03:32
*** ccard_ has quit IRC03:34
*** Marga_ has joined #openstack-nova03:35
*** aginwala has quit IRC03:36
*** Marga_ has quit IRC03:36
*** vilobhmm has joined #openstack-nova03:39
*** salv-orlando has quit IRC03:41
*** jwcroppe has quit IRC03:44
*** ccard_ has joined #openstack-nova03:46
*** shuquan has joined #openstack-nova03:48
*** ccard__ has quit IRC03:49
*** ccard__ has joined #openstack-nova03:50
*** achanda has quit IRC03:50
*** ccard_ has quit IRC03:50
*** shuquan has quit IRC03:53
*** gcb has joined #openstack-nova03:58
*** Yanyanhu has joined #openstack-nova04:01
*** dims has quit IRC04:06
*** klkumar has quit IRC04:06
*** Yanyanhu has quit IRC04:10
*** Yanyanhu has joined #openstack-nova04:11
*** otter768 has joined #openstack-nova04:11
*** otter768 has quit IRC04:16
*** alejandrito has quit IRC04:16
*** liusheng has quit IRC04:17
*** liusheng has joined #openstack-nova04:17
*** Yanyanhu has quit IRC04:18
*** chenli has joined #openstack-nova04:18
*** klkumar has joined #openstack-nova04:18
*** achanda has joined #openstack-nova04:21
oomichialex_xu: hi, can you take a look at https://review.openstack.org/#/c/222068/ ?04:22
oomichialex_xu: you are familar with this, I believe :-)04:23
*** sacharya has joined #openstack-nova04:29
*** zhangjn has joined #openstack-nova04:35
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Support hypervisor_type option  https://review.openstack.org/24615404:37
*** Yanyanhu has joined #openstack-nova04:38
*** takashin has joined #openstack-nova04:39
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Volume support for multi hypervisors  https://review.openstack.org/24598604:42
*** sudipto has joined #openstack-nova04:42
*** Yanyan has joined #openstack-nova04:43
*** Yanyanhu has quit IRC04:44
*** zhangjn has quit IRC04:45
*** dave-mccowan has quit IRC04:45
*** jwang__ has joined #openstack-nova04:50
*** artom_ has joined #openstack-nova04:50
*** jwcroppe has joined #openstack-nova04:51
openstackgerritSudipta Biswas proposed openstack/nova: libvirt: Add ppc to supported arch for NUMA  https://review.openstack.org/24576504:52
*** liusheng has quit IRC04:53
*** mari0jv has joined #openstack-nova04:53
*** salv-orlando has joined #openstack-nova04:53
*** liusheng has joined #openstack-nova04:53
*** nikhil|afk has joined #openstack-nova04:54
*** Nakato_ has joined #openstack-nova04:55
*** baffle has joined #openstack-nova04:55
*** doffm has joined #openstack-nova04:56
*** sshen_ has joined #openstack-nova04:57
*** _d34dh0r53_ has joined #openstack-nova04:57
*** jyuso has joined #openstack-nova04:58
*** jhesketh_ has joined #openstack-nova04:58
*** cfarquhar_ has joined #openstack-nova04:58
*** Vek_ has joined #openstack-nova04:59
*** cburgess_ has joined #openstack-nova04:59
*** artom has quit IRC04:59
*** baoli has quit IRC05:00
*** mariojv has quit IRC05:00
*** jwang_ has quit IRC05:00
*** serverascode has quit IRC05:00
*** flaper87 has quit IRC05:00
*** xnox has quit IRC05:00
*** jyuso1 has quit IRC05:00
*** cburgess has quit IRC05:00
*** codebauss has quit IRC05:00
*** baffle_ has quit IRC05:00
*** Vek has quit IRC05:00
*** keekz has quit IRC05:00
*** leseb has quit IRC05:00
*** jhesketh has quit IRC05:00
*** sshen has quit IRC05:00
*** d34dh0r53 has quit IRC05:00
*** Nakato has quit IRC05:00
*** nikhil has quit IRC05:00
*** jamielennox has quit IRC05:00
*** mgagne has quit IRC05:00
*** adam_g has quit IRC05:00
*** jroll has quit IRC05:00
*** sbadia has quit IRC05:00
*** wasmum has quit IRC05:00
*** doffm_ has quit IRC05:00
*** kbyrne has quit IRC05:00
*** afazekas has quit IRC05:00
*** cfarquhar has quit IRC05:00
*** stevemar_ has quit IRC05:00
*** jroll has joined #openstack-nova05:00
*** jhesketh_ is now known as jhesketh05:00
*** adam_g` has joined #openstack-nova05:00
*** klkumar has quit IRC05:00
*** klkumar has joined #openstack-nova05:00
*** jhesketh has quit IRC05:00
*** mgagne has joined #openstack-nova05:00
*** jhesketh has joined #openstack-nova05:01
*** flaper87 has joined #openstack-nova05:01
*** keekz has joined #openstack-nova05:01
*** jamielennox has joined #openstack-nova05:01
*** wasmum- has joined #openstack-nova05:01
*** leseb has joined #openstack-nova05:01
*** jerrygb has joined #openstack-nova05:01
*** kbyrne has joined #openstack-nova05:01
*** edtubill has joined #openstack-nova05:01
*** jroll has quit IRC05:02
*** jroll has joined #openstack-nova05:02
*** jroll is now known as Guest8904005:02
*** xnox has joined #openstack-nova05:02
*** afazekas has joined #openstack-nova05:02
*** codebauss has joined #openstack-nova05:03
*** sudipto has quit IRC05:03
*** serverascode has joined #openstack-nova05:04
*** sbadia has joined #openstack-nova05:04
*** edtubill has quit IRC05:05
*** salv-orlando has quit IRC05:05
*** mdrabe has joined #openstack-nova05:06
*** jamielennox is now known as jamielennox|away05:06
*** edtubill has joined #openstack-nova05:06
*** Marga_ has joined #openstack-nova05:10
*** Marga_ has quit IRC05:11
*** Marga_ has joined #openstack-nova05:11
*** suro-patz has joined #openstack-nova05:12
*** mdrabe has quit IRC05:13
alex_xuoomichi: hi, how about we change the test name for https://review.openstack.org/#/c/222068/05:13
alex_xuoomichi: that is a little confuse for what it is test for05:14
oomichialex_xu: nice comment for fitting the above test_validate_v2compatible_request :-)05:16
alex_xuoomichi: thanks05:16
*** achanda_ has joined #openstack-nova05:16
*** betherly has quit IRC05:17
*** rpodolyaka1 has joined #openstack-nova05:17
*** annasort_ has joined #openstack-nova05:17
*** junjie has quit IRC05:18
*** auggy has quit IRC05:18
*** kgalanov has quit IRC05:18
*** greghaynes has quit IRC05:19
*** tpeoples_away has quit IRC05:19
*** bnemec has joined #openstack-nova05:19
*** rubasov1 has joined #openstack-nova05:19
*** doffm_ has joined #openstack-nova05:20
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Add microversions schema unit test for None  https://review.openstack.org/22206805:21
*** rpodolyaka1 has quit IRC05:22
oomichialex_xu: ^^^ done05:22
alex_xuoomichi: done :)05:23
oomichialex_xu: :)05:23
*** sirushti_ has joined #openstack-nova05:24
*** doffm has quit IRC05:25
*** achanda has quit IRC05:25
*** yonglihe has quit IRC05:25
*** beekneemech has quit IRC05:25
*** rubasov has quit IRC05:25
*** maurosr has quit IRC05:25
*** annasort has quit IRC05:25
*** garyk has quit IRC05:25
*** DuncanT has quit IRC05:25
*** sballe has quit IRC05:25
*** andrewbogott has quit IRC05:25
*** markmcclain has quit IRC05:25
*** sirushti has quit IRC05:25
*** jbernard has quit IRC05:25
*** onder has quit IRC05:25
*** sirushti_ is now known as sirushti05:25
*** maurosr has joined #openstack-nova05:26
*** DuncanT has joined #openstack-nova05:26
*** garyk has joined #openstack-nova05:27
*** tpeoples_away has joined #openstack-nova05:28
*** andrewbogott has joined #openstack-nova05:29
*** auggy has joined #openstack-nova05:30
*** sballe has joined #openstack-nova05:30
*** jbernard has joined #openstack-nova05:30
*** onder has joined #openstack-nova05:30
*** markmcclain has joined #openstack-nova05:31
*** suro-patz has quit IRC05:31
*** gus has quit IRC05:33
*** gus has joined #openstack-nova05:34
*** junjie has joined #openstack-nova05:34
*** greghaynes has joined #openstack-nova05:35
*** yonglihe has joined #openstack-nova05:38
*** eliqiao has joined #openstack-nova05:38
*** jerrygb has quit IRC05:38
*** betherly has joined #openstack-nova05:39
*** kgalanov has joined #openstack-nova05:39
*** eliqiao_ has quit IRC05:41
*** rpodolyaka1 has joined #openstack-nova05:42
*** adam_g` is now known as adam_g05:42
*** adam_g has quit IRC05:42
*** adam_g has joined #openstack-nova05:42
*** rushiagr_away is now known as rushiagr05:42
*** klkumar has quit IRC05:44
*** jmccrory has quit IRC05:48
*** sacharya has quit IRC05:50
*** jmccrory has joined #openstack-nova05:51
*** aswadr has joined #openstack-nova05:54
*** yangyapeng has quit IRC05:59
*** moshele has joined #openstack-nova06:00
*** edtubill has quit IRC06:02
*** edtubill has joined #openstack-nova06:05
*** rushil has joined #openstack-nova06:06
*** unicell has joined #openstack-nova06:07
*** sirushti has quit IRC06:11
*** sirushti has joined #openstack-nova06:11
*** otter768 has joined #openstack-nova06:12
*** otter768 has quit IRC06:16
openstackgerritVincent Untz proposed openstack/python-novaclient: Fix typo in error message  https://review.openstack.org/24619906:19
*** gcb has quit IRC06:19
*** abhishekk has joined #openstack-nova06:21
*** rcernin has joined #openstack-nova06:22
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Imported Translations from Zanata  https://review.openstack.org/24272706:23
*** aginwala has joined #openstack-nova06:24
*** vilobhmm has quit IRC06:25
*** korzen has joined #openstack-nova06:25
*** aginwala_ has joined #openstack-nova06:30
*** aginwala has quit IRC06:32
*** gcb has joined #openstack-nova06:32
openstackgerritZhenyu Zheng proposed openstack/nova: Add project-id and user-id when list server-groups  https://review.openstack.org/20901806:32
*** rushil has quit IRC06:33
*** markmc has quit IRC06:33
*** markmc has joined #openstack-nova06:34
*** salv-orlando has joined #openstack-nova06:34
*** aginwala_ has quit IRC06:35
*** edtubill has quit IRC06:35
*** pratikmallya has joined #openstack-nova06:36
*** aginwala has joined #openstack-nova06:36
*** salv-orlando has quit IRC06:37
*** Jeffrey4l has quit IRC06:39
*** jerrygb has joined #openstack-nova06:39
*** ankit_ag has joined #openstack-nova06:41
*** venkat_p has left #openstack-nova06:41
*** eliqiao has quit IRC06:44
*** jerrygb has quit IRC06:45
*** eliqiao has joined #openstack-nova06:45
*** sudipto has joined #openstack-nova06:47
*** sileht has joined #openstack-nova06:49
*** josecastroleon has joined #openstack-nova06:50
*** Jeffrey4l has joined #openstack-nova06:51
*** angdraug has joined #openstack-nova06:53
*** aswadr has quit IRC06:55
*** rpodolyaka1 has quit IRC06:56
*** Yanyan has quit IRC07:00
*** unicell1 has joined #openstack-nova07:01
*** dhellmann has quit IRC07:01
*** dhellmann has joined #openstack-nova07:02
*** unicell has quit IRC07:02
openstackgerritWen Zhi Yu proposed openstack/nova: Use SHA-256 algorithm in nova/crypto.py  https://review.openstack.org/24621707:03
*** pkoniszewski has quit IRC07:04
*** yonglihe has quit IRC07:04
*** unicell has joined #openstack-nova07:05
*** unicell1 has quit IRC07:05
*** penick has joined #openstack-nova07:07
*** rpodolyaka1 has joined #openstack-nova07:07
*** sahid has joined #openstack-nova07:08
*** mjura has joined #openstack-nova07:09
*** moshele has quit IRC07:10
*** mleroy has quit IRC07:12
openstackgerritTang Chen proposed openstack/nova-specs: Support triggering crash dump in a server.  https://review.openstack.org/22925507:13
*** deepakcs has joined #openstack-nova07:15
tangchen__alex_xu: Hi alex, on November 19, we are going to review spec, right ?07:18
tangchen__alex_xu: What time shall we start on that day ?07:19
openstackgerritPawel Koniszewski proposed openstack/nova-specs: Pause VM during live migration  https://review.openstack.org/22904007:19
*** moshele has joined #openstack-nova07:19
alex_xutangchen__: anytime in that day07:20
*** jyuso has quit IRC07:20
*** jyuso has joined #openstack-nova07:21
alex_xutangchen__: that is for local time07:21
sahidjaypipes: hello, any chance you take a look in my PR, also please when you have moment let me know how i can try to involve myself better, don't know what was the concensus about design07:21
*** pratikmallya has quit IRC07:22
*** rpodolyaka1 has quit IRC07:23
*** paul-carlton has joined #openstack-nova07:25
*** rpodolyaka1 has joined #openstack-nova07:25
*** Jeffrey4l has quit IRC07:26
*** garyk1 has joined #openstack-nova07:28
openstackgerritTom Patzig proposed openstack/nova: Add 'insecure' option to ironic sect in nova.conf  https://review.openstack.org/23256007:28
*** lpetrut has joined #openstack-nova07:29
tangchen__alex_xu: Thanks. :)07:32
alex_xutangchen__: np :)07:32
*** alexschm has joined #openstack-nova07:34
openstackgerritAlex Xu proposed openstack/nova: POC: generate swagger spec from api code base  https://review.openstack.org/23344607:35
*** achanda_ has quit IRC07:36
*** nkrinner has joined #openstack-nova07:37
*** Jeffrey4l has joined #openstack-nova07:39
alex_xuoomichi: ^ something you expected :)07:39
oomichialex_xu: thanks, will see it :-)07:40
*** artom__ has joined #openstack-nova07:40
alex_xuoomichi: thanks :)07:40
*** artom_ has quit IRC07:40
*** artom__ has quit IRC07:41
*** artom__ has joined #openstack-nova07:41
*** mleroy has joined #openstack-nova07:45
*** aginwala has quit IRC07:45
*** aginwala has joined #openstack-nova07:49
*** longsube has joined #openstack-nova07:50
*** Jeffrey4l has quit IRC07:53
*** longsube has quit IRC07:54
*** irenab has quit IRC07:54
*** irenab has joined #openstack-nova07:54
*** romainh has joined #openstack-nova07:56
*** tyagiprince has joined #openstack-nova07:57
tyagiprincehey everyone.. I am facing problem in booting up my instance.. It gives the following error.. ERROR (ClientException): The server has either erred or is incapable of performing the requested operation. (HTTP 500)07:57
tyagiprinceunable to find the cause root of this problem..07:57
*** gongysh has joined #openstack-nova07:58
*** rotbeard has joined #openstack-nova07:58
tangchen__tyagiprince: Hi, what is the status of your instance ?07:59
tangchen__tyagiprince: If it is ERROR, you can use nova reset-state --active.08:00
tangchen__tyagiprince: And try to boot it again.08:00
tangchen__tyagiprince: And also, please try to see the logs. Maybe you can find something.08:01
*** rotbart has joined #openstack-nova08:02
*** rotbeard has quit IRC08:04
openstackgerritAlex Xu proposed openstack/nova-specs: Make block_migration and host flags in live-migration API  https://review.openstack.org/24554308:04
*** tyagiprince has quit IRC08:06
*** lpetrut has quit IRC08:08
moshelejohnthetubaguy, danp: can you please review https://review.openstack.org/#/c/199488/ and https://review.openstack.org/#/c/227160/08:13
*** otter768 has joined #openstack-nova08:13
*** jichen has joined #openstack-nova08:14
*** matrohon has joined #openstack-nova08:15
*** scheuran has joined #openstack-nova08:15
*** otter768 has quit IRC08:18
*** deepakcs has quit IRC08:22
*** ildikov has quit IRC08:28
*** mpavone has joined #openstack-nova08:30
*** penick has quit IRC08:33
*** rdopiera has joined #openstack-nova08:37
*** liverpooler has joined #openstack-nova08:38
*** lpetrut has joined #openstack-nova08:38
*** jerrygb has joined #openstack-nova08:40
*** liverpoo1er has joined #openstack-nova08:41
*** jlanoux has joined #openstack-nova08:41
*** liverpooler has quit IRC08:43
*** mpavone has quit IRC08:43
jichenalex_xu: hi, another look at https://review.openstack.org/#/c/232721/?08:43
*** korzen has quit IRC08:43
markus_ztonyb: I guess we both have the most terrible timezones when trying to communicate in real time :)08:44
*** deepakcs has joined #openstack-nova08:44
tonybmarkus_z: Yeah, but as it turns out I'm on now ;p08:44
tonybmarkus_z: how can I help?08:44
markus_ztonyb: woah, hey :)08:44
markus_ztonyb: about the bleeding edge libvirt gate job08:44
alex_xujichen: emm...just looks like strange, we have some fake at db layer and one at compute api layer08:45
tonybmarkus_z: Ahh okay.08:45
*** tyagiprince has joined #openstack-nova08:45
*** salv-orlando has joined #openstack-nova08:45
tyagiprince<tangchen>: instance is unable to boot.. when i boot an instance with --nic net-id="something-here", it gives an error after a while.. If I dont pass this argument the instance gets booted up and comes in active state08:45
markus_ztonyb: I haven't created a gate job before and was unsure where exactly to start08:45
tonybmarkus_z: So my idea is to make a devstack plugin that will add $repo which contains newwer libvirt and qwemu packages08:46
jichenum... maybe, but I don't think it has much effect, right? we focus on unit test for the code path instead of others, right?08:46
tyagiprince<tangchen__>: instance is unable to boot.. when i boot an instance with --nic net-id="something-here", it gives an error after a while.. If I dont pass this argument the instance gets booted up and comes in active state08:46
tonybmarkus_z: To be honnest I'm not 100% certain myself ;P08:46
tonybmarkus_z: but I figure the plugin is a good place then we'll work out how to use it ;P08:46
alex_xujichen: yea, ideally we should use mock instead of those stubs08:46
*** jerrygb has quit IRC08:46
alex_xujichen: the stub of key_pair_create is still needed? as we already have mock at compute api layer?08:47
*** rpodolyaka1 has quit IRC08:47
markus_ztonyb: Yeah, that's a better start than what I head in mind. I wanted to create a node which already has the latest libvirt/qemu.08:47
jichenalex_xu: for other test cases ...08:47
jichenfor simplility and less changes needed08:47
tonybmarkus_z: Yeah that'll be hard.08:47
markus_ztonyb: yeah, sounds pretty dumb after I typed it :D08:48
tonybmarkus_z: Although it might not be terrible to get a nodepool image for Ubuntu 15.1008:48
tonybthat would be newer, but the custom repo could be build daily or whenever we thought there was something of interest to test08:48
bauzasgood morning Nova08:48
tonybheck the devstack plugin can even try different repos08:49
markus_ztonyb: Yepp, sounds more flexible08:49
tonybI feel like having conttol over it would be good for testing the latest and greatest.08:49
openstackgerritPawel Koniszewski proposed openstack/nova: Live migration failure in API doesn't revert task_state of vm  https://review.openstack.org/16891608:50
tonybmarkus_z: I don't have a good handle on how we test the serial console stuff with tempest.08:50
tonybmarkus_z: We should probably talk to mtreinish and sdague about that.08:50
markus_ztonyb: OK, assumed we have a node with U1510 and the devstack plugin, the gate job will then trigger the download from $repo and build it from source?08:51
*** deepakcs has quit IRC08:52
tonybmarkus_z: No I was thinking that $repo would contain Ubuntu packages for qemu/libvirt08:52
*** sahid_ has joined #openstack-nova08:53
*** mpavone has joined #openstack-nova08:53
markus_ztonyb: I used "novaconsole" from larsks back then to test it manually: https://github.com/larsks/novaconsole/08:53
*** mpavone has quit IRC08:53
tonybmarkus_z: I we *could* build form source but thsat creates a massive delay in the dsvm jobs and an additional failure mode08:53
alex_xujichen: strange...I can pass the test if I remove the line 36408:53
*** sahid has quit IRC08:53
tyagiprincehey people need help with nova boot.. instance is unable to boot.. when i boot an instance with --nic net-id="something-here", it gives an error after a while.. If I dont pass this argument the instance gets booted up and comes in active state08:53
tyagiprincedoes anyone know what might be the problem08:53
jichenalex_xu: ok only 364, might, I removed 360-367 for test08:53
alex_xujichen: ok08:54
bauzastyagiprince: hi, I feel you should get more help in the #openstack channel08:54
bauzastyagiprince: see that channel $topic08:54
tyagiprincethanks bauzas08:54
markus_ztonyb: Right, the job would take longer. I was just thinking about the delay when upstream libvirt provides the new version until Ubuntu has it ins packages.08:54
bauzastyagiprince: although I guess you'd get more support in the general channel by US working hours08:54
markus_z*in its08:55
tonybmarkus_z: Well I was thinking that we'd build the packages ourselves08:55
markus_ztonyb: OK, got it08:55
tonybmarkus_z: initially I was thinking I'd build $latest version.08:55
*** ihrachys has joined #openstack-nova08:56
tonybmarkus_z: but we could have nightly builds08:56
tonybmarkus_z: of course I'll look again to see if anyone is already doing these builds08:56
tonybmarkus_z: I know they used for for Fedora but Fedora in the gate can be problematic08:56
openstackgerritJames Page proposed openstack/nova: Add 'lxd' to the list of recognized hypervisors  https://review.openstack.org/24581108:56
markus_ztonyb: "nightly" gets blurry for me with all the different timezones :p08:57
bauzastonyb: why Fedora in the gate seems problematic ?08:57
bauzastonyb: I thought there were some efforts for running it08:57
tonybmarkus_z: Well checkout git master at 00:00 UTC ;P08:57
*** sahid_ has quit IRC08:57
*** sahid has joined #openstack-nova08:58
tonybbauzas: Sure, its there.  Last time I looked it was F21 (which is now EOLd)08:58
tonybbauzas: I'd be happy if I was wrong and it was all F23 and all signing and dancing :)08:58
* markus_z will skip the dancing part08:59
*** salv-orlando has quit IRC08:59
bauzastonyb: I mean, I just think that using Fedora would help, right?08:59
jichenalex_xu: I will try this and update patch if it pass08:59
bauzastonyb: if so, I can discuss internally to see how we can help you09:00
tonybbauzas: As long as we didn't have to close a backlog of other issues sure.09:00
tonybbauzas: Cool :)09:00
*** markus_z has quit IRC09:00
bauzastonyb: fersur, you not happy with getting lots of regressions ? nah.09:00
tonybbauzas: For the record I'm a fedora package maintainer and use it a lot at home so I have no problem with Fedora09:00
bauzastonyb: I mean, I'm not really opiniated09:01
tonybbauzas: :)09:01
bauzastonyb: but it seems Fedora is a bit more goodness for getting the latest libvirt releases, right?09:01
*** exploreshaifali has joined #openstack-nova09:02
bauzastonyb: so, I can try to see how I can bring back some help for you09:02
tonybbauzas: Yeah There was someone doing builds for supported Fedora releases on each libvirt release09:02
tonybbauzas: That'd be very welcome.09:02
*** markus_z has joined #openstack-nova09:03
*** xek has joined #openstack-nova09:03
bauzastonyb: if that doesn't match because of X, then feel free to play with Ubuntu :D09:03
markus_ztonyb: sorry, my network kicked me out09:03
*** lucas-afk is now known as lucasagomes09:03
tonybbauzas: I can build .debs but .rpms make more sense to me :)09:04
tonybmarkus_z: np09:04
*** deepakcs has joined #openstack-nova09:04
*** eglynn_ has joined #openstack-nova09:06
*** ildikov has joined #openstack-nova09:07
*** Marga_ has quit IRC09:07
markus_ztonyb: bauzas: Just to double-check if I understand it. Please correct me if I'm wrong. We have 4 layers now. 1) $repo will contain the $latest libvirt version 2) We need a node image which can run $latest libvirt version (maybe Ubuntu15.10) 3) Devstack pulls $latest libvirt from $repo in a gate test 4) tempest needs addional test cases to test "serial console"09:07
markus_zIt makes sense in my head, which confuses me...09:08
*** liverpoo1er has quit IRC09:11
*** liverpooler has joined #openstack-nova09:11
tonybmarkus_z: 1) We create $repo to contain latest released libvirt and qemu compatible with $nodepool image 2) we create a devstack job that ultilises $repo 3) We modifiy tempest to be capable of dealing with both serial console and normal "log to file" consoles09:12
openstackgerritjichenjc proposed openstack/nova: use NoDBTestCase for KeypairPolicyTest  https://review.openstack.org/23272109:12
tonybmarkus_z: eventually 3 goes away once we have a unified way of dealing with consoles regardless of how the operator has configured the system.09:12
tonybmarkus_z: does that make sense?09:12
*** deepakcs has quit IRC09:13
openstackgerritShaoHe Feng proposed openstack/nova: do not log twice with different level.  https://review.openstack.org/24626909:14
markus_ztonyb: Yepp, it does. I can work on 3. I don't think that it goes away, because we have 2 separate REST API's we have to fulfill.09:14
*** yassine has joined #openstack-nova09:15
*** salv-orlando has joined #openstack-nova09:15
*** salv-orlando has quit IRC09:15
markus_zI also think 3 can be implemented independently from 1+209:15
tonybmarkus_z: You'll want to sync up with mtreinish and sdague on that.09:15
*** klkumar has joined #openstack-nova09:15
*** salv-orlando has joined #openstack-nova09:15
tonybmarkus_z: I could be very wrong about needing it.09:15
*** deepakcs has joined #openstack-nova09:16
*** moshele has quit IRC09:16
*** Marga_ has joined #openstack-nova09:16
markus_ztonyb: OK, I'm jumping to conclusions right now.09:17
markus_ztonyb: I think I get the flow and potential work items. I would post this on the ML to get also feedback from the infra folks?09:17
tonybmarkus_z: Sounds good.09:17
*** lpetrut has quit IRC09:18
markus_ztonyb: Awesome, thanks a lot! Just chime in on the ML post when I say something crazy09:18
openstackgerritPawel Koniszewski proposed openstack/nova: Live migration failure in API leaves VM in MIGRATING state  https://review.openstack.org/16891609:18
tonybmarkus_z: Sure :)09:18
markus_zcool, have nice evening09:19
tonybHas anyone seen 'TypeError: Cannot mock a MockAnything! Did you remember to call UnsetStubs in your previous test?' before?09:19
*** salv-orlando has quit IRC09:20
*** salv-orlando has joined #openstack-nova09:20
*** moshele has joined #openstack-nova09:21
*** klkumar has quit IRC09:21
openstackgerritli,chen proposed openstack/nova: Add unit tests for nova.virt.firewall.IpTablesFirewallDriver  https://review.openstack.org/24627409:22
*** ekarlso has quit IRC09:22
*** rpodolyaka1 has joined #openstack-nova09:23
markus_ztonyb: I could only imagine a missing teardown in the test case class09:24
*** ihrachys has quit IRC09:24
tonybmarkus_z: Hmm perhaps.09:24
sudiptomarkus_z, +109:24
tonybThe same code passed the check pipline and failed in the gate so I was thinking that something other than the code was to blame ...09:25
*** ihrachys has joined #openstack-nova09:26
*** mgoddard has joined #openstack-nova09:27
markus_zCould still be the code. AFAIK the test methods get executed in a random order. Which results in different starting positions when a teardown is missing.09:27
markus_zWhich would explain why it passes sometimes.09:28
openstackgerritPavel Kholkin proposed openstack/nova: WIP: enginefacade: 'instance'  https://review.openstack.org/24414609:29
tonybmarkus_z: Oh.  I have to admit I didn't know that. (about the order of execution)09:30
markus_ztonyb: wohoo, I could help :D09:30
*** eglynn_ is now known as eglynn09:31
*** openstackgerrit has quit IRC09:31
*** ihrachys has quit IRC09:31
* tonyb now debates if I shoudl just recheck this review.09:31
*** openstackgerrit has joined #openstack-nova09:32
markus_zWhich review is it?09:32
*** klkumar has joined #openstack-nova09:33
tonybmarkus_z: it09:34
*** aginwala has quit IRC09:34
tonybs kinda slowing things down09:34
johnthetubaguytonyb: i did see that one before, but not in the gate09:35
*** ihrachys has joined #openstack-nova09:35
bauzastonyb: I remember that story, sec09:35
bauzasjohnthetubaguy: yeah that09:35
*** chenli has quit IRC09:35
* tonyb goes to put the chickens in for the night ....09:35
bauzasit's a py34 job failing09:36
bauzassince py34 is only running a few tests, I wonder if that's just now a brand new test09:36
bauzasthat said, it should have failed in the check pipeline, not in gate09:36
*** ekarlso has joined #openstack-nova09:37
*** achanda has joined #openstack-nova09:38
*** ekarlso has quit IRC09:38
*** ekarlso has joined #openstack-nova09:38
kashyapmarkus_z: (Just catching up on scroll.)  There was a long discussion previously about running newest libvirt on the Gate, trying to fish that. . .09:39
tonybbauzas: I checked and nothing merged between the 2 runs so it's not like the failing test was added in between09:39
bauzastonyb: yup09:39
tonybkashyap: We're only talking about doing that for *some* runs not all.09:40
oomichialex_xu: thanks for https://review.openstack.org/#/c/233446 , that is very cool :-)09:40
bauzastonyb: checking e-r09:40
tonybkashyap: Which is different to last time it came up ...09:40
kashyaptonyb: Still reading scroll, what are these "some" runs?  (If you still have time.)09:40
openstackgerritPavel Kholkin proposed openstack/nova: WIP: enginefacade: 'instance'  https://review.openstack.org/24414609:40
tonybkashyap: well intially it'd be only on the experimental queue, then (clearly) a dedicated job09:40
tonybkashyap: it's not in the scrool back, but basically we want to be able test code paths that have a requirement on libvirt that's newer than in $distro09:41
kashyaptonyb: Yes, experimental-only sounds reasonable.  And, we're diligent in analyzing the results :-)09:42
*** danpb has joined #openstack-nova09:42
tonybkashyap: yeah to start with :)09:43
kashyaps/we're/if we're/09:43
tonybkashyap: perhaps for all of mitaka, that parts not very well defined yet ;P09:43
*** rpodolyaka1 has quit IRC09:43
kashyapCool, is there an intention to start a thread about it on ML?09:44
*** claudiub has joined #openstack-nova09:44
*** rpodolyaka1 has joined #openstack-nova09:45
tonybkashyap: I believe that markus_z will commited to doing that.09:45
markus_zkashyap: tonyb: Yepp, I'm going to post that on the ML today to get more feedback.09:46
*** exploreshaifali has quit IRC09:47
kashyapmarkus_z: Cool.09:48
*** e0ne has joined #openstack-nova09:48
johnthetubaguytonyb: totally sounds like a random test re-ordering bug09:49
tonybjohnthetubaguy: okay.09:50
*** Marga_ has quit IRC09:50
* tonyb recheck's it09:50
*** mgoddard has quit IRC09:50
johnthetubaguytonyb: how is the pass rate, its quite possible not that patch09:51
bauzasjohnthetubaguy: +109:51
bauzastonyb: johnthetubaguy: I checked logstash and it only occurred as a spike09:51
tonybHuh, how does a review get into the gate pipeline and fail pep8?09:51
*** belmoreira has joined #openstack-nova09:52
bauzasmeh, the py34 output formatting breaks logstash...09:52
bauzasit makes the message very big09:52
johnthetubaguytonyb: thats oddd09:52
bauzastonyb: which one ?09:52
bauzastonyb: someone the PEP8 check fails because of some other causes (timeout or file upload failing etc.)09:53
tonybit went throught check in August to it's possible we changed the rules in between09:53
johnthetubaguyyeah, change in rules would do it09:54
bauzastonyb: 218644 is okay09:54
bauzasunless I'm blind09:54
johnthetubaguyso we do merge lots of stuff on top of each other09:54
johnthetubaguyso it could be a patch underneath that fails them all, I guess09:55
tonybjohnthetubaguy: Yeah that was it.09:55
*** ihrachys has quit IRC09:55
tonybjohnthetubaguy, bauzas: It just got re-queued in zuul ....09:55
bauzastonyb: haha, got it09:55
johnthetubaguywe used to recheck before joining gate, but it did waste a lot of resources I guess09:55
johnthetubaguy(well and multiplied the issue of failure rates)09:56
tonybAh so much history I don't have ... that's good and bad ;P09:56
bauzasjohnthetubaguy: right https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L53-L5609:56
*** ekarlso has quit IRC09:57
bauzasjohnthetubaguy: tonyb: there is no longer a newer-than tag09:58
*** achanda has quit IRC10:00
*** rk4n has joined #openstack-nova10:00
bauzasoooooh, the liberty gate is also impacted by the o.vo issue...10:01
*** Sree_ has joined #openstack-nova10:01
tonybbauzas: Yeah.10:01
bauzastonyb: heard of any backport ?10:02
*** Sree_ is now known as Guest5027310:02
bauzastonyb: I left the chan before the issue popped up, so I need to scroll back10:02
tonybbauzas: Yeah but I can't approve it and melwitt had a question about it ....10:02
*** tyagiprince has quit IRC10:02
johnthetubaguywe don't pin on unit tests yet, as I understand it10:02
johnthetubaguynot sure why we don't do that for stable/liberty10:02
bauzaslinks ?10:02
openstackgerritBob Ball proposed openstack/nova: XenAPI: Cope with more Cinder backends  https://review.openstack.org/24478910:03
bauzasFWIW, the reno liberty changes are now stuck because of that - which blocks the reno CI job - which blocks us approving relnote changes10:03
tonybbauzas: melwitt's question is a comment on your backport:10:04
bauzastonyb: which backport ?10:04
johnthetubaguyhmm, I am unsure now, maybe they are pinned by upper constraints10:05
tonybbauzas: https://review.openstack.org/#/c/246022/ Oh dansmith did it not you10:05
*** markvoelker has quit IRC10:05
*** Sree has quit IRC10:05
bauzastonyb: okay, I got the change-id10:05
tonybbauzas: Yeah.10:05
tonybbauzas: I was checking ....10:05
openstackgerritlyanchih proposed openstack/nova: NUMATopologyFilter raise exception and not continue filter next node  https://review.openstack.org/20063010:06
bauzasokay, I need to understand what was the problem and what was doing the fix10:06
openstackgerritBob Ball proposed openstack/nova: XenAPI: Fix race in rotate_xen_guest_logs  https://review.openstack.org/24589810:07
*** e0ne has quit IRC10:07
*** mgoddard has joined #openstack-nova10:08
*** ociuhandu has quit IRC10:09
*** achanda has joined #openstack-nova10:12
*** ildikov has quit IRC10:12
*** daemontool has joined #openstack-nova10:12
*** ihrachys has joined #openstack-nova10:13
*** eliqiao has quit IRC10:13
*** ptm_away is now known as PaulMurray10:13
*** otter768 has joined #openstack-nova10:14
*** daemontool has quit IRC10:14
*** alex_klimov has joined #openstack-nova10:14
openstackgerritZhongcheng Lao proposed openstack/nova: VMware: Omit exceptions during snapshot cleanup  https://review.openstack.org/21608610:14
*** daemontool has joined #openstack-nova10:15
alex_xuoomichi: thanks, glad to see you like it10:15
oomichialex_xu: yeah, very nice for me. can you check my comment again?10:15
*** eliqiao has joined #openstack-nova10:16
oomichialex_xu: and we need to build a consensus for the direction soon for making it in Mitaka10:16
alex_xuoomichi: yea, I checked, I remember swagger support something for include10:16
openstackgerritMaxime Leroy proposed openstack/nova: Libvirt: Support fp plug in vhostuser vif  https://review.openstack.org/24536910:16
openstackgerritMaxime Leroy proposed openstack/nova: Libvirt: Support ovs fp plug in vhostuser vif  https://review.openstack.org/23768010:16
alex_xuoomichi: but let me check more to ensure that10:16
*** achanda has quit IRC10:16
alex_xuoomichi: yea10:16
*** Guest50273 has quit IRC10:17
oomichialex_xu: api thing needs a lot of time/patches always ;-)10:17
alex_xuoomichi: what suggestion you have for build a consensus direction10:17
*** Sree has joined #openstack-nova10:17
alex_xuoomichi: yea...a lot10:17
*** EinstCra_ has quit IRC10:18
*** otter768 has quit IRC10:18
openstackgerritAdelina Tuvenie proposed openstack/nova: Added support for new block device format in Hyper-V  https://review.openstack.org/24629810:18
openstackgerritAdelina Tuvenie proposed openstack/nova: Added support for new block device format in vmops  https://review.openstack.org/24629910:18
oomichialex_xu: need to update the spec for getting a consensus10:18
alex_xudanpb: PaulMurray, good morning, hope you can give some opinion about whether this propose is useful for live-migration works https://review.openstack.org/245543, thanks in advance!10:19
alex_xuoomichi: yea, also need talk with doc team, I missed that in this week10:19
alex_xuoomichi: I have a little worry about this https://github.com/russell/fairy-slipper didn't very active recently10:20
johnthetubaguyalex_xu: is that the bit that converts swagger into html output?10:20
alex_xujohnthetubaguy: yea10:21
*** lpetrut has joined #openstack-nova10:21
oomichialex_xu: yeah, I am imaging russell's script need to be merged into tempest-lib in long-term I guess10:21
PaulMurrayalex_xu, looking10:22
oomichialex_xu: but it should be out of scope now from swagger thing10:22
alex_xuPaulMurray: thanks :)10:22
alex_xuoomichi: ok10:22
openstackgerritPavel Kholkin proposed openstack/nova: WIP: enginefacade: 'security_group'  https://review.openstack.org/24630210:24
sahidhello johnthetubaguy i saw that you made moving branch realtime but unfortunatly you stopped to review, did you find a problem or something like that ?10:25
johnthetubaguysahid: ran out of hours in the day, and I think that went deep into libvirt bits I don't really know about10:26
*** exploreshaifali has joined #openstack-nova10:27
*** ihrachys has quit IRC10:27
sahidjohnthetubaguy: yes unfortunatly only danpb is involved on libvirt driver and he already have acked them10:28
*** subscope has joined #openstack-nova10:30
*** alexschm has quit IRC10:31
*** haomaiwa_ has quit IRC10:31
*** openstackgerrit has quit IRC10:31
*** ildikov has joined #openstack-nova10:31
*** thorst has joined #openstack-nova10:32
PaulMurrayalex_xu, seems related to https://review.openstack.org/#/c/227278/10:32
*** openstackgerrit has joined #openstack-nova10:32
*** haomaiwang has joined #openstack-nova10:32
PaulMurrayI think ultimately there may be no need for the block-migration flag altogether, but maybe not just yet10:32
PaulMurrayPersonally I think the spec is in scope for live migration subteam10:32
PaulMurrayI'll add it to the pages for review10:33
alex_xuPaulMurray: thanks :)10:33
alex_xuPaulMurray: I will take a look at the patch you pointed out10:33
openstackgerritClaudiu Belu proposed openstack/nova: Hyper-V: Fixes disk overhead claim issue  https://review.openstack.org/23428310:33
openstackgerritClaudiu Belu proposed openstack/nova: Hyper-V: Fixes disk overhead claim issue  https://review.openstack.org/24630710:33
*** thorst has quit IRC10:35
*** ekarlso has joined #openstack-nova10:36
johnthetubaguyPaulMurray: +1 for removing that as soon as we can10:36
*** ociuhandu has joined #openstack-nova10:37
*** Sree has quit IRC10:38
*** achanda has joined #openstack-nova10:38
*** zhenguo has quit IRC10:38
*** Sree has joined #openstack-nova10:39
*** ekarlso has quit IRC10:39
PaulMurrayjohnthetubaguy, do you know if its needed for non-libvirt hypervisors?10:39
claudiubwhat is needed by other hypervisors?10:40
johnthetubaguyPaulMurray: as far as I know, its mostly as stupid with the other hypervisors10:40
johnthetubaguyblock_migrate flag for live-migrate10:40
alex_xujohnthetubaguy: PaulMurray I thought, there will have different shared storage pools in the env, so block_migration can be used control for whether the live-migration will across the shared storage pool10:40
*** ihrachys has joined #openstack-nova10:40
claudiubah, I see.10:41
johnthetubaguyalex_xu: possibly, but we should know that an just to the right thing10:41
johnthetubaguynow there is a use cases where we might want to say, make sure no block migration, error out if that needed10:41
johnthetubaguybut I much prefer the idea of "doing whats needed" by default10:42
johnthetubaguyleaking too many internal details out the API with the current approach10:42
johnthetubaguy... I have been meaning to sort out that migrate/live-migrate API for years, just never had the time to do it, so happy to see movement in that direction10:43
*** jerrygb has joined #openstack-nova10:43
PaulMurrayjohnthetubaguy, I thought same, but to be honest, its a bit like saying "do X but tell me if you think I'm wrong because I can't be bothered to figure it out" :)10:43
alex_xujohnthetubaguy: yea, agree10:43
*** Sree has quit IRC10:43
openstackgerritClaudiu Belu proposed openstack/nova: Compute: Adds driver disk_gb instance overhead estimation  https://review.openstack.org/23428310:44
johnthetubaguyPaulMurray: heh10:44
PaulMurrayjohnthetubaguy, I tended to agree with jaypipes  comment in the pause VM spec https://review.openstack.org/#/c/229040/10:44
PaulMurrayI know we started with a "force" name and decided against it10:45
PaulMurraybut there does seem to be somehting more generic that can be done instead of pause10:45
PaulMurrayin the future10:45
PaulMurraylike: invoke agressive migration algorithm even though its bad for the user10:45
PaulMurrayAny htoughts?10:46
johnthetubaguy"force-complete" makes more sense to me10:46
johnthetubaguyits clearer about the intention10:46
*** haomaiwang has quit IRC10:47
*** jerrygb has quit IRC10:47
PaulMurrayYes (although I can never settle on a name so I'll leave that to others)10:48
PaulMurrayJust seems to be far from normal pause API in my mind now10:48
PaulMurrayAppart from the exact name I think the spec is ready10:48
*** haomaiwang has joined #openstack-nova10:49
*** haomaiwang has quit IRC10:49
*** haomaiwa_ has joined #openstack-nova10:49
*** anshprat_ has joined #openstack-nova10:49
*** rpodolyaka1 has quit IRC10:50
*** subscope has quit IRC10:50
*** rk4n has quit IRC10:50
anshprat_hi, I am wondering how do I see the commit https://github.com/openstack/nova/commit/1cfcfa5aa8ff2437f289c9af0d1ded5047388c11 in openstack git? I mean its not an upstream change via gerrit etc so how did it land in there..?10:50
*** yassine has quit IRC10:51
*** rk4n has joined #openstack-nova10:52
andrearosaPaulMurray: My point about using Force is that you can never guarantee that the live migration will complete even if you pause the instance.10:52
PaulMurraydanpb, I looked at your comment on the pause vm spec - what do you think about using a more generic name for live-migrate-pause-vm - thinking that it could10:53
PaulMurrayoptions in future for different approaches?10:53
johnthetubaguyandrearosa: really? what goes wrong after the pause?10:53
*** ihrachys has quit IRC10:54
johnthetubaguyandrearosa: I am fine if the pause then makes the operation error out quickly, just wondering what could go into an infinite loop after the pause?10:54
andrearosajohnthetubaguy: I think paul-carlton1 has more use cases than me, I can think to some network issue or the target host goes down or other stuff like that10:55
andrearosajohnthetubaguy: not infinite loop but failed migration10:55
johnthetubaguyandrearosa: yeah, that stuff is fine, its the loop for ever bit that it needs to stop10:55
johnthetubaguyforce operations can often fail, so I think its OK10:56
andrearosajohnthetubaguy: ok so when you think to force live miggration to finish you do not assume anything about the final result. It can fail or success but it reaches a final state10:56
*** ihrachys has joined #openstack-nova10:57
*** e0ne has joined #openstack-nova10:57
johnthetubaguyandrearosa: thats a good way of putting it, force a final state10:58
PaulMurrayjohnthetubaguy, andrearosa I think what we will need to decide is if there is any intent to make this a more generic command or if it will only ever do a pause10:58
andrearosajohnthetubaguy: force a final state makes sense10:59
johnthetubaguyPaulMurray: I am fine with the body of the API call getting a method argument eventually10:59
johnthetubaguyandrearosa: does live-migrate-force-end sound better?11:00
*** takashin has quit IRC11:01
*** aysyd has joined #openstack-nova11:02
johnthetubaguyPaulMurray: +1 your previous statement about the spec, it does seem ready apart from the name thing, and the details around that11:02
andrearosajohnthetubaguy: to me yes, it is clear what that means.11:02
*** liusheng has quit IRC11:03
PaulMurrayjohnthetubaguy, I'm also fine with setting the intent that this API would be extended if another alternative to pause comes up11:04
*** shz has quit IRC11:05
*** markvoelker has joined #openstack-nova11:06
*** shz has joined #openstack-nova11:06
johnthetubaguyandrearosa: just looking at this nova-manage add, I thought we already had a blueprint for that somewhere?11:06
*** dims has joined #openstack-nova11:08
johnthetubaguyandrearosa: thinking about this one: https://review.openstack.org/#/c/184537/2411:08
*** rfolco has joined #openstack-nova11:08
andrearosajohnthetubaguy: not sure as it is supposed to be a bug fixing11:08
andrearosajohnthetubaguy: let me know if we need a blueprint11:08
johnthetubaguyandrearosa: I think we should add one, just to advertise it, folks have been wanting this for a long time, its good to shout about having added it11:09
johnthetubaguyandrearosa: if you can create a blueprint, give me the link and I can approve that for you11:09
*** eliqiao has quit IRC11:10
andrearosajohnthetubaguy: ok I will write a BP not now but hopefully later today. Thanks11:10
johnthetubaguyandrearosa: not sure its worth re-spinning the patch though, and loosing the +211:10
*** daemontool has quit IRC11:11
johnthetubaguyandrearosa: found it: https://blueprints.launchpad.net/nova/+spec/add-force-detach-to-nova11:11
*** markvoelker has quit IRC11:11
*** daemontool has joined #openstack-nova11:11
*** eliqiao has joined #openstack-nova11:12
*** zhangjn has joined #openstack-nova11:12
*** zhangjn has quit IRC11:12
andrearosajohnthetubaguy: oh good, I knew we had a spec (abandoned) not sure about the BP11:12
*** zhangjn has joined #openstack-nova11:13
*** zhangjn has quit IRC11:14
*** zhangjn has joined #openstack-nova11:14
*** zhangjn has quit IRC11:15
*** salv-orl_ has joined #openstack-nova11:15
*** zhangjn has joined #openstack-nova11:15
*** zhangjn has quit IRC11:15
*** zhangjn has joined #openstack-nova11:16
*** zhangjn has quit IRC11:16
*** salv-orlando has quit IRC11:17
*** zhangjn has joined #openstack-nova11:17
openstackgerritSergey Nikitin proposed openstack/nova: Replaced private field to public method in enginefacade  https://review.openstack.org/24633711:17
tangchen__alex_xu: Hi, alex, when is the deadline of the doc review work ?11:18
*** zhangjn has quit IRC11:18
*** zhangjn has joined #openstack-nova11:18
alex_xutangchen__: I guess there isn't deadline11:18
johnthetubaguyandrearosa: didn't we need to call cinder's force-detach in the end?11:18
*** zhangjn has quit IRC11:19
tangchen__alex_xu: Oh, I see. Thx. :)11:19
alex_xutangchen__: np11:19
*** zhangjn has joined #openstack-nova11:20
*** zhangjn has quit IRC11:20
*** zhangjn has joined #openstack-nova11:21
Kevin_ZhengNova jenkins python27 and python34 failed all day?11:21
tangchen__<tyagiprince>: Sorry, I didn't see your reply. Hope you have fixed your problem.11:21
andrearosajohnthetubaguy: nope, because of two main reason. 1 cinder force-detach has a bug and doens't do what it claims. I know that the cinder guys are wokring on a fix. 2. we have decided that nova needs to clean up its bits and don't bother about trying to fix cinder side. It's up to the operator to use/call the cinder tools to fix the situation in cinder.11:23
*** rmart04 has joined #openstack-nova11:23
andrearosajohnthetubaguy: not sure if I explained it very well11:23
*** daemontool has quit IRC11:23
bauzasKevin_Zheng: https://bugs.launchpad.net/nova/+bug/151680511:23
openstackLaunchpad bug 1516805 in OpenStack Compute (nova) liberty "nova unit tests failing with infinite recursion with oslo.versionedobjects 0.13.0" [Critical,In progress] - Assigned to Dan Smith (danms)11:23
*** EinstCrazy has joined #openstack-nova11:23
*** romainh has quit IRC11:23
*** daemontool has joined #openstack-nova11:24
*** sudipto has quit IRC11:24
bauzasPaulMurray: FYI https://review.openstack.org/#/c/245993/11:24
Kevin_ZhengAh, thanks11:24
Kevin_Zhengbauzas: Thanks11:24
*** daemontool has quit IRC11:25
bauzasKevin_Zheng: the fix for master is in the gate pipeline, hoping it could be merged by the next hours11:26
*** xnox has left #openstack-nova11:26
*** daemontool has joined #openstack-nova11:26
Kevin_Zhengbauzas: cool11:26
*** flaper87 has quit IRC11:27
*** flaper87 has joined #openstack-nova11:27
openstackgerritBob Ball proposed openstack/nova: XenAPI: Cope with more Cinder backends  https://review.openstack.org/24478911:28
PaulMurraybauzas, thanks11:29
*** atuvenie has joined #openstack-nova11:30
bauzasPaulMurray: btw. any idea when you'll get quotes for room blocks in Bristol ?11:31
*** salv-orl_ has quit IRC11:31
*** salv-orlando has joined #openstack-nova11:32
atuveniejohnthetubaguy: Hey, I have a trivial bug fix to make create configdrive work with py3. It's currently stuck due to a -1, can you take a look and see what you think? https://review.openstack.org/#/c/237786/11:33
*** josecastroleon has quit IRC11:36
*** ihrachys has quit IRC11:37
openstackgerritShaoHe Feng proposed openstack/nova: do not log twice with different level.  https://review.openstack.org/24626911:37
*** jhesketh has quit IRC11:37
*** paul-carlton has quit IRC11:40
*** jhesketh has joined #openstack-nova11:40
*** pixelbeat has joined #openstack-nova11:40
johnthetubaguyatuvenie: have you reached out to sdague on IRC? he is usually online quite soon11:43
*** ihrachys has joined #openstack-nova11:43
*** shz has quit IRC11:43
*** rook has quit IRC11:43
*** achanda has quit IRC11:43
openstackgerritMark McLoughlin proposed openstack/nova: servicegroup: remove the zookeeper driver  https://review.openstack.org/24634311:43
*** shz has joined #openstack-nova11:43
alex_xubauzas: currently we only use reno add upgrade note?11:44
bauzasalex_xu: you mean ?11:44
bauzasalex_xu: for the moment, we're holding off reno relnotes because we're not gating them11:44
alex_xubauzas: just want to ask whether should ask feature patch add feature note11:44
johnthetubaguyatuvenie: we also have a blueprint for all the python3 work, attaching it to the blueprint might get the attention of the folks who focus on python3 stuff: https://blueprints.launchpad.net/nova/+spec/nova-python3-mitaka11:45
alex_xubauzas: ok, got it11:45
bauzasalex_xu: yup, you should11:45
bauzasalex_xu: eg. https://review.openstack.org/#/c/197912/11:45
claudiubjohnthetubaguy: ahem, I have like 7-8 python 3 related patches that haven't been reviewed. :D11:46
alex_xubauzas: cool, thanks11:46
johnthetubaguyclaudiub: agreed, we have 600 outstanding patches right now11:46
PaulMurraybauzas, Just asked - there was a correction we had to make, will try to get it out today. It will work out around £97 per night for the hotel near us11:46
johnthetubaguyclaudiub: just giving hints for folks to reach your patches sooner11:46
atuveniejohnthetubaguy: Thanks :) I will try to reach Sean and will attach the patch to that BP11:47
PaulMurrayapprox - detail to come11:47
alex_xubauzas: for such single patch feature https://review.openstack.org/#/c/209018/, I should ask add feature note, right?11:47
johnthetubaguyclaudiub: FWIW, I saw a jenkins -1 at the base of one chain on py34 patches, so skipped that chain till thats fixed11:47
bauzasPaulMurray: I got a ? for the currency, I bet it's pounds?11:47
PaulMurraybauzas, really, ha. Yes pounds11:48
bauzasPaulMurray: cool11:48
PaulMurraywhich is not good for you Euro folks right now11:48
openstackgerritSergey Nikitin proposed openstack/nova: Replaced private field in get_session/engine with public method  https://review.openstack.org/24633711:48
bauzasPaulMurray: well, UK will get Euro soon, right? :p11:48
*** deepakcs has quit IRC11:48
*** smatzek has joined #openstack-nova11:48
*** vladikr has quit IRC11:49
bauzasPaulMurray: thanks for sorting that out11:49
PaulMurraybauzas, will send out a web site link with everything you need11:50
PaulMurrayvery soon11:50
bauzasPaulMurray: well, tbh, I'm not that worried11:50
*** ZZelle has quit IRC11:50
*** ZZelle has joined #openstack-nova11:51
*** markus_z has quit IRC11:51
*** ihrachys has quit IRC11:51
*** ihrachys has joined #openstack-nova11:52
* alex_xu reminders api meeting in 8 mins11:52
*** ihrachys has quit IRC11:52
*** markus_z has joined #openstack-nova11:53
*** jlanoux has quit IRC11:53
*** tyagiprince has joined #openstack-nova11:53
*** ndipanov has joined #openstack-nova11:54
*** paul-carlton has joined #openstack-nova11:54
sdagueatuvenie: looking, ok, explanation seems fine11:55
tyagiprinceHey nova people.. I need some help.. I created a 2 node setup and when I boot an instance, it goes into the error state.. and when i try to delete the same, it gives this error "an error occurred while refreshing the network cache."11:55
tyagiprinceThis error comes when i boot an instance..11:55
tyagiprinceERROR nova.compute.manager [req-9669c1e0-7afd-4810-988e-83f505ece99d f24656405a8648399c079e26ed0b81e6 7d5099e4dc2e4269b4632adfcc94c132 - - -] [instance: f559751c-dcd4-4ac3-a4ba-69f99a09ad5f] Failed to deallocate network for instance.11:55
atuveniesdague: Thanks!11:55
*** rpodolyaka1 has joined #openstack-nova11:57
*** jwcroppe has quit IRC11:57
*** ParsectiX has joined #openstack-nova11:57
ParsectiXHi guys just a small question. Did openstack deletes old instance info from MariaDB ?11:58
*** rpodolyaka1 has quit IRC11:59
alex_xunova api meeting is running12:00
*** venkat_p has joined #openstack-nova12:02
*** Sree_ has joined #openstack-nova12:03
*** lpetrut1 has joined #openstack-nova12:03
*** Sree_ is now known as Guest2404012:03
*** romainh has joined #openstack-nova12:04
*** lpetrut has quit IRC12:04
*** lpetrut1 is now known as lpetrut12:04
*** ndipanov has quit IRC12:05
*** ParsectiX has quit IRC12:06
*** achanda has joined #openstack-nova12:06
*** stevemar_ has joined #openstack-nova12:06
*** ParsectiX has joined #openstack-nova12:06
*** josecastroleon has joined #openstack-nova12:07
*** ParsectiX has quit IRC12:08
*** rpodolyaka1 has joined #openstack-nova12:09
*** achanda has quit IRC12:09
mdboothdanpb: I'm thinking about virStorageVolUpload in the context of a potential performance impact on https://review.openstack.org/#/c/232053/9/specs/mitaka/approved/migrate-libvirt-volumes.rst12:10
*** rushiagr is now known as rushiagr_away12:10
mdboothDo you happen to know anything about its performance?12:11
*** venkat_p has quit IRC12:11
mdboothA major concern is the lack of efficient sparse handling, which is usually a huge win when copying disk images.12:12
*** links has joined #openstack-nova12:12
mdboothRaw disk images, anyway. qcow2 is probably less affected.12:12
*** links has quit IRC12:12
mdboothI notice that you call it and it returns a stream pointer. I believe the transfer channel is multiplexed on the libvirt connection, right?12:13
*** haomaiwa_ has quit IRC12:14
*** otter768 has joined #openstack-nova12:15
mdboothIs setting up that multiplex an expensive operation? It occurs to me that if it was cheap you could simply do sparse detection at source and only transfer non-sparse segments. However, it it's expensive that would have a very high overhead, as this would result in massively more calls.12:15
*** jyuso has quit IRC12:15
*** rubasov1 has quit IRC12:17
*** rubasov has joined #openstack-nova12:19
*** otter768 has quit IRC12:19
*** ndipanov has joined #openstack-nova12:21
*** tojuvone has joined #openstack-nova12:21
mdboothdanpb: Thinking about it, if libvirt exposed a raw data channel, you could just attach rsync at both ends...12:23
*** shaohe_feng has joined #openstack-nova12:24
shaohe_fengjohnthetubaguy:  hi12:25
johnthetubaguyshaohe_feng: hi, how can I help?12:26
shaohe_fengjust want to check the meeting time12:26
danpbmdbooth: as the name suggests, the API is designing for streaming, so it doesn't do sparse detection12:26
shaohe_fenglive migration12:26
*** fawadkhaliq has quit IRC12:26
shaohe_fengjohnthetubaguy:  ^12:26
*** fawadkhaliq has joined #openstack-nova12:27
mdboothdanpb: Right, just thought I'd check. There's a flags argument, though. Don't suppose anybody's proposed anything which might be relevant in this area?12:27
*** chhavi has joined #openstack-nova12:28
johnthetubaguyshaohe_feng: not sure about the time, I would have to check the ML for details, but PaulMurray will know12:28
danpbmdbooth: at a protocol level we're just asynchronously sending the data - there's no offsets present - its assumed to be a continuous stream of data12:29
*** klkumar has quit IRC12:30
danpbmdbooth: we could theoretically extend the protocol to add the notion of a "data hole" packet which just provided a length, but no data12:30
mdboothdanpb: So, the api seems to specify an offset that the stream refers to. If it's cheap to create a stream we could simply create lots of streams.12:30
danpbso if the source detected sparseness it would put  a data hole into the stream12:30
mdboothThat would also be good.12:30
mdboothMuch better, in fact.12:30
shaohe_fengjohnthetubaguy: thank you.   let me  check it with PaulMurray12:30
shaohe_fengPaulMurray:  ping12:31
danpbnot sure off hand how hard that would be to wire up but its certainly something that could be explored12:31
*** fawadkhaliq has quit IRC12:31
*** jerrygb has joined #openstack-nova12:31
*** kylek3h has joined #openstack-nova12:31
*** kylek3h has quit IRC12:31
*** pkoniszewski has joined #openstack-nova12:31
*** kylek3h has joined #openstack-nova12:31
johnthetubaguyshaohe_feng: ah, there is a new message on the ML: http://lists.openstack.org/pipermail/openstack-dev/2015-November/079631.html12:32
danpbmdbooth:  NB I get the feeling that most clouds without shared storage are likely to be using qcow2, rather than raw12:32
danpbso while its desirable to improve performance, I don't think its the end of the world for openstack if we don't detect sparseness12:32
mdboothdanpb: Yeah, that's why I'm not totally against the idea.12:32
danpboh wait no, i'm stupid12:32
mdboothI think we should be explicit about that, though. There are going to be pathological cases.12:33
danpbwe fallocate the qcow2 file to the full size12:33
mdboothOk, it's all bad :)12:33
kashyapdanpb: Yeah, I think there's a patch to change that to use 'fallocate' option of qemu-img?12:33
* mdbooth would nak it on that basis :(12:33
mdboothIf I had a nak, that is.12:33
danpbkashyap: that's not helpful12:33
danpbmdbooth: on the flip side there's many clouds which 100% refuse to deploy rsync+ssh12:34
*** gcb has quit IRC12:34
mdboothdanpb: Yup. PaulMurray was saying.12:34
kashyapAny reason why they refuse to use rsync+ssh?12:34
danpbmdbooth: so with the choice between no feature at all, or the slow method in current libvirt, they might choose the slow method12:34
danpbkashyap: remote root access is a security risk12:34
shaohe_fengjohnthetubaguy:  got it. the time has changed. It is not  the  poll time  http://doodle.com/poll/rbta6n3qsrzcqfbn.12:35
mdboothdanpb: I don't think we can force it, though. It would be a severe performance regression for anybody who was previously using rsync.12:35
kashyapAh, right.12:35
danpbmdbooth: yeah, looks like we'd need to make choice of transports configurable12:35
danpbso people can opt in to continuing to use rsync if desired12:36
*** Guest89040 is now known as jroll12:36
danpbarguably we need to make it configurable regardless, as we need a deprecation period of at least one cycle for rsync method before we remove it12:36
danpbwe can't just force people to switch methods with no warning on upgrade12:37
mdboothThe argument in the spec is that the new method is functionally equivalent, but simpler to use, which I agree with.12:37
*** markvoelker has joined #openstack-nova12:37
mdboothOn that basis alone I don't think there's a reason to keep it. It's just the performance impact.12:37
kashyapmdbooth: You might find some commentary here relevant (comments on the left hand side) - https://review.openstack.org/#/c/171098/1/nova/virt/libvirt/driver.py,cm12:39
mdboothkashyap: Reading, thanks.12:39
kashyapYou might know most of it already.12:39
*** annegentle has joined #openstack-nova12:39
*** ihrachys has joined #openstack-nova12:40
kashyapshaohe_feng: That meeting is at 14:00 UTC.12:41
* mdbooth finds it odd that we automatically use every new version of every python package as soon as it hits pypi, but we continue using libvirt that's years old.12:42
*** markvoelker has quit IRC12:42
shaohe_fengkashyap: got it. thantk you.12:42
*** alejandrito has joined #openstack-nova12:42
kashyapmdbooth: Yeah, earlier in the day there was a topic about running newest libvirt on Gate for features that require new enough libvirt12:42
*** jgjhgsonchn88523 has joined #openstack-nova12:44
*** annegentle has quit IRC12:44
*** klkumar has joined #openstack-nova12:45
*** shaohe_feng1 has joined #openstack-nova12:45
*** ihrachys has quit IRC12:46
*** shaohe_feng has quit IRC12:46
*** annegentle has joined #openstack-nova12:47
*** pkoniszewski is now known as pkoniszewski_12:50
*** lucasagomes is now known as lucas-hungry12:52
*** baoli has joined #openstack-nova12:52
*** korzen has joined #openstack-nova12:55
*** thorst has joined #openstack-nova12:55
*** achanda has joined #openstack-nova12:56
*** ihrachys has joined #openstack-nova12:57
*** ihrachys has quit IRC12:58
*** exploreshaifali has quit IRC12:59
*** tyagiprince has quit IRC12:59
*** tyagiprince has joined #openstack-nova12:59
jichenalex_xu: maybe a quick talk about https://bugs.launchpad.net/nova/+bug/1516158?12:59
openstackLaunchpad bug 1516158 in OpenStack Compute (nova) "os-instance_usage_audit_log is used instead of os-instance-usage-audit-log" [Undecided,New] - Assigned to jichenjc (jichenjc)12:59
sdaguejichen: can you explain this more - https://bugs.launchpad.net/nova/+bug/1516158 - like what's the setup condition to get there on a devstack and what's going on?12:59
sdaguebecause I'm not really understanding the expected vs. actual behavior13:00
*** jgjhgsonchn88523 has quit IRC13:00
*** venkat_p has joined #openstack-nova13:01
jichensdague: the end point is os-instance_usage_audiog, we always use x-x-x-x in the resource type13:01
*** paul-carlton has quit IRC13:01
jichensdague: I am curious why we did so for this one only13:01
sdagueI do not know, but that's definitely the code from what I can see13:02
*** doug-fish has joined #openstack-nova13:02
jichensdague: yeah, this is from http://developer.openstack.org/api-ref-compute-v2.1.html#listusageaudits13:03
alex_xujichen: looks like we can't change that13:04
alex_xuif just for consistent13:04
sdagueyeh, I would fix the documentation13:04
sdaguethis is what it is13:05
jichenok, I can fix the doc but this is really weird compare to other resource we have13:05
*** achanda has quit IRC13:05
sdagueit is, I agree, but it's what we have13:05
jichensdague: ok, I see, I will close the bug13:06
markus_zdanpb: Hi, just to be sure, the libvirt subteam meeting is no more, right? It's still listed (with you as host) here: eavesdrop.openstack.org/#Libvirt_Meeting13:06
markus_zdanpb: Apologies If I already asked that some time ago, I'm not sure anymore13:07
sdaguejichen: well, there is still a doc change needed13:07
sdagueso don't close the bug until the doc patch is landed13:08
jichenok, I thought we tracked this in doc API change bug , and you are right, I will keep it ,thx13:08
lbelivea_ndipanov: No SR-IOV meeting today ?13:09
kashyapmarkus_z: I think it was disbaneded because not enough people were showing up consistently I guess.  But a lot of libvirt related topics gets discussed on this channel anyway.13:11
*** nkrinner has quit IRC13:12
*** gcb has joined #openstack-nova13:12
*** dave-mccowan has joined #openstack-nova13:12
markus_zkashyap: I see. I wanted to ask a question there without to make some noise here.13:14
sdaguemarkus_z: here is fine13:15
kashyapmarkus_z: It's absolutely on-topic here, not noise at all.  Type away all you want :-)13:15
*** paul-carlton has joined #openstack-nova13:15
*** pradk has joined #openstack-nova13:15
*** pradk has quit IRC13:15
markus_zOK, cool :)13:15
*** nkrinner has joined #openstack-nova13:15
markus_zI wanted to ask if someone (or more people) like to be a contact point when it comes to untriaged libvirt bugs.13:16
*** ctrath has joined #openstack-nova13:16
markus_zI guess the entry here is not up to date anymore: https://wiki.openstack.org/wiki/Nova/BugTriage#Step_2:_Triage_Tagged_Bugs13:16
*** ctrath has quit IRC13:16
*** jlanoux has joined #openstack-nova13:17
*** rushil has joined #openstack-nova13:17
*** signed8bit has joined #openstack-nova13:17
*** lbelivea_ is now known as lbelivea13:17
markus_zThe ML post 2 weeks ago didn't get much feedback: http://lists.openstack.org/pipermail/openstack-dev/2015-November/078252.html13:17
*** ff has joined #openstack-nova13:18
*** ff has quit IRC13:18
*** rook has joined #openstack-nova13:18
*** jlanoux_ has joined #openstack-nova13:19
openstackgerritjichenjc proposed openstack/nova: Change Invalid exception to a specified exception  https://review.openstack.org/23250613:20
*** rushil_ has joined #openstack-nova13:20
markus_zThat's all I wanted to say, thanks for listening :D13:20
kashyapmarkus_z: You can add me there.13:20
kashyapmarkus_z: I'm supposed to be more diligent with them, so feel free to shame me here if I'm not being so in future13:21
*** jlanoux has quit IRC13:21
*** yassine has joined #openstack-nova13:22
markus_zkashyap: Yeah, very cool, thanks!13:22
kashyapmarkus_z: I did read that email (it was not ignored).13:22
kashyapmarkus_z: I'll update the entry13:22
sdaguejichen: I uploaded a doc fix - https://review.openstack.org/#/c/246383/ - let me know what you think13:22
jichensdague: ok, I will take a look at that, thanks~13:23
*** rushil has quit IRC13:23
sdagueI think that's the right doc fix, though wadl is so hard to know13:23
*** ankit_ag has quit IRC13:26
*** gcb has quit IRC13:27
markus_zkashyap: Thanks a lot! I think you know that this doesn't mean you have to solve it by your own.13:27
kashyapmarkus_z: Of course.13:27
kashyapmarkus_z: Some of the bugs are only reproducible in OpenStack Gate env.13:27
*** markvoelker has joined #openstack-nova13:27
kashyapmarkus_z: Some have very poor info, or special storage hardware that I don't have access to, etc.13:28
*** raildo-afk is now known as raildo13:28
kashyapmarkus_z: Realize the goal is to actively triage, without letting things rot. . .13:28
markus_zkashyap: Yeah, I think I'm still to hesitant to close bugs when they don't provide enough information.13:28
kashyapmarkus_z: Well, if the reporter doesn't provide the details in a month or two, can be closed as INSUFFICIENT_INFO13:29
openstackgerritPavel Kholkin proposed openstack/nova: WIP: enginefacade: 'security_group'  https://review.openstack.org/24630213:30
sdaguehey everyone, please stop approving nova patches until https://review.openstack.org/#/c/246013/ lands, they will all fail13:30
PaulMurrayshaohe_feng1, meeting https://wiki.openstack.org/wiki/Meetings/NovaLiveMigration13:30
PaulMurraywas at lunch13:30
kashyapmarkus_z: If you haven't already seen, see some notes from sdague here: http://lists.openstack.org/pipermail/openstack-dev/2014-September/046517.html13:30
openstackgerritSean Dague proposed openstack/nova: Replace "all_mappings" variable by "block_device_mappings"  https://review.openstack.org/20851113:30
kashyapFWIW, I agree with a lot of things there.13:30
openstackgerritClaudiu Belu proposed openstack/nova: Compute: Adds driver disk_gb instance overhead estimation  https://review.openstack.org/23428313:31
openstackgerritSean Dague proposed openstack/nova: Use flavor instead of flavour  https://review.openstack.org/21864413:31
openstackgerritSean Dague proposed openstack/nova: VMware: fix docstring for cluster management  https://review.openstack.org/21864113:31
openstackgerritSean Dague proposed openstack/nova: Refactor get_filtered_hosts in filter host manager  https://review.openstack.org/21130713:31
openstackgerritSean Dague proposed openstack/nova: Sync L3Driver, NullL3 interface with LinuxNetL3  https://review.openstack.org/18900113:31
markus_zkashyap: Yepp, thanks, I've read it and it makes perfectly sense for me. My fear is to close a valid critical bug because I didn't know better.13:31
sdagueyeh, I guess mikal didn't look at the current state of the world before approving a bunch of things13:31
*** Guest24040 has quit IRC13:32
kashyapmarkus_z: Yeah, I hear you.  If it's critical enough, I'm sure enough people will yell :-)13:32
*** burgerk has joined #openstack-nova13:32
*** Sree has joined #openstack-nova13:32
openstackgerritjichenjc proposed openstack/nova-specs: Add display-flavor-id-in-migration.rst  https://review.openstack.org/21943113:33
*** rubasov has quit IRC13:33
sdaguefyi, dansmith, next time there is a break like this it should also get an email to the list to get people to stop approving code. The fix bounced and ruined everyone else's merge overnight13:33
*** rubasov has joined #openstack-nova13:33
markus_zkashyap: A previous manager I had often said "just imaging Sisyphus was a happy person" :D13:33
*** alexschm has joined #openstack-nova13:34
*** amotoki has joined #openstack-nova13:34
markus_zkashyap: Again, I'm glad you offered your help here. I'm heading back to some code stuff.13:34
*** jichen has quit IRC13:34
*** rpodolyaka1 has quit IRC13:35
kashyapmarkus_z: :-)13:35
danpbmarkus_z: correct, i just don't have the time or organization skills to rnu it13:36
*** artom__ has quit IRC13:36
danpbbest to just ask questons here, or on the mailing list13:36
*** Sree has quit IRC13:37
*** gcb has joined #openstack-nova13:40
*** cdent has joined #openstack-nova13:41
*** cdent has left #openstack-nova13:41
*** pradk has joined #openstack-nova13:43
*** oomichi has quit IRC13:43
*** pkoniszewski has joined #openstack-nova13:45
*** abhishekk has quit IRC13:45
*** markus_z_bnc has joined #openstack-nova13:45
*** haomaiwang has joined #openstack-nova13:47
sdaguemarkus_z: hey, a bug triage question. Any chance we could get per tag stats reported into graphana? It would be super nice to have some burndown stats on bugs, especially per tag.13:48
sdaguemarkus_z: also, I triaged those 4 api bugs13:48
*** paul-carlton has left #openstack-nova13:49
*** paul-carlton has joined #openstack-nova13:49
*** yassine has quit IRC13:49
*** diegows has joined #openstack-nova13:49
markus_zsdague: I haven't used graphana before. I have a poc at home which uses j3ds. That's a stretch goal I had but didn't announce yet.13:49
bauzasdo we have grafana already?13:50
markus_zsdague: And I agree that a bug dashboard would be helpful IMO.13:50
bauzasooooh http://grafana.openstack.org/13:50
markus_zbauzas: ohhhh, shiny, I'll take a look :)13:51
markus_zsdague: Thanks for the triaging of the api bugs!13:51
*** eharney has quit IRC13:52
bauzassdague: seems that the LP OAuth is not for grafana ?13:52
*** haomaiwang has quit IRC13:52
markus_zgod I loooove data visualization13:52
*** atuvenie_ has joined #openstack-nova13:53
*** samueldmq has joined #openstack-nova13:53
*** sacharya has joined #openstack-nova13:55
*** rpodolyaka1 has joined #openstack-nova13:56
*** signed8bit is now known as signed8bit_ZZZzz13:56
*** atuvenie has quit IRC13:56
*** claudiub has quit IRC13:56
openstackgerritPavel Kholkin proposed openstack/nova: enginefacade: 'bw_usage', 'vol_usage' and 's3_image'  https://review.openstack.org/24639713:57
*** jgjhgsonchn88523 has joined #openstack-nova13:57
*** claudiub has joined #openstack-nova13:57
*** markus_z is now known as markus_z_meeting13:58
sdaguemarkus_z_bnc: well, I was chatting with jeblair the other day, and that basically just consumes anything in our graphite13:58
*** moshele has quit IRC13:58
sdagueso if there was a periodic job that fed in graphite counters it could be pulled back in there13:59
*** mdrabe has joined #openstack-nova13:59
*** ijuwang has joined #openstack-nova13:59
sdagueso chatting with some infra folks about what's needed to do that, it could be a replacement for http://status.openstack.org//bugday/13:59
*** sacharya has quit IRC13:59
PaulMurrayNova Live Migration meeting on #openstack-meeting-314:00
*** jlanoux_ has quit IRC14:00
*** signed8bit_ZZZzz is now known as signed8bit14:00
*** obondarev_ is now known as obondarev14:01
*** klkumar has quit IRC14:01
*** ekarlso has joined #openstack-nova14:01
*** jlanoux has joined #openstack-nova14:01
ndipanovlbelivea, I think we said every second week?14:02
*** lucas-hungry is now known as lucasagomes14:03
kashyapmdbooth: ^ (from Paul) Guess you wanted to join it, you asked here the other day?14:03
kashyapAh, you waved there, disregard me.14:03
*** rubasov has quit IRC14:04
*** vladikr has joined #openstack-nova14:04
*** cdent has joined #openstack-nova14:04
*** _gryf has quit IRC14:04
*** claudiub has quit IRC14:04
*** _gryf has joined #openstack-nova14:05
*** _gryf has quit IRC14:05
*** claudiub has joined #openstack-nova14:05
*** rubasov has joined #openstack-nova14:05
*** _gryf has joined #openstack-nova14:06
*** rushil_ has quit IRC14:07
*** READ10 has joined #openstack-nova14:07
*** rlrossit has joined #openstack-nova14:10
*** venkat_p has quit IRC14:13
lbeliveandipanov: Ok my bad I taught it was every week14:13
*** rushiagr_away is now known as rushiagr14:14
*** burgerk has quit IRC14:15
*** otter768 has joined #openstack-nova14:15
*** kfarr has quit IRC14:16
openstackgerritjichenjc proposed openstack/nova: Add DiskFormat as Enum in fields  https://review.openstack.org/22985414:16
openstackgerritjichenjc proposed openstack/nova: Make ImageMeta:disk_format use DiskFormatField  https://review.openstack.org/23146014:16
*** ctrath has joined #openstack-nova14:16
*** cdent has left #openstack-nova14:17
lbeliveandipanov: What is your plan for sriov-physical-function-passthrough ?  We might be interested to participate in the design/implementation.14:17
*** signed8bit is now known as signed8bit_ZZZzz14:19
*** moshele has joined #openstack-nova14:19
*** Marga_ has joined #openstack-nova14:19
*** otter768 has quit IRC14:20
*** signed8bit_ZZZzz is now known as signed8bit14:21
ajoobondarev , rdopiera14:23
ajoon live migration VM resume / RARP supression14:23
ajowhat's the nova status?14:23
ajoobondarev , I think you told me there were patches already?14:24
ndipanovlbelivea, that is a good question14:24
*** vlaza has joined #openstack-nova14:24
*** pratikmallya has joined #openstack-nova14:25
*** venkat_p has joined #openstack-nova14:25
dansmithsdague: okay14:25
*** tyagiprince has quit IRC14:26
*** yassine has joined #openstack-nova14:28
ndipanovso my current plan is to just get the Nova internal bits ready for the fact that pfs can be assigned14:29
ndipanovlbelivea, ^14:29
ndipanovthere is some work to be done on top of that, and on the neutron side as well14:29
*** liverpooler has quit IRC14:29
*** smatzek has quit IRC14:29
openstackgerritGábor Antal proposed openstack/nova: Replaced deprecated timeutils methods  https://review.openstack.org/24117914:29
*** klkumar has joined #openstack-nova14:30
ndipanovlbelivea, vladikr, was looking into that14:31
*** mriedem has joined #openstack-nova14:32
moshelendipanov: can you review https://review.openstack.org/#/c/227160/ ?14:32
mriedemhttps://review.openstack.org/#/c/246013/ is still in the gate?14:33
*** neiljerram_bb has joined #openstack-nova14:33
*** dustins has joined #openstack-nova14:33
*** yassine has quit IRC14:34
ndipanovmoshele, hmmm I thought that sriov_numfs was only an ixgbe thing...14:34
*** yassine has joined #openstack-nova14:34
*** ttx has quit IRC14:36
*** breitz has quit IRC14:36
*** breitz has joined #openstack-nova14:36
moshelendipanov: I think it is a standard we have it on the mellanox driver and also baoli +1 it so I guess it good for cisco14:37
ndipanovmoshele, fair enough...14:38
*** haomaiwang has joined #openstack-nova14:39
*** nikhil_k is now known as nikhil14:39
*** mrkz has joined #openstack-nova14:40
*** ekarlso has quit IRC14:40
*** ttx has joined #openstack-nova14:40
*** doug-fish has quit IRC14:42
*** eharney has joined #openstack-nova14:43
*** artom has joined #openstack-nova14:44
*** salv-orl_ has joined #openstack-nova14:45
lbeliveandipanov: Let's talk about it next SR-IOV meeting.  I'll put together some more comments about the BP by then.14:46
*** burgerk has joined #openstack-nova14:47
*** fawadkhaliq has joined #openstack-nova14:48
ndipanovlbelivea, ok but what did you have in mind when it comes to that particular BP?14:48
*** salv-orlando has quit IRC14:48
ndipanovdo you see some issues with the proposed solution, or are you looking to add more enhancements?14:49
* ndipanov maybe we do need to have a meeting every week in the beginning?14:49
lbeliveandipanov: I had a chat with our architect and we might have some customers that are interested in something similar.  I need to gather input first.14:49
ndipanovlbelivea, ah ok awesome! more use cases sound excellent!14:50
*** burt has joined #openstack-nova14:50
*** jlanoux has quit IRC14:50
lbeliveandipanov: I would not disagree with weekly meetings.14:50
cfriesen__hopefully a quick question.  Do we expect to be allowed to delete a flavor at any point, regardless of whether it's in use by an instance?14:51
*** jlanoux has joined #openstack-nova14:52
cfriesen__I'm looking at nova.conductor.manager.ComputeTaskManager.migrate_server() and it does a call to objects.Flavor.get_by_id().  If the flavor is deleted, I think that could cause problems.14:52
*** Marga_ has quit IRC14:52
*** hamzy has quit IRC14:53
lbeliveandipanov: Side question - do you know the historic reason why nova networking information wrt to ports are stored in neutron and not in nova (like PCI address, etc.) ?  You mention something about nova interface with neutron ...14:54
*** doug-fish has joined #openstack-nova14:54
*** mriedem has quit IRC14:55
*** rubasov has quit IRC14:55
ndipanovlbelivea, so wrt weekly meetings - let's leave it like this for this week, and if we see next week that we are hitting the hour with plenty of stuff to talk about - we'll increase the freq14:56
*** rubasov has joined #openstack-nova14:57
*** yonglihe has joined #openstack-nova14:57
*** doug-fish has quit IRC14:57
*** sneti has joined #openstack-nova14:58
*** achanda has joined #openstack-nova14:58
*** jwcroppe has joined #openstack-nova14:58
*** doug-fish has joined #openstack-nova14:59
*** zhangjn has quit IRC14:59
*** doug-fish has quit IRC14:59
*** smatzek has joined #openstack-nova14:59
*** doug-fish has joined #openstack-nova15:00
*** ctrath has quit IRC15:00
*** doug-fish has quit IRC15:01
*** jwcroppe has quit IRC15:01
*** jwcroppe has joined #openstack-nova15:02
*** jwcroppe has quit IRC15:02
moshelendipanov: just to understand the process now that we have focus of SR-IOV in nova, should  we had our bp/commits  in https://etherpad.openstack.org/p/mitaka-nova-priorities-tracking ?15:02
ndipanovmoshele, yes15:02
ndipanovmoshele, the idea is though15:02
ndipanovthat you put there only stuff that is "ready"15:02
moshelendipanov: also who is include in the sub team? can I add myself ?15:02
ndipanovto be reviewed by cores15:03
ndipanovso don't add the patches there as soon as you post them15:03
ndipanovbut let's get some team folks to review them first15:03
ndipanovmoshele, yes of course15:03
ndipanovand lbelivea15:03
*** mriedem has joined #openstack-nova15:03
ndipanovand I assume there will be more interested folks that just missed the first meeting15:04
moshelendipanov: ok I will update the etherpad15:04
lbeliveasame here15:04
*** venkat_p has quit IRC15:04
*** venkat_p has joined #openstack-nova15:05
*** tdurakov has joined #openstack-nova15:06
PaulMurraytdurakov, go ahead15:06
*** ccarmack has joined #openstack-nova15:07
tdurakovupdate for live-migration job: initial part is merged to project-config and it's possible to schedule it with check experimental. There is currently bug in devstack-gate, that i'm going to submit in 15 minutes, after that everything should be ok. next step would be shared storage15:07
mriedemperformance team meeting in #openstack-performance if anyone is interested,15:07
mriedemthere is a topic on nova-conductor in there15:07
mriedemalaski: dansmith: ^15:08
tdurakovi'd prefer to address nfs first, but it's up to you to choose one15:08
tdurakovPaulMurray, ^15:08
*** ctrath has joined #openstack-nova15:08
pkoniszewskiwell, nfs is a tricky one, sounds good to me15:09
pkoniszewskigreat to hear that we are close15:09
pkoniszewskicould you add to me review when you submit a fix?15:09
PaulMurraytdurakov, to be honest, you're doing the work so you get to pick the order15:10
*** amotoki_ has joined #openstack-nova15:10
tdurakovpkoniszewski sure, it's easy one, just string reordering15:10
tdurakovPaulMurray, ok, then it gonna be nfs first:)15:10
*** annegentle has quit IRC15:11
PaulMurraytdurakov, glad to make you happy :)15:11
*** blaisebool has joined #openstack-nova15:11
*** amotoki has quit IRC15:11
*** tongli has joined #openstack-nova15:11
bauzasalaski: had a few questions on https://review.openstack.org/#/c/211753/2915:11
*** annegentle has joined #openstack-nova15:11
PaulMurraytdurakov, let jlanoux know if there is anything he can do to help - but be warned, we don't generally use shared storage15:11
openstackgerritMark McLoughlin proposed openstack/nova: servicegroup: remove the zookeeper driver  https://review.openstack.org/24634315:11
bauzasalaski: in particular about setting the id field15:12
tdurakovbtw, here is the problem, first site to check timezone forget about daylight saving:http://www.timeanddate.com/worldclock/fixedtime.html?month=10&day=14&year=2005&hour=14&min=0&sec=0&p1=015:12
*** READ10 has quit IRC15:13
*** venkat_p has quit IRC15:13
*** amotoki_ is now known as amotoki15:13
tdurakovPaulMurray, i've already set up nfs for devstack, so it should be ok, jlanoux already helped me:) so yep15:13
*** venkat_p has joined #openstack-nova15:14
*** sneti has quit IRC15:14
*** sneti_ has joined #openstack-nova15:14
PaulMurraytdurakov, good - glad you are getting this moving - its very important15:14
alaskibauzas: okay, looking15:15
*** pumaranikar has joined #openstack-nova15:15
tdurakovPaulMurray, yes15:15
obondarevrdopiera: ping15:15
*** mdrabe has quit IRC15:15
rdopieraobondarev: pong15:16
* PaulMurray heads off in search of coffee15:16
obondarevrdopiera: ajo said you're looking into https://bugs.launchpad.net/neutron/+bug/1414559 ?15:16
openstackLaunchpad bug 1414559 in neutron "OVS drops RARP packets by QEMU upon live-migration - VM temporarily disconnected" [Undecided,Invalid] - Assigned to Oleg Bondarev (obondarev)15:16
obondarevrdopiera: do you?15:16
*** Guest80875 is now known as mfisch15:16
*** mfisch is now known as Guest2059415:17
rdopieraobondarev: not at this exact moment, but I do want to fix it15:17
obondarevrdopiera: just wanted to sync with you, as I have neutron and nova patches which I'm testing right now15:17
*** Guest20594 is now known as mfisch15:18
obondarevrdopiera: should publish them by tomorrow eod15:18
*** mfisch has quit IRC15:18
*** mfisch has joined #openstack-nova15:18
rdopieraobondarev: that's awesome, please let me know when you do and I will test too15:18
obondarevrdopiera: thanks, will do15:19
*** venkat_p has quit IRC15:19
*** jaypipes has quit IRC15:19
*** venkat_p has joined #openstack-nova15:19
*** claudiub has quit IRC15:20
rdopieraobondarev: funny how it is marked as "invalid"15:20
*** doug-fish has joined #openstack-nova15:21
obondarevrdopiera: not easy to reproduce I guess15:21
*** dustins has quit IRC15:22
*** mtanino has joined #openstack-nova15:23
*** obondarev is now known as obondarev_afk15:23
*** jwcroppe has joined #openstack-nova15:24
*** andreykurilin__ has joined #openstack-nova15:24
alaskibauzas: commented.  I'll double check on 'id', but I was under the impression that sqlalchemy handled that15:25
bauzasalaski: that's my open question :)15:25
bauzasalaski: if the SQLA object is updated by calling .save(), I'm fine15:25
bauzasalaski: tbh, I was lazy, I should have verified by myself :/15:26
*** venkat_p has quit IRC15:26
*** venkat_p has joined #openstack-nova15:26
*** dustins has joined #openstack-nova15:27
alaskiit's my code, I should be able to defend what it does15:27
*** tonytan4ever has joined #openstack-nova15:28
*** garyk has quit IRC15:29
*** doug-fish has quit IRC15:29
*** ctrath has quit IRC15:29
*** doug-fish has joined #openstack-nova15:30
*** moshele has quit IRC15:31
*** doug-fis_ has joined #openstack-nova15:31
*** doug-fi__ has joined #openstack-nova15:33
*** doug-fi__ has quit IRC15:33
*** mc_nair has quit IRC15:33
*** doug-fi__ has joined #openstack-nova15:34
*** ctrath has joined #openstack-nova15:34
*** dkliban has joined #openstack-nova15:34
*** doug-fish has quit IRC15:34
BobBallAnyone know of a new bug for oslo.versionobjects recursion depth failures? https://bugs.launchpad.net/nova/+bug/1470302 It's constantly killing https://review.openstack.org/#/c/244789/ and seems to be showing up in logstash quite a lot ATM (based on query from 1470302).  I've not found a new bug yet, but it seems to be happening too regularly for there not to be a bug already?15:34
openstackLaunchpad bug 1470302 in OpenStack Compute (nova) "gate-nova-python27 fails with "RuntimeError: maximum recursion depth exceeded"" [High,Fix released] - Assigned to Dan Smith (danms)15:34
*** daemontool has quit IRC15:34
dimsBobBall : https://review.openstack.org/#/c/246013/ is about to merge15:35
BobBallah awesome15:35
bauzasdims: hopefully :)15:36
*** doug-fis_ has quit IRC15:36
*** vlaza has quit IRC15:36
bauzas20 mins left :)15:36
*** klkumar has quit IRC15:36
*** dkliban has left #openstack-nova15:36
* BobBall will be patient15:36
*** markus_z_meeting is now known as markus_z15:37
*** mdrabe has joined #openstack-nova15:37
tdurakovpkoniszewski, added you for review15:37
*** pratikma_ has joined #openstack-nova15:38
*** klkumar has joined #openstack-nova15:39
*** pratikmallya has quit IRC15:41
*** paul-carlton has quit IRC15:43
*** amotoki_ has joined #openstack-nova15:44
*** garyk1 has quit IRC15:45
*** kfarr has joined #openstack-nova15:45
*** amotoki has quit IRC15:47
*** pratikma_ is now known as pratikmallya15:48
*** bnemec has quit IRC15:48
*** electrocucaracha has joined #openstack-nova15:49
*** armax has quit IRC15:49
*** Piet has quit IRC15:50
*** mc_nair has joined #openstack-nova15:51
*** venkat_p has quit IRC15:51
*** diegows has quit IRC15:51
*** venkat_p has joined #openstack-nova15:51
*** dustins has quit IRC15:52
*** dustins has joined #openstack-nova15:52
*** venkat_p has quit IRC15:53
*** tdurakov has quit IRC15:56
*** vilobhmm has joined #openstack-nova15:56
*** daemontool has joined #openstack-nova15:57
*** EinstCrazy has quit IRC15:57
openstackgerritJim Rollenhagen proposed openstack/nova-specs: Tenant networking support for Ironic driver.  https://review.openstack.org/23706715:58
mriedemdansmith: i was going to add a note to the kilo release notes that the flavor migrations will put extra load on conductor so you may need to scale those out until liberty, are you ok with that?15:58
*** vilobhmm1 has joined #openstack-nova15:58
*** atuvenie_ has quit IRC15:58
dansmithmriedem: yeah15:58
*** paul-carlton has joined #openstack-nova15:58
*** daemontool has quit IRC15:59
*** jaypipes has joined #openstack-nova15:59
*** daemontool has joined #openstack-nova15:59
*** kfarr has quit IRC16:00
mriedemdone https://wiki.openstack.org/wiki/ReleaseNotes/Kilo#Upgrade_Notes_216:00
mriedemmay not need to be bold, idk16:01
*** mgoddard_ has joined #openstack-nova16:01
*** mjura has quit IRC16:01
*** vilobhmm has quit IRC16:01
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: review guide: add location details for config options  https://review.openstack.org/24578916:02
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: config options: centralize section "serial_console"  https://review.openstack.org/24417716:02
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: config options: enhance help test of section "serial_console"  https://review.openstack.org/24646516:02
dansmithmriedem: yep, seems fine16:02
*** bnemec has joined #openstack-nova16:03
*** mgoddard has quit IRC16:03
*** tdurakov has joined #openstack-nova16:03
*** diegows has joined #openstack-nova16:04
*** rcernin has quit IRC16:04
dansmithklindgren: now that you're all done with your migrations on kilo, you have your [upgrade_levels] all cleared out right?16:04
bauzasdansmith: mriedem: correct me if I'm wrong but getting a GuruMeditationReport for the conductor could help finding if most of the green threads are running the flavor mig, nope ?16:04
*** diegows has quit IRC16:04
dansmithbauzas: if you get lucky, maybe16:05
dansmithbauzas: depends on what the problem is16:05
bauzasdansmith: for sure, I'm thinking of long-running tasks16:05
johnthetubaguyPaulMurray: re your questions on error handling in those live-migrate specs, I think we should use this: http://developer.openstack.org/api-ref-compute-v2.1.html#os-instance-actions-v2.116:05
dansmithbauzas: yeah, I just don't know where the problem is16:05
bauzasdansmith: agreed, that's just something really quick to do without really needing to profile16:06
bauzasI meant, not sure if klindgren is knowing that16:06
dansmithyep, if they're all in the same function it might help16:06
*** mriedem has quit IRC16:06
*** tdurakov has quit IRC16:06
markus_zjohnthetubaguy: That's the split up in the "cofig options" change we talked about yesterday: https://review.openstack.org/#/c/244177/4  Let me know if you think this is the right granularity.16:07
*** gcb has quit IRC16:07
johnthetubaguymarkus_z: ah, so you caught me at a good point, I will take a quick peak now16:07
*** ildikov has quit IRC16:07
klindgrendansmith, yea we made sure of that a few months ago.  We don't have any lingering upgradelevels16:09
dansmithklindgren: okay16:09
dansmithklindgren: can you check your system_metadata table to see  if you have any instances with non-deleted instance_type_* keys?16:10
sdaguecburgess_: how's that fix / workaround coming for ebtables?16:10
*** mriedem has joined #openstack-nova16:11
klindgrendansmith, sure - one second16:12
PaulMurrayjohnthetubaguy, I think I wrote a confusing comment16:12
lxslisdague: mind checking https://review.openstack.org/#/c/215568 please? Has a few +1s, been open for a while. The followup patch too if you'd be so kind :)16:13
*** Yingxin1 has joined #openstack-nova16:13
PaulMurrayjohnthetubaguy, I meant that a failure of the migration itself would not need to be reported because it is already being reported by existing code16:13
*** belmoreira has quit IRC16:13
PaulMurrayjohnthetubaguy, the failure to initiate the pause would need to be reported as you say16:13
PaulMurrayjohnthetubaguy, the question was about should a failure to pause also fail the migration16:14
johnthetubaguyPaulMurray: ah, it sort of is, but I think we might be missing the instance action at the moment.16:14
PaulMurrayjohnthetubaguy, and I think the answer is no16:14
johnthetubaguyPaulMurray: so it depends, the pause might trigger the live-migrate to fail, in which case you might end up with both failing, and the server in an ERROR state16:15
openstackgerritYingxin Cheng proposed openstack/nova: Use stevedore for scheduler host manager  https://review.openstack.org/24647616:15
johnthetubaguyPaulMurray: if the pause fails, but live migration succeeds, thats fine, only report the failed pause16:15
openstackgerritAlexis Lee proposed openstack/nova: Remove SoftDeleteMixin from BandwidthUsage  https://review.openstack.org/24036116:15
*** rubasov has quit IRC16:16
*** mlavalle has joined #openstack-nova16:16
mlavallejohnthetubaguy: ping16:16
PaulMurrayjohnthetubaguy, Is it part of this spec if we need to add instance action to the live migration16:16
*** otter768 has joined #openstack-nova16:16
johnthetubaguyPaulMurray: that can probably be a separate spec, probably best to concentrate on the pause16:16
johnthetubaguyPaulMurray: same argument for cancel of course16:17
sdaguelxsli: looking16:17
PaulMurrayjohnthetubaguy, sounds good16:17
johnthetubaguymarkus_z: thats it, got a few nits I am adding, but yeah, I think thats easier to review16:17
johnthetubaguymlavalle: hows things?16:17
bauzasso, the nova patch for o.vo got rebalanced :(16:18
mlavallejohnthetubaguy: great, thanks for the review to https://review.openstack.org/#/c/90150/21/specs/mitaka/approved/internal-dns-resolution.rst16:18
*** mwagner has quit IRC16:18
vilobhmm1johnthetubaguy : ping this spec has couple of reviews already https://review.openstack.org/#/c/222423/ if the cores find it ok…and if this gets merged I can submit the code16:18
mlavallejohnthetubaguy: do you have a few minutes to discuss comment to line 122?16:18
johnthetubaguymlavalle: sure thing16:19
*** amotoki_ has quit IRC16:19
mlavallejohnthetubaguy: I don't quite understand your comment. Would you elaborate a little bit?16:19
johnthetubaguyvilobhmm1: sorry, not sure I understand the question16:19
johnthetubaguymlavalle: ah, right16:19
dansmithbauzas: rebalanced?16:20
vilobhmm1johnthetubaguy : https://review.openstack.org/#/c/222423/  needs one more review from the core member16:20
johnthetubaguymlavalle: so I am not keen on us detected when the passed in host name is a FQDN16:20
johnthetubaguymlavalle: it feels a little odd somehow16:20
vilobhmm1to be merged16:20
vilobhmm1this is the servicegroup spec16:20
openstackgerritBalazs Gibizer proposed openstack/nova: Add service status notification  https://review.openstack.org/24567816:20
sdaguebauzas: yeh, the libvirt ebtables race is the primary culprit (also folks approving nova patches overnight, which basically halted all gate progress until we kicked them all out this morning)16:20
bauzasdansmith: I meant back in the zuul queue16:20
johnthetubaguyvilobhmm1: ah, yes, it does, hopefully folks will get to that sooner now it has a +2 on it16:21
sdaguethe heat job that just failed, failed on the ebtables issue16:21
sdaguewhich reset us again16:21
johnthetubaguymlavalle: so I had a few follow on thoughts from that, which go on a bit16:21
*** otter768 has quit IRC16:21
bauzassdague: okay, thanks for the heads-up16:21
mlavallejohnthetubaguy: I am not proposing any change to nova's behaviour. All I am going to do is to take the data in the instance.hostname and send it to Neutron. Are we clear on that?16:21
vilobhmm1johnthetubaguy : ok..lets hope so :)16:21
johnthetubaguymlavalle: first was, well its not really part of this spec, as thats neutron behaviour, so I guess Nova just passes that stuff to Neutron16:22
*** _d34dh0r53_ is now known as d34dh0r5316:22
dansmithsdague: you haven't sent out an email about not approving nova jobs.. do you still think we should do that?16:22
johnthetubaguymlavalle: yep, the basic idea seems sound16:22
bauzasdansmith: all the changes are impacted, right?16:22
sdaguewell, this is the top nova patch right now, let's give it an hour and see what happens16:22
lxslijohnthetubaguy: I discussed an oslo.config opt flag "becoming default in next release" with someone, was it you?16:23
*** thangp has joined #openstack-nova16:23
dansmithsdague: I guess I didn't see the point since nothing should be getting into the gate queue because this is 100% fail16:23
*** alex_klimov has quit IRC16:23
johnthetubaguylxsli: unsure16:23
sdaguedansmith: because we don't check for fresh test results16:23
sdagueso mikal went through and approved a bunch of things that were jenkins +116:23
dansmithsdague: okay so a few things can make it in16:23
mlavallejohnthetubaguy: on the neutron side there is a domain defined. All I am saying is that if the trailing part of the hostname from nova matches the domain in Neutron, I am going to consider it a FQDN.16:23
johnthetubaguymlavalle: sorry, not making this very clear yet...16:23
johnthetubaguymlavalle: yeah, the problem is the official list of top level domains is expanding these days, etc, it seems a little odd16:24
johnthetubaguymlavalle: what does cloud-init do with that today? I guess it just sets the hostname to what is specified, its possible we trim the hostname before passing it there, I don't remember now16:24
*** rpodolyaka1 has quit IRC16:24
sdagueyeh, about ~30 patches - https://review.openstack.org/#/q/project:openstack/nova+label:Verified%253E%253D1%252Cjenkins+status:open+label:Code-Review%253E%253D2,n,z16:25
johnthetubaguymlavalle: anyways, I think the easy fix is to remove that detail from the nova spec, as its technically neutron behaviour, but honestly it sounds/feels like it will create a whole heap of bugs and issues with miss-matches between what happens today with cloud-init, etc, and what will happen with the configured DNS. I could be over thinking it though.16:26
*** salv-orl_ has quit IRC16:26
*** hemnafk is now known as hemna16:26
dansmithsdague: okay well I didn't think it was that big of a deal because I expected check to run on most everything16:26
*** mspreitz has joined #openstack-nova16:26
*** nkrinner has quit IRC16:27
bauzasnope, it's no longer the case AFAICS16:27
bauzasit only verifies a +1 from Jenkins, not the age16:27
dansmithI'm pretty sure it considers the age16:28
dansmithsdague: ?16:28
*** READ10 has joined #openstack-nova16:28
sdagueit does not16:28
*** armax has joined #openstack-nova16:28
bauzasdansmith: I can't see https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L34-L6816:28
dansmithI can't keep it straight16:28
sdaguethat was changed sometime last spring16:28
*** shaohe_feng1 has quit IRC16:28
klindgrendansmith, select * from instance_system_metadata where deleted_at is null and `key` LIKE 'instance_type_%';16:28
klindgren= 15670 rows16:29
mlavallejohnthetubaguy: so, the host name set by cloud-init in the instance might not be the same thing as instance.hostname in nova?16:29
dansmithklindgren: are those instances deleted?16:29
dansmithklindgren: I don't think you want deleted for the sysmeta row, but for the instance16:29
johnthetubaguymlavalle: unsure, I was just checking this logic: https://github.com/openstack/nova/blob/aa141f0d8ea90a082558a9969340ba655a07c417/nova/utils.py#L774 going a bit deeper16:30
mlavallejohnthetubaguy: but is that the possibility that worries you?16:30
*** rpodolyaka1 has joined #openstack-nova16:30
johnthetubaguymlavalle: right now, yeah16:30
mlavallejohnthetubaguy: ok, let me dig a little deeper. I'll get back to you in a bit16:31
PaulMurraypkoniszewski, are you still around16:31
lxslijohnthetubaguy: looking at https://wiki.openstack.org/wiki/Nova/Process again - it's OK for me to just edit it?16:31
*** Piet has joined #openstack-nova16:31
lxslito get https://review.openstack.org/#/c/206070 off my plate16:32
johnthetubaguylxsli: yeah, honestly I have a TODO to add that into devref so its in the tree rather than the wiki16:32
lxslijohnthetubaguy: probably easiest for me to edit the wiki then convert to devref than try to do both at once though16:33
johnthetubaguylxsli: +116:33
lxslijohnthetubaguy: cool, thanks16:33
johnthetubaguylxsli: edit the wiki, then convert, makes sense I think16:33
*** josecastroleon has quit IRC16:33
*** rk4n has quit IRC16:33
*** mspreitz has quit IRC16:34
*** rk4n has joined #openstack-nova16:34
*** gokrokve has joined #openstack-nova16:35
*** pkoniszewski has quit IRC16:36
johnthetubaguymlavalle: seems like cloud-init considers everything after the . as an FQDN: http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/util.py#L93216:37
* mlavalle looking16:37
*** armax has quit IRC16:37
bauzasjohnthetubaguy: FWIW, will respin https://review.openstack.org/#/c/221280/16:38
*** rk4n has quit IRC16:38
bauzasalaski: I'm also tempted to resubmit the very old Tasks spec, you happy with that?16:38
johnthetubaguymlavalle: seems like this bit is also important: http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/view/head:/cloudinit/sources/__init__.py#L17116:39
*** armax has joined #openstack-nova16:40
*** rk4n has joined #openstack-nova16:40
mlavallejohnthetubaguy: thanks, going through it16:40
*** e0ne has quit IRC16:41
*** browne has joined #openstack-nova16:42
johnthetubaguymlavalle: cool, I am leaning towards treating everything with a "." in it as an FQDN, but I think for the sake of the Nova spec, you could just say Neutron is planning to be consistent with cloud-init's use, or something like that?16:42
sdaguedoes anyone know netfilter folks?16:43
sdagueit seems a little odd that ebtables requires all users of it to say "please don't punch me in the face" to not be punched in the face16:43
*** doug-fi__ is now known as doug-fish16:43
sdaguei.e. why is --concurrent even an option vs. just the default16:43
*** rk4n has quit IRC16:44
*** rk4n has joined #openstack-nova16:44
*** rmart04 has quit IRC16:45
mlavallejohnthetubaguy: ah, ok. Upon encountering the first '.', then we assume it is a fqdn. right? the hostname is the string before the first '.'. correct?16:46
*** gjayavelu has joined #openstack-nova16:46
*** rk4n has quit IRC16:48
danpbsdague: looking at the code they wanted to be able to consider failure to lock to be a fatal error when --concurrent is used16:49
*** sahid has quit IRC16:49
danpbsdague: if they had made it the default, then this would have caused a behaviour change for existing users in some cases16:50
sdaguedanpb: hmmm, yeh, it's just unfortunate, because it means if another tool shows up that uses it, then you get a race again16:51
*** njirap has joined #openstack-nova16:51
alaskibauzas: go for it16:51
bauzasalaski: you made good points in the manifesto, I first need to respin it16:52
sdagueand there is no way to specify at a system level, always lock16:52
bauzasalaski: I just had one question for you, with "This is only true for API instantiated actions.  Something like an evacuate would not be tracked."16:52
*** vilobhmm1 has quit IRC16:52
bauzasalaski: I don't understand, evacuate is API-triggered, right?16:52
*** electrocucaracha has quit IRC16:52
*** unicell has quit IRC16:52
johnthetubaguymlavalle: yeah, that sounds about right16:53
*** e0ne has joined #openstack-nova16:53
bauzasalaski: sorry, the context being https://review.openstack.org/#/c/221280/3..4/doc/source/tasks.rst,cm L3616:54
bauzashttps://review.openstack.org/#/c/221280/3/doc/source/tasks.rst,cm L3616:54
alaskibauzas: yeah, not sure what I was thinking there.  what I meant to say was that anything initiated by a periodic task wouldn't be tracked16:54
bauzasalaski: oh yeah, for sure16:54
mlavallejohnthetubaguy: it actually makes that whole section about fqdn and pqdn irrelevant in the nova spec. in hostname I am just getting the first label. I should remove that section and just indicate that we are going to comply with cloud-init. Makes sense?16:55
bauzasalaski: you probably meant when the compute is restarted, and the evacuations are checked16:55
alaskibauzas: yeah, that.  and checks for destroyed instances as well16:55
*** vishwana_ is now known as vishwanathj16:55
bauzasalaski: okay, I will make that point clear16:56
bauzasalaski: but I don't remember periodic tasks changing an instance state ?16:56
alaskibauzas: it could confirm a resize16:58
alaskior offload a shelved instance16:58
alaskiand I just noticed the virt lifecycle events which aren't covered16:58
openstackgerritArtom Lifshitz proposed openstack/nova: Fix max concurrent builds's unlimited semaphore  https://review.openstack.org/23905416:58
*** mriedem is now known as mriedem_meeting16:59
bauzasalaski: could you please leave a comment accordingly ?16:59
*** angdraug has quit IRC16:59
*** njirap has quit IRC16:59
*** ctrath1 has joined #openstack-nova17:00
*** ljxiash has joined #openstack-nova17:00
*** njirap has joined #openstack-nova17:00
*** korzen_ has joined #openstack-nova17:01
*** korzen has quit IRC17:02
*** ctrath has quit IRC17:02
*** mwagner has joined #openstack-nova17:03
*** daemontool has quit IRC17:04
*** sacharya has joined #openstack-nova17:04
*** daemontool has joined #openstack-nova17:04
*** tpeoples_away is now known as tpeoples17:04
openstackgerritArtom Lifshitz proposed openstack/nova: Fix max concurrent builds's unlimited semaphore  https://review.openstack.org/23905417:05
*** rushil has joined #openstack-nova17:05
mlavallejohnthetubaguy: in light of this exchange, I realize that the whole section about PQDN and FQDN was really rambling. It is irrelevant on the Nova side. I'll remove it. Thanks for the guidance! :-)17:05
*** ctrath1 has quit IRC17:05
*** rlrossit has quit IRC17:05
johnthetubaguymlavalle: no problems, yeah, I think just dropping it is probably best. Less is often more when it comes to specs.17:06
*** mgoddard_ has quit IRC17:06
*** jlanoux has quit IRC17:07
*** yassine has quit IRC17:07
cburgess_sdague Working on it today, should have it up soon. Did have 1 question when you get a minute.17:08
edleafemarkus_z: still around? Got a question about the centralized conf stuff17:09
dansmithklindgren: are you still working on trying to do that query?17:09
markus_zedleafe: yepp, I'm here17:09
openstackgerritMoshe Levi proposed openstack/nova: libvirt: mlnx_direct vif type removal  https://review.openstack.org/18167117:10
klindgrenI was but - fires came up - I am not the best at cross table joins.  By best I mean - its been a long time since i had to do something more complex than a select <stuff> from <table> where <blah>;17:11
klindgrenif you have a querry I can run it - but working on a prod issue17:11
edleafemarkus_z: so should all conf options under, say, nova/scheduler/ be moved to one file: nova/conf/scheduler.py? Or should it be nova/conf/scheduler/manager.py, etc?17:11
edleafemarkus_z: IOW, one conf file for each original file?17:12
*** rubasov has joined #openstack-nova17:12
dansmithklindgren: well, if you want to make sure that something else is not going on, it'd be good to get that when you have a chance17:13
openstackgerritSamuel Matzek proposed openstack/nova: Reschedule boot when attach volume fails  https://review.openstack.org/24650517:13
*** blaisebool has quit IRC17:13
*** rlrossit has joined #openstack-nova17:13
markus_zedleafe: The name of the "nova.conf" section (like: "[DEFAULT]") determines the name of the new module17:13
*** e0ne has quit IRC17:13
markus_zI see now that there is no "[scheduler]" section, is that right?17:13
dansmithklindgren: I don't really want to try to make a query for you to run that could have issues, but basically joining instances with sysmeta on uuid, where instance is not deleted and the key is as you had it..17:13
edleafemarkus_z: not following17:14
markus_zedleafe: sorry, I look to your patch set again to pick a better example17:14
sdaguecburgess_: sure, fire away17:14
sdaguealso, I wrote the terrible part of the work around for devstack right now17:14
cburgess_sdague OK17:15
sdaguebecause this is hurting us pretty hard in landing patches17:15
cburgess_sdague So... there is no timeout for ebtables. What I mean is if you invoke it with --concurrent then you will wait.. forever.17:15
dansmithsdague: can you make that only happen if some config is set? so we don't all destroy our machines if we run that?17:15
*** unicell has joined #openstack-nova17:15
cburgess_sdague Oh sorry I thought it was one of those occassionaly thing happens things. Not OMG hair on fire.17:16
markus_zedleafe: http://docs.openstack.org/developer/nova/sample_config.html17:16
*** korzen has joined #openstack-nova17:16
sdaguedansmith: yes, that would be reasonable. Though it doesn't destroy your machine any more than the rest of devstack.17:17
*** pratikmallya has quit IRC17:17
sdagueand it's only set that way for affected platforms of the bug17:17
*** korzen_ has quit IRC17:17
dansmithsdague: we move things in /sbin elsewhere in devstack/17:17
edleafemarkus_z: I guess I'm not asking clearly17:17
markus_zedleafe: The idea was to have one python module per "nova.conf" section. Sections are the "[...]" things in the config file.17:17
cburgess_sdague So the question is.. do we need a timeout? I ask because while its not hard to set a timer in python to do this, due to how we invoke things via the oslo libraries and rootwrap vs the new fancy daemon thing its actually a bit of work to get all the bits to properly create a timeout.17:18
sdaguedansmith: we modify so much of /etc17:18
edleafemarkus_z: so nearly everything is in [DEFAULT]17:18
cburgess_dansmith Yeah sdague has a point. devstack pretty much stomp /etc17:18
dansmithsdague: yeah, but dicking around in /sbin is a different thing I think17:18
sdaguecburgess_: because you are concerned that ebtables deadlocks?17:18
*** shaohe_feng has joined #openstack-nova17:19
cburgess_sdague I have no evidence of that. Just thinking/coding defensibly.17:19
markus_zedleafe: Yes, I didn't know that before.17:19
sdaguecburgess_: right, that's a good question17:19
sdaguecburgess_: let me look for a sec17:19
cburgess_sdague We can try it without a timeout. If it proves to be an issue we can fix it later.17:20
edleafemarkus_z: for reviewing simplicity, I've been thinking about 1:1 files in nova/conf/ to the location of the original CONF declaration17:20
edleafemarkus_z: IOW, if I move an option from nova/scheduler/manager.py, it will go in nova/conf/scheduler/manager.py17:20
*** pradk has quit IRC17:21
*** e0ne has joined #openstack-nova17:21
johnthetubaguyedleafe: can't we just group all the scheduler options in one file?17:22
markus_zedleafe: I think I get you.17:22
cburgess_sdague There are a few ways to do a timeout, but all of them would require us to muck with oslo_concurrency17:22
markus_zedleafe: what johnthetubaguy said17:22
edleafejohnthetubaguy: that was the other option I was asking about17:22
sdaguecburgess_: yeh, honestly, for all the kwargs that oslo_concurrency takes, the fact that it doesn't support command timeout is a thing17:22
dimscburgess_ : help us fix things in oslo_concurrency? :)17:22
*** njirap has quit IRC17:23
sdaguethat being said, is this going to be in an interuptable state?17:23
johnthetubaguyedleafe: possibly keep the filters and weights ones in their own files, thinking about it17:23
mdboothUpdating a spec here. Remind me, can I add random sections, or will a bot complain at me?17:23
edleafemarkus_z: johnthetubaguy: so everything under nova/scheduler/ will be in nova/conf/scheduler.py, and everything under nova/foo/ will be in nova/conf/foo.py17:23
cburgess_sdague dims Right so... we can fix oslo but it takes even more time to land the fix. We could maybe land a fix no timeout, then add timeout to oslo, then land a time fix with timeout.17:23
sdaguecburgess_: yeh, lets do that path17:23
cburgess_sdague Yeah as I recall its just sitting attempting to lock a file.17:23
sdagueland nova without the timeout17:23
sdaguethat makes things better17:24
openstackgerritTom Patzig proposed openstack/nova: Add 'insecure' option to ironic sect in nova.conf  https://review.openstack.org/23256017:24
johnthetubaguyedleafe: not sure we need to be that clear cut, see what the largest natural groupings turn out to be17:24
cburgess_sdague I think I wrote a test along time ago around this and determined that it was possible to kill it.17:24
cburgess_sdague OK17:24
sdaguecburgess_: might be worth seeing if libvirt is keeping an eye on a timeout here as well, danpb, any ideas there?17:24
sdaguebecause if we are as good as libvirt, we're probably fine17:24
cburgess_sdague So the gate has a new enough ebtables and libvirt right? We just aren't going to do any version testing and just assume we have the right versions moving forwad right?17:24
johnthetubaguyedleafe: like, common stuff, lots of optional stuff, might be common, like scheduler.py and scheduler_filter_weights.py maybe? not sure17:25
sdagueit has new enough ebtables17:25
markus_zedleafe: johnthetubaguy: The grouping most people agreed on was using the sections ("[libvirt]", "[default]", ...) for the grouping.17:25
cburgess_sdague I don't think there is a timeout in libvirt, its been a while since looked at the code though.17:25
cburgess_sdague OK17:25
*** tdurakov has joined #openstack-nova17:25
*** unicell has quit IRC17:25
cburgess_This should be fairly easy.17:25
sdagueit does *not* have new enough libvirt on ubunut17:25
*** rushiagr is now known as rushiagr_away17:25
*** matrohon has quit IRC17:25
edleafejohnthetubaguy: that's a very subjective grouping, no?17:25
johnthetubaguymarkus_z: yeah, that makes sense, but I don't want all of the default ones in one file, I suspect17:25
sdaguethe fedora jobs (which are non voting) do, so you could probably see results there17:25
cburgess_sdague Right so thats the ugly work around thing you put into devstack. I'll merge a similiar thing to nova in a latter patch for the ugly workaround.17:25
edleafemarkus_z: we're gonna have one ginormous default.py file!17:25
johnthetubaguyedleafe: yes, we want a logical grouping17:26
sdaguecburgess_: right, and I'll make devstack use the nova ugly work around17:26
cburgess_sdague Right17:26
sdagueI just wanted to try to deal with the bleeding here17:26
danpbcburgess_: there's no timeout in libvirt for ebtables and i'm not really seing a reason to add one either17:26
*** gjayavel_ has joined #openstack-nova17:26
johnthetubaguymarkus_z: edleafe: this is a horrible answer, but I think it will be easier to tell whats correct when we see the true horror of the specific change.17:26
cburgess_danpb Didn't think there was.17:26
markus_zedleafe: johnthetubaguy: Right, we wanted also figure out what the natural grouping is.17:27
edleafejohnthetubaguy: markus_z: ok, well I've started putting all into one file. When that's ready I'll push it and we can argue on gerrit, ok?17:27
danpbIME adding timeouts on this kind of thing has only ever led to use *causing* failures where something would have in fact succeeeded with a longer timeout17:27
cburgess_sdague So.. land nova no timeout, unrelated we can go add a timeout to oslo but.. without a timeout we are as good as libvirt.17:27
johnthetubaguyedleafe: makes sense17:27
openstackgerritAndrew Laski proposed openstack/nova: Add persistence to the RequestSpec object  https://review.openstack.org/21175317:27
sdaguecburgess_: yeh17:27
cburgess_danpb I'm totally fine with that.17:27
markus_zedleafe: I need a second to think about it17:27
johnthetubaguyedleafe: maybe just do a few models at first, to see how it looks?17:27
danpbany ebtables command should complete eventually unless the kernel is so fubar that you are doomed regardless17:27
cburgess_danpb As in, if we are as a good as libvirt thats good enough for nova.17:27
cburgess_sdague ^^17:27
danpb(ie only time i've seen ebtables hang is when the kernel has already reported  mild panics and tried to carry on)17:28
sdagueI think the only fail scenario is if ebtables gets killed before it's lock file is cleaned out17:28
sdaguethen starts stacking up on stale lock file17:28
johnthetubaguymarkus_z: sorry, I forgot to submit my questions earlier, done that now: https://review.openstack.org/#/c/244177/417:28
sdagueI don't know how well it recovers from said condition17:28
danpbif correctly implemented there should be no chance of stale lock files17:28
cburgess_sdague Yeah so I don't think thats possible the way it does locking. I will look at the code to confirm but I think we are good.17:28
johnthetubaguyedleafe: its tempting to get one merged before we start on uploading new ones17:29
*** rk4n has joined #openstack-nova17:29
danpbie, if it holds a fcntl lock, those get released when the process exits17:29
markus_zedleafe: yeah, ok, let's discuss in code. Makes probably more sense.17:29
openstackgerritAlexis Lee proposed openstack/nova: Move Process and Mentoring pages to devref  https://review.openstack.org/20607017:29
cburgess_Pretty sure when I looked last year it was using fcntl17:29
danpbit certainly shouldn't rely on mere existance of the file17:29
johnthetubaguymarkus_z: I am wondering about explicitly importing the used options, to help grep make it clear where things are coming from? or is that overkil?17:29
markus_zjohnthetubaguy: thanks, I'll take a look17:29
edleafejohnthetubaguy: sure, mine is dependent on markus_z's initial patch17:29
johnthetubaguyedleafe: makes sense17:29
*** gjayavelu has quit IRC17:29
cburgess_sdague Just as soon as I can remember how to use nova-network in devstack I'll get this patch up. Been spending way to much time with neutron.17:30
*** salv-orl_ has joined #openstack-nova17:32
*** romainh has left #openstack-nova17:32
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Volume support for multi hypervisors  https://review.openstack.org/24598617:32
klindgrendansmith, select * as inst_row_count from instance_system_metadata im join instances i   on im.instance_uuid = i.id where im.deleted_at is null    and im.key LIKE 'instance_type_%'   and i.vm_state = 'deleted';17:32
sdaguecburgess_: heh17:32
klindgrenresult == 2200 rows17:32
cburgess_sdague You can't wash that out.17:32
openstackgerritEd Leafe proposed openstack/nova: DO NOT MERGE - WIP  https://review.openstack.org/24589117:33
edleafejohnthetubaguy: markus_z: ^^17:33
dansmithklindgren: you need d.deleted != null not vm_state to be sure17:33
*** ssurana has joined #openstack-nova17:33
dansmithklindgren: sorry, i.deleted17:33
*** jdurgin has quit IRC17:33
*** mgoddard_ has joined #openstack-nova17:33
edleafejohnthetubaguy: markus_z: I posted what I have so far17:33
*** scheuran has quit IRC17:34
*** stackdump has joined #openstack-nova17:34
*** ctrath has joined #openstack-nova17:34
*** rcernin has joined #openstack-nova17:36
klindgrendansmith, select * from instance_system_metadata im join instances i   on im.instance_uuid = i.id where im.deleted_at is null    and im.key LIKE 'instance_type_%'   and i.deleted != null;17:37
klindgrenEmpty set (0.20 sec)17:37
dansmithklindgren: okay cool17:37
dansmithklindgren: I'm not positive about that im.deleted bit, tbh17:38
dansmithklindgren: but since you were getting results before and now not, I think that's probably fine17:38
dansmithklindgren: so are you interested in trying something hacky?17:38
dansmithklindgren: if I could get devstack on kilo to work, I'd be testing it, but alas17:38
markus_zjohnthetubaguy: I answered your questions in https://review.openstack.org/#/c/244177/317:39
mriedem_meetingnot deleted is 0 i think17:39
*** mriedem_meeting is now known as mriedem17:39
mriedemi.deleted != 017:39
klindgrenI am willing to do a moderate amount of hackyness, balanced by the fact this is a production environment17:39
sdaguecburgess_ / danpb : http://git.netfilter.org/ebtables/tree/libebtc.c#n187 which I think means we're fine without a timeout17:39
*** ljxiash has quit IRC17:40
klindgreni.deleted != 0 = 220017:40
dansmithklindgren: deleted!=0 means deleted17:40
dansmithso those you don't care about17:40
*** pratikmallya has joined #openstack-nova17:42
cburgess_sdague danpb Well they call open with O_EXCL and only mask SIGINT. SIGKILL won't unlock the file. But.. I think the kernel will still clean the FD up on kill.17:43
*** sacharya_ has joined #openstack-nova17:43
*** alexschm has quit IRC17:43
*** Piet has quit IRC17:44
*** jmccrory has quit IRC17:44
*** electrocucaracha has joined #openstack-nova17:45
danpbcburgess_: sigh, yes, it is stupidly implemented - they really ought to use fcntl locks as they're the only guaranteed race-free auto-cleaned impl17:45
*** electrocucaracha has quit IRC17:45
*** electrocucaracha has joined #openstack-nova17:45
*** sacharya has quit IRC17:45
cburgess_danpb But I think the FD should get cleaned up anyways right? Its a dirty implementation but I *think* its safe.17:45
*** e0ne has quit IRC17:46
*** korzen has quit IRC17:47
*** apoorvad has joined #openstack-nova17:47
*** salv-orl_ has quit IRC17:47
*** browne has quit IRC17:47
*** ctrath has quit IRC17:48
danpbcburgess_: nothing ensures the lockfile is deleted in SIGKILL, only in clean exit17:48
markus_zedleafe: I get it now. code helps me a lot :) I'll make comments17:48
*** ZZelle_ has joined #openstack-nova17:48
*** salv-orlando has joined #openstack-nova17:48
danpbof course if admins use sigkill they get what they deserve to some extent17:48
*** garyk has joined #openstack-nova17:48
*** e0ne has joined #openstack-nova17:49
*** greghaynes has quit IRC17:49
dansmithklindgren: you don't have any preprod to run this in?17:49
*** auggy has quit IRC17:49
dansmithklindgren: to measure potential performance improvement?17:49
*** yonglihe has quit IRC17:50
cburgess_danpb /sigh yeah you are right. I missed that.17:50
*** junjie has quit IRC17:50
*** betherly has quit IRC17:51
klindgrendansmith, we do - but its no where near what we have in prod17:51
*** READ10 has quit IRC17:51
klindgrenand I cant drive the same load in it - because I dont run all the same crap that the users do17:52
dansmithklindgren: well, obviously, but surely you can hamper your conductor, hammer your preprod, then apply this and see how it goes?17:52
klindgrenIE 95% sure load is coming from metadata17:52
klindgrenyea - like I said I am willing to do some hacky stuffs17:52
*** tpeoples has quit IRC17:53
*** kgalanov has quit IRC17:53
*** rpodolyaka1 has quit IRC17:55
*** haomaiwang has quit IRC17:55
mhorban_Hi sdague, johnthetubaguy, Could you please look at https://review.openstack.org/#/c/237588/ ?17:55
*** yonglihe has joined #openstack-nova17:56
*** _RuiChen_ has joined #openstack-nova17:56
*** artom_ has joined #openstack-nova17:56
*** thangp has quit IRC17:57
dansmithklindgren: https://review.openstack.org/24653017:57
*** markvoelker_ has joined #openstack-nova17:57
*** burgerk_ has joined #openstack-nova17:57
*** ctrath has joined #openstack-nova17:58
*** lpetrut1 has joined #openstack-nova17:58
*** lpetrut has quit IRC17:58
*** lpetrut1 is now known as lpetrut17:58
dansmithalaski: ^^ in case you want to try it too17:58
*** jmccrory has joined #openstack-nova17:58
*** pkoniszewski_ has quit IRC17:59
*** baffle has quit IRC17:59
cburgess_dansmith LOL I love that option name and commit message17:59
*** rlrossit has quit IRC17:59
*** bnemec has quit IRC17:59
*** mdrabe has quit IRC17:59
*** keekz has quit IRC17:59
*** agireud has quit IRC17:59
dansmithcburgess_: I was going to go for more obscenity, but decided against it17:59
*** _gryf has quit IRC17:59
*** ZZelle has quit IRC17:59
*** artom has quit IRC18:00
*** markvoelker has quit IRC18:00
*** sbadia has quit IRC18:00
*** _RuiChen has quit IRC18:00
*** yonglihe has quit IRC18:00
cburgess_dansmith Yeah its public so have to maintain a certain level of decorum.18:00
*** bnemec has joined #openstack-nova18:00
*** burgerk has quit IRC18:00
*** xek has quit IRC18:00
*** maurosr has quit IRC18:00
*** baffle has joined #openstack-nova18:00
*** whenry has joined #openstack-nova18:01
*** lpetrut has quit IRC18:02
*** pratikma_ has joined #openstack-nova18:02
*** thangp has joined #openstack-nova18:04
*** unicell has joined #openstack-nova18:05
*** sbadia has joined #openstack-nova18:05
*** keekz has joined #openstack-nova18:05
dansmithalaski: klindgren: had to fix a typo, updated now.18:05
*** pratikmallya has quit IRC18:05
*** ihrachys has joined #openstack-nova18:05
*** agireud has joined #openstack-nova18:05
dansmithklindgren: if you hammer that real good, I'd be willing to at least propose it seriously for kilo as an opt-in band-aid18:06
*** maurosr has joined #openstack-nova18:06
*** ctrath has quit IRC18:06
*** armax has quit IRC18:06
*** pkoniszewski has joined #openstack-nova18:07
*** ZZelle has joined #openstack-nova18:07
alaskidansmith: we're past that point now so I don't have a good place to test anymore18:08
dansmithalaski: oh, even better :)18:09
dansmithalaski: and congrats :)18:09
alaskiyep, and thanks :)18:09
*** ctrath has joined #openstack-nova18:10
*** _gryf has joined #openstack-nova18:11
*** sacharya has joined #openstack-nova18:11
*** mdrabe has joined #openstack-nova18:11
*** ihrachys has quit IRC18:11
cburgess_sdague danpb So another question, we currently use utils.sync on ebtables call. If we are using --concurrent should we remove that?18:12
*** sacharya_ has quit IRC18:14
*** ihrachys has joined #openstack-nova18:14
*** neiljerram_bb has quit IRC18:14
*** tdurakov has quit IRC18:16
mriedem246013 is at the top of the gate queue18:16
*** vilobhmm has joined #openstack-nova18:17
danpbcburgess_: yeah, iirc there's a comment in code saying it should be removed when using --concurrent18:17
mriedemdansmith: bluebox was just moving to kilo i think, so jlk might be interested18:17
*** otter768 has joined #openstack-nova18:17
* mriedem jumps to -operators18:17
*** rdopiera has quit IRC18:18
*** tdurakov has joined #openstack-nova18:18
dansmithmriedem: merged now18:18
*** tonytan4ever has quit IRC18:19
*** achanda has quit IRC18:21
*** lucasagomes is now known as lucas-dinner18:21
*** otter768 has quit IRC18:22
*** ihrachys has quit IRC18:25
*** fawadkhaliq has quit IRC18:26
*** armax has joined #openstack-nova18:26
sdaguefinally that nova objects fix merged18:27
*** liverpooler has joined #openstack-nova18:27
*** Vek_ has quit IRC18:28
*** baoli has quit IRC18:29
*** chhavi has quit IRC18:29
*** ihrachys has joined #openstack-nova18:30
*** penick has joined #openstack-nova18:31
*** baoli has joined #openstack-nova18:32
*** junjie has joined #openstack-nova18:34
*** kfarr has joined #openstack-nova18:35
vilobhmmmelwitt : ping18:36
melwittbauzas: what I meant on the test_objects.py backport was, it won't merge until we do something about the cells tests, right? it's not passing py27 and py34 unit tests because of what I commented about18:36
vilobhmmmelwitt : why did this didn't merge https://review.openstack.org/#/c/241016/ any idea ?18:36
*** tyagiprince has joined #openstack-nova18:37
*** PaulMurray is now known as ptm_away18:37
markus_zsdague: That was the fix which blocked the gate, right?18:37
*** dims has quit IRC18:37
sdaguemarkus_z: yes, for nova18:37
markus_zsdague: Ah, cool, good to head.18:37
*** gjayavel_ has quit IRC18:38
markus_zsdague: Can I see the start of the issue at http://status.openstack.org/openstack-health/#/project/openstack%252Fnova ?18:38
melwittvilobhmm: there was a problem blocking the gate yesterday, you can recheck it now18:38
*** jichen has joined #openstack-nova18:38
sdaguemarkus_z: maybe, it was only the nova project affected18:39
markus_zsdague: I'm asking because I still fail to detect gate failures early.18:39
sdagueit wasn't a cross project thing that was breaking everyone18:39
sdaguealso, that's only tempest jobs18:39
*** lpetrut has joined #openstack-nova18:39
sdagueit doesn't handle unit tests18:39
*** danpb has quit IRC18:40
*** gokrokve has quit IRC18:40
*** browne has joined #openstack-nova18:40
markus_zoh, you're right18:40
markus_zI thought that could be the single source of truth if something is burning18:40
*** dims has joined #openstack-nova18:40
*** junjie has quit IRC18:40
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Virt image properties boot time override via flavor  https://review.openstack.org/24654818:40
*** tonytan4ever has joined #openstack-nova18:41
*** mriedem has quit IRC18:42
openstackgerritMaxime Leroy proposed openstack/nova: Libvirt: Support fp plug in vhostuser vif  https://review.openstack.org/24536918:42
openstackgerritMaxime Leroy proposed openstack/nova: Libvirt: Support ovs fp plug in vhostuser vif  https://review.openstack.org/23768018:42
samueldmqis nova microversioned already ?18:42
*** ihrachys has quit IRC18:42
*** e0ne has quit IRC18:43
*** ihrachys has joined #openstack-nova18:43
*** suro-patz has joined #openstack-nova18:43
*** jaypipes has quit IRC18:43
*** unicell1 has joined #openstack-nova18:44
*** daemontool has quit IRC18:44
*** mriedem has joined #openstack-nova18:44
*** pratikma_ has quit IRC18:45
cburgess_sdague danpb so given that ebtables isn't smart about its locking.. are we still ok with no timeout? Any admin that issues a SIGKILL against ebtables gets what they deserve?18:46
*** rlrossit has joined #openstack-nova18:46
*** andreykurilin__ has quit IRC18:46
*** unicell has quit IRC18:46
*** nic has joined #openstack-nova18:47
*** w_verdugo has quit IRC18:47
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Volume support for multi hypervisors  https://review.openstack.org/24598618:48
*** sacharya_ has joined #openstack-nova18:48
*** boris-42 has quit IRC18:48
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo VersionPredicateField  https://review.openstack.org/24532718:49
openstackgerritSergey Nikitin proposed openstack/nova: Removed extra spaces from double line strings  https://review.openstack.org/24298918:49
*** EinstCrazy has joined #openstack-nova18:49
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo FlexibleBooleanField  https://review.openstack.org/24533618:49
*** ociuhandu has quit IRC18:49
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo DictOfListOfStringsField  https://review.openstack.org/24534418:50
*** sacharya has quit IRC18:51
openstackgerritRyan Rossiter proposed openstack/nova: Use ContainerFormat instead of strings  https://review.openstack.org/24583518:51
openstackgerritRyan Rossiter proposed openstack/nova: Add ContainerFormat field  https://review.openstack.org/24583418:51
*** tpeoples has joined #openstack-nova18:51
*** betherly has joined #openstack-nova18:52
*** lpetrut has quit IRC18:52
*** armax has quit IRC18:53
*** lpetrut has joined #openstack-nova18:53
edleafesamueldmq: yes: http://docs.openstack.org/developer/nova/api_microversions.html18:54
*** armax has joined #openstack-nova18:55
samueldmqedleafe: nice, I was wondering how useful that could be for keystone18:55
samueldmqedleafe: I will take a look at it, thanks for the link18:55
*** EinstCrazy has quit IRC18:56
*** gokrokve has joined #openstack-nova18:57
*** pratikmallya has joined #openstack-nova18:57
*** sc68cal_ is now known as sc68cal19:00
*** irenab has quit IRC19:00
*** gokrokve has quit IRC19:02
*** rotbart has quit IRC19:03
*** rushil has quit IRC19:04
*** ociuhandu has joined #openstack-nova19:05
*** penick has quit IRC19:06
edleafesamueldmq: np19:07
*** ihrachys has quit IRC19:07
*** andrewbogott has quit IRC19:09
*** andrewbogott has joined #openstack-nova19:09
*** sacharya_ has quit IRC19:11
markus_ztonyb: kashyap: bauzas: FYI: http://lists.openstack.org/pipermail/openstack-dev/2015-November/079679.html19:11
bauzasmelwitt: heya sorry was afk19:12
mriedemmarkus_z: fwiw this has come up before, and russellb had ideas on the same kind of thing19:12
*** Marga_ has joined #openstack-nova19:13
russellbi think i even built a job at one point?19:13
*** sacharya has joined #openstack-nova19:13
russellbor maybe i'm crazy19:13
mriedemcould be the neutron talking19:13
russellbvery well could be.19:13
melwittbauzas: np19:14
russellbmarkus_z: i believe i was messing with having a custom nodepool image like you talk about19:14
russellbbut i didn't want us to have to maintain packages19:14
russellbFedora has a repo with bleeding edge virt components that i used19:14
russellband then a similar thing was being developed for centos that i was going to switch to19:15
russellband that's roughly where my memory gets fuzzy and i don't remember what happened to it all19:15
*** gokrokve has joined #openstack-nova19:15
russellbbut i'm not working on it :)19:15
*** pixelbeat has quit IRC19:15
markus_zrussellb: When did that happen?19:15
russellbcouple years ago19:16
*** pushkaru has joined #openstack-nova19:16
*** pumaranikar has quit IRC19:16
*** ihrachys has joined #openstack-nova19:16
*** rushil has joined #openstack-nova19:18
markus_zrussellb: Hm, wanna do it again? ;P19:18
*** kgalanov has joined #openstack-nova19:18
*** jdurgin1 has quit IRC19:18
mriedemsdague: would you agree with the granularity comment here https://review.openstack.org/#/c/240401/ ?19:18
mriedemi.e. lumping too many changes into a single microversion19:18
russellbmarkus_z: virt-preview for fedora is still around ... there was something being built for CentOS but i can't find status this minute19:19
*** paul-carlton has left #openstack-nova19:19
*** tdurakov has quit IRC19:19
russellbmarkus_z: my only advice is to try to piggy back on one of the efforts for maintaining packages like that, seems like less work19:19
*** gokrokve has quit IRC19:20
markus_zrussellb: yeah, it makes sense to re-use the efforts from others. Thanks for your pointers.19:20
*** ildikov has joined #openstack-nova19:20
markus_zI have to go home now, see you folks tomorrow19:20
russellbnp, good luck19:20
*** markus_z has quit IRC19:22
*** signed8__ has joined #openstack-nova19:23
*** ekarlso has joined #openstack-nova19:24
*** signed8bit has quit IRC19:25
*** Piet has joined #openstack-nova19:26
*** ctrath has quit IRC19:26
*** jaypipes has joined #openstack-nova19:27
*** e0ne has joined #openstack-nova19:27
*** ctrath has joined #openstack-nova19:29
*** RichardRaseley has joined #openstack-nova19:30
*** tpeoples_ has joined #openstack-nova19:30
bauzasmelwitt: sorry, again afk, okay I see your point, my bad19:31
*** tyagiprince has quit IRC19:33
openstackgerritBalazs Gibizer proposed openstack/nova: Add service status notification  https://review.openstack.org/24567819:33
*** josecastroleon has joined #openstack-nova19:34
*** electrocucaracha has quit IRC19:34
openstackgerritMitsuhiro Tanino proposed openstack/nova-specs: Virt image properties boot time override via flavor  https://review.openstack.org/24654819:35
*** greghaynes has joined #openstack-nova19:36
*** junjie has joined #openstack-nova19:36
*** ihrachys has quit IRC19:36
*** yonglihe has joined #openstack-nova19:36
openstackgerritShaoHe Feng proposed openstack/nova: HyperV: do not log twice with different level.  https://review.openstack.org/24626919:37
*** josecastroleon has quit IRC19:38
*** klkumar has quit IRC19:38
*** auggy has joined #openstack-nova19:41
*** guilh has joined #openstack-nova19:41
*** sneti_ has quit IRC19:42
*** w_verdugo has joined #openstack-nova19:42
*** sneti has joined #openstack-nova19:43
guilhHello all,19:43
guilhI have create an aggregate and flavor and I also create a new cinder volume-type called 'fast-iscsi'. I can create new volume from that volume-type without problem, I can also attach that volume to any VM, and it work.19:43
guilhI still have a lot of trouble to just set the default volume-type to my flavor called Client.fast-iSCSI.19:43
guilhDoes anyone know how to dedicated a volume-type to a flavor?19:43
guilhGuillaume HARVEY19:43
*** aginwala has joined #openstack-nova19:43
*** ihrachys has joined #openstack-nova19:43
guilhHello all, I have create an aggregate and flavor and I also create a new cinder volume-type called 'fast-iscsi'. I can create new volume from that volume-type without problem, I can also attach that volume to any VM, and it work. I still have a lot of trouble to just set the default volume-type to my flavor called Client.fast-iSCSI. Does anyone know how to dedicated a volume-type to a flavor? Regards, Guillaume HARVEY19:45
*** jamielennox|away is now known as jamielennox19:45
RichardRaseleyguilh: This is a development channel. Please use #openstack for support issues.19:46
guilhoh thx19:46
*** penick has joined #openstack-nova19:46
*** exploreshaifali has joined #openstack-nova19:47
*** angdraug has joined #openstack-nova19:48
*** salv-orlando has quit IRC19:48
*** ccarmack has quit IRC19:49
*** ccarmack has joined #openstack-nova19:49
*** e0ne has quit IRC19:50
*** eglynn has quit IRC19:51
*** ihrachys has quit IRC19:51
*** jamielennox is now known as jamielennox|away19:52
*** baoli has quit IRC19:53
*** stackdump has quit IRC19:54
*** sacharya has quit IRC19:55
*** doug-fis_ has joined #openstack-nova19:57
*** armax has quit IRC19:57
*** ed___ has joined #openstack-nova19:58
*** ed___ is now known as edleafe_19:59
*** eliqiao has quit IRC19:59
gibi_nova notification subteam meeting starting now on openstack-meeting-alt20:00
openstackgerritChet Burgess proposed openstack/nova: Use --concurrent with ebtables  https://review.openstack.org/24658020:00
openstackgerritChet Burgess proposed openstack/nova: ebtables/libvirt workaround  https://review.openstack.org/24658120:00
*** aginwala has quit IRC20:00
*** doug-fish has quit IRC20:00
cburgess_sdague Reviews for --concurrent posted20:01
*** doug-fis_ has quit IRC20:01
*** eliqiao has joined #openstack-nova20:01
cburgess_Heading to lunch back in a bit20:01
sdaguecburgess_: thanks!20:01
*** doug-fish has joined #openstack-nova20:01
*** ccesario has quit IRC20:01
sdagueI will review later today, TC meeting now20:01
cburgess_sdague First review is proper fix, 2nd one is using your workaround. I need to ammend the commit to indicate you wrote that. my bad.20:01
sdaguecburgess_: meh, I don't care that much20:01
sdagueunless you are just trying to spread the blame :)20:02
*** artom_ has quit IRC20:02
cburgess_sdague BINGO!20:02
openstackgerritChet Burgess proposed openstack/nova: ebtables/libvirt workaround  https://review.openstack.org/24658120:02
cburgess_sdague Plus I need a reason why I'm commiting something with the HP copyright.20:03
*** baoli has joined #openstack-nova20:03
*** pratikmallya has quit IRC20:05
openstackgerritMark McLoughlin proposed openstack/nova: servicegroup: remove the zookeeper driver  https://review.openstack.org/24634320:06
*** claudiub has joined #openstack-nova20:07
sdaguecburgess_: I think there is an issue with the first patch, comment left20:07
cburgess_sdague reading20:08
cburgess_sdague Good catch20:08
*** jamielennox|away is now known as jamielennox20:09
openstackgerritChet Burgess proposed openstack/nova: Use --concurrent with ebtables  https://review.openstack.org/24658020:10
cburgess_sdague fixed20:10
*** aginwala has joined #openstack-nova20:10
*** tdurakov has joined #openstack-nova20:10
openstackgerritChet Burgess proposed openstack/nova: ebtables/libvirt workaround  https://review.openstack.org/24658120:11
sdagueoh, I'm going to hit you for one more administrative issue, we should probably add a release note that nova requires ebtables >= 2... (whatever it is)20:11
*** Marga_ has quit IRC20:11
sdaguethat should be part of that patch, then I'm +220:11
cburgess_sdague OK thats cool. I'll fix that after lunch.20:11
sdaguecburgess_: great, thanks again20:11
*** doug-fish has quit IRC20:12
cburgess_sdague So just to be clear, I update the commit message to indicate we now require ebtables blah blah blah and then I follow this release note process you linked me?20:12
*** ijw has joined #openstack-nova20:12
cburgess_Or this process is part of the same commit?20:12
*** doug-fish has joined #openstack-nova20:12
sdagueit's part of the same commit20:12
sdagueit's a tool that builds release notes into the tree20:12
cburgess_sdague OK I'll give it a shot. Thanks.20:12
*** penick has quit IRC20:14
*** sacharya has joined #openstack-nova20:16
*** Marga_ has joined #openstack-nova20:17
*** otter768 has joined #openstack-nova20:18
bauzassdague: I'm tempted to respin the stable reno change with a different Change-Id so we could merge the CI job patch20:20
bauzassdague: because for the moment, stable/liberty is wedged20:20
bauzasdue to an ovo issue that can't be fixed easily as melwitt spotted in https://review.openstack.org/#/c/246022/120:21
bauzassdague: thoughts on that?20:21
sdagueoh, the ovo thing wreaked havoc there as well?20:21
dansmithbauzas: it's not an ovo issue20:21
dansmithbauzas: the ovo fix is blocked on a cells issue20:21
bauzasdansmith: sorry I was unclear20:21
dansmithsdague: just a backport of the other fix20:22
bauzasdansmith: the backport fix for the ovo release is -1'd by Jenkins because of a cells job issue, right (to be clear)20:22
*** pixelbeat has joined #openstack-nova20:23
*** otter768 has quit IRC20:23
sdaguebauzas: on cells unit tests actually20:23
melwittI'm not sure it's cells-specific in that somehow with the backport it makes a remote call of objects.InstanceV1.get_by_uuid() return an instance 2.020:23
bauzascorrect, not a cells job, dammit me20:23
dansmithmelwitt: I thought I tried it on base stable and it failed there too20:24
melwittdansmith: by base stable do you mean without the backport?20:25
sdaguehttp://logs.openstack.org/22/246022/1/check/gate-nova-python27/45d1d61/console.html#_2015-11-16_23_20_19_491 hmmmm it says it's returning a V1 object there, though what was the thing it was returning before20:25
sdaguewhich actually has a real repr20:25
*** rook has quit IRC20:25
melwittdansmith: without the backport on base stable with o.vo 0.13.0 I was getting the recursion depth exceeded20:25
dansmithmelwitt: yeah, I forced the package20:25
bauzasmelwitt: dansmith: that was my original thought, I don't get how the change can be related to the problem - even if it's a bit close20:26
dansmithbut let me re-do that and see if I did what I thought I did20:26
*** aginwala has quit IRC20:26
bauzasI can play too20:27
dansmithhmm, nope,20:27
dansmithit works without the backport20:27
dansmithI must've not done the right thing when I was rushing last night20:27
*** doug-fish has quit IRC20:27
*** doug-fish has joined #openstack-nova20:28
dansmithso, let me work on that20:28
*** sacharya_ has joined #openstack-nova20:28
melwittdansmith: yeah, sorry I don't understand. you mean it works without the backport with o.vo 0.13.0 by some modification you did?20:29
ansiwenmriedem: I addressed your comments, if you wanna have a look: https://review.openstack.org/#/c/243616/20:29
dansmithI mean it must be related to the backport20:29
bauzasdansmith: okay, so I'm holding off my possible respin of the reno stable backport with a different change-id20:29
dansmithit's weird that it's just cells tests tho20:30
dansmithwait a sec20:30
*** sacharya has quit IRC20:31
*** apoorvad has quit IRC20:31
melwittsdague: that's the expected call, wanting InstanceV1 which it used to get before the backport20:32
sdagueI thought + meant it was what it got and - is what it expected?20:32
sdagueor do I have that totally backwards20:32
sdaguegod I had mox20:32
*** rfolco has quit IRC20:33
melwittit says unexpected:-  expected:+ but I guess that depends on what order you put them which is why people have done patches to change order of assertEquals for example?20:33
melwittI put a bunch of prints in the test and verified it gets Instance 2.0 after the get_by_uuid call where objects.Instance has been mocked as objects.InstanceV120:34
*** edtubill has joined #openstack-nova20:35
melwittI just don't understand how that's happening. don't know how the version makes its way from the class to the remote side, and then sent back20:35
*** baoli has quit IRC20:36
dansmithI expect it's just a race to setup the test mocks20:37
dansmithI'm looking at it20:37
*** apoorvad has joined #openstack-nova20:37
*** pratikmallya has joined #openstack-nova20:37
*** Vek_ has joined #openstack-nova20:38
*** greghaynes has quit IRC20:38
*** greghaynes_ is now known as greghaynes20:38
*** jichen has quit IRC20:38
*** pratikmallya has quit IRC20:39
*** pratikmallya has joined #openstack-nova20:39
dansmithlemme run this a couple more times20:41
*** doug-fish has quit IRC20:41
dansmiththis cells save test is kinda hariy20:41
dansmithalso hairy20:41
*** doug-fish has joined #openstack-nova20:42
dansmithupdated, let's see if that looks better20:43
cburgess_sdague Which section should this be in? prelude?20:44
*** e0ne has joined #openstack-nova20:44
bauzasdansmith: IIUC, that call is not first checked https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L330-L33120:45
bauzasdansmith: because of the mock call you make in the IndirectionAPIFixture20:45
bauzasdansmith: when calling https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L32920:45
sdaguecburgess_: I'd actually put it in the upgrade section I think20:46
cburgess_sdague Cool20:46
sdaguebecause it's a new prereq before you deploy this code20:46
bauzasand that's also the same problem with https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L324-L32720:46
*** doug-fish has quit IRC20:46
wznoinskdims: are you around?20:46
sdagueand I'll leave it as an exercise for you about figuring out what the actual min required version is here20:46
bauzasdansmith: but I could be wrong20:46
dimswznoinsk pong20:46
*** e0ne has quit IRC20:47
wznoinskdims: hi, a (possibly) quick question for you20:47
dimswznoinsk sure20:47
dansmithbauzas: what call?20:47
wznoinsknova-docker, was it not getting any traction or had some fundamental limitation and didn't evolve?20:48
sdaguewznoinsk: it was broken all the time, no one cared enough to keep it working for more than a day20:48
wznoinskI heard about lack of interest for sure20:48
sdagueor to address some of the limitations20:49
dimswznoinsk : some people are using it. i keep it alive when i get a chance especially around release boundaries.20:49
wznoinsksdague: sure, but was lack of interest cause by some major (design) issues ?20:49
bauzasdansmith: for example wiht https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L36220:49
bauzasdansmith: it goes to https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L329-L33120:49
dimswznoinsk : if you see launchpad for nova-docker you can see the tempest tests yet to be fixed and some other work needed20:49
bauzasdansmith: then you have http://logs.openstack.org/22/246022/1/check/gate-nova-python27/45d1d61/console.html#_2015-11-16_23_20_19_30720:50
dansmithbauzas: ... I'm not sure what your point is20:50
*** tonytan4ever has quit IRC20:50
wznoinsksdague, dims: I'm trying to understand what was the reason it got dropped... without spending couple of days trying to figure out there were some major flows in how it was designed/implemented to work20:50
*** rmart04 has joined #openstack-nova20:50
*** salv-orlando has joined #openstack-nova20:51
dimswznoinsk : folks switched to magnum as it uses docker swarm, kubernetes etc20:51
dansmithwznoinsk: it's not a good fit inside nova20:51
bauzasdansmith: sorry... if you see the above log, that's because https://github.com/openstack/nova/blob/master/nova/tests/unit/objects/test_instance.py#L332-L333 is not called IIUC20:51
wznoinskif it's lack of use cases I hope I may have one but if it's fundamental flow of some sort that made people saying 'nah... that won't work' then it's a bigger problem20:51
dansmithwznoinsk: and like sdague said, it was broken all the time and never getting fixed20:51
*** salv-orlando has quit IRC20:51
wznoinskdims: sure, slightly different use case when using magnum tho20:51
*** rmart04 has left #openstack-nova20:52
*** salv-orlando has joined #openstack-nova20:52
dimswznoinsk : one blocker was mounting of cinder volumes last i checked needed some more work on docker side too20:52
dansmithbauzas: it's called, just with different than expected arguments20:52
dansmithbauzas: did you see my change?20:52
dimswznoinsk : right20:52
wznoinskdansmith: I can understand not a good fit inside nova where 'broken all the time/nobody interested' is probably the result of that20:52
dansmithwznoinsk: I don't think they were related tbh20:53
dansmithwznoinsk: the broken all the time was just a matter of nobody giving it attention for the things that it could do (which was a small subset)20:53
bauzasdansmith: https://review.openstack.org/#/c/246022/1/nova/tests/unit/objects/test_objects.py,cm ? sure20:53
dimswznoinsk : it worked fine for a specific set of scenarios for some folks. then they all went away20:53
dansmithbauzas: this: https://review.openstack.org/#/c/246022/1..2/nova/tests/unit/objects/test_instance.py,cm20:54
*** penick has joined #openstack-nova20:54
wznoinskdansmith: re fitting nova, does that mean supporting different features/request (the ones for containers) than nova was designed for originally?20:54
dansmithwznoinsk: didn't provide things that nova generally considers table stakes, and nova doesn't expose the things that makes docker useful for most people20:55
dansmithwznoinsk: thus, the intersection between the two is pretty small20:55
wznoinskor because nova scheduling is good for VMs but not necessarily for quick spinning up containers? (religion discussion alert)20:55
melwittwznoinsk: fwiw, I thought I remember dims reached out on the ML to ask if anyone was interested in bringing the feature back and it was discussed at the nova midcycle before last, and we couldn't find anyone interested in the driver20:55
bauzasdansmith: okay, fair enough...20:56
wznoinskdansmith: ok, understood20:56
* bauzas hides20:56
wznoinskmelwitt: I'm interested in that only recently, don't feel guilty :-)20:56
dimswznoinsk : we were looking for help implementing all the mandatory features for a nova driver20:56
wznoinsks/, /, I20:56
*** mari0jv is now known as mariojv20:57
*** ccesario has joined #openstack-nova20:57
dimslike melwitt mentioned, folks did not show up20:57
wznoinskdims: I would need to look deeper on what exactly we can do there still using nova api, I'm thinking having docker containers as compute nodes may have it's valid use cases but only if you can make use of their full potential20:58
dansmithwell, you can't and you never will, if that helps :)20:59
dansmithwznoinsk: and you mean s/nodes/instances/ right?20:59
wznoinskdansmith: that's my impression after your point few lines back ;-)20:59
dansmithbecause nova-docker was not about running compute nodes in docker containers20:59
dansmithwznoinsk: I don't suffer from subtlety :)21:00
*** dims_ has joined #openstack-nova21:00
openstackgerritChet Burgess proposed openstack/nova: Use --concurrent with ebtables  https://review.openstack.org/24658021:00
melwittwznoinsk: yeah, nova-docker means it will put a barebones container on a compute node when you request an instance. so you get a container instead of a VM21:00
wznoinskdansmith: I was using nodes interchangeable with instances because a container to me is closer to a baremetal but I guess that might have been a wrong wording21:01
*** ctrath has quit IRC21:01
wznoinskdansmith, melwitt: sure I know that, bad wording21:02
dansmithwznoinsk: nodes is not the right term here and comes with a lot of other meanings :)21:02
melwittwell, "compute node" anyway21:02
*** aginwala has joined #openstack-nova21:02
dansmithwell, nodes too21:02
*** dims has quit IRC21:02
dansmithbut anyway :)21:02
melwittokay :P21:03
cburgess_sdague So... I think maybe I did the release notes thing correct, please take a look at let me know21:03
wznoinskI'd consider 'node' more general term - 'a node' of any kind, where compute instance is of some specific type like instatiation of nova compute21:03
sdaguecburgess_: yeh, I think that's right21:03
sdagueexcept for the libvirt version you wrote :)21:03
dansmithmelwitt: https://github.com/openstack/nova/blob/master/nova/db/api.py#L48121:03
cburgess_sdague Whats the version?21:03
dansmithwznoinsk: but it's used in nova for a specific thing21:03
cburgess_sdague OMG21:04
cburgess_sdague Thats what I get for doing this while on an annoying conference call /sigh21:04
mriedemso we're brining nova-docker back?!21:04
dansmithwznoinsk: see what you did?21:04
* wznoinsk sets eyes down21:04
openstackgerritChet Burgess proposed openstack/nova: Use --concurrent with ebtables  https://review.openstack.org/24658021:05
cburgess_sdague fixed /sigh21:05
*** pixelbeat has quit IRC21:05
wznoinskmriedem: if you ask me, probably (for sure) not21:05
melwittwznoinsk: I personally see some usefulness there, even if you can't control the container much with nova api after you create it. but I thought we abandoned the idea for lack of interest from people wanting to use it when there is magnum out there. and we do have lxc driver, so maybe that covered the usecase21:05
*** doug-fish has joined #openstack-nova21:06
mriedemdansmith: ctrath and doffm_ want to integrate CIM with nova :)21:06
melwittdansmith: oh yeah, host + node!21:07
mriedemmelwitt: there is also lxd out of tree21:07
dansmithmriedem: come on dude, am I not having a bad enough day already?21:07
mriedemdansmith: heh21:07
wznoinskmelwitt: with support of docker container format you get some use cases... my thinking was slightly further tho... I'm getting my head around lkvm, clear container have unofficial docker support which I wanted to try in Openstack21:07
*** ctrath has joined #openstack-nova21:07
mriedemmelwitt: we also have virtuozzo in tree21:08
melwittmriedem: ah, right21:08
mriedemthose are all hidden in the libvirt driver though21:08
*** rushil has quit IRC21:08
wznoinskso in short while I see use cases for compute (hic!) instances being containers, that's not necessarily native docker containers I'm interested in21:08
dansmiththank goodness21:08
*** salv-orlando has quit IRC21:08
sdaguecburgess_: +221:09
cburgess_sdague cool21:09
sdagueanyone else for that?21:09
openstackgerritChet Burgess proposed openstack/nova: ebtables/libvirt workaround  https://review.openstack.org/24658121:09
cburgess_sdague Well so there is the ugly workaround too https://review.openstack.org/#/c/246581/21:09
cburgess_sdague So we need to figure out what to do with this.21:10
sdagueyeh, I'll review in the morning, I need to do an errand21:10
cburgess_sdague Cool21:10
cburgess_sdague Once that merges we probably should update your devstack fix to actually directly use the nova stuff.21:10
*** salv-orlando has joined #openstack-nova21:10
*** sacharya_ has quit IRC21:11
*** sacharya has joined #openstack-nova21:11
wznoinskwith that scheduler refactoring and possibly supporting more schedulers, should there be nova api extended too?21:12
mriedemnative VMs modeled with CIM...i can see it all now21:12
ctrathmriedem: already working on it21:12
*** lpetrut has quit IRC21:12
wznoinsksdague, dansmith ^21:13
dansmithwznoinsk: no21:13
mriedemlike....nova api extensions...21:13
mriedemwhere is jaypipes when you want him21:13
mriedemprobably writing nova api extensions21:13
wznoinskguys, apologies if some parts are not making sense ... I'm not close to nova's and it's code21:15
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE  https://review.openstack.org/24505021:16
mriedemyou're not by chance a phd student are you?21:16
*** Nakato_ is now known as Nakato21:16
wznoinskmriedem: I didn't know I sound that bad21:16
mriedemheh, no offence of course21:17
mriedemthere is just usually this pattern of, 'hey, i'm new, have you guys ever thought about doing x? no. oh, well, i'm thinking about doing it for my thesis.'21:17
mriedemwznoinsk: there are some people running nova-docker thingies, but it's really not maintained past probably kilo21:18
*** electrocucaracha has joined #openstack-nova21:21
*** smatzek has quit IRC21:22
*** rpodolyaka1 has joined #openstack-nova21:23
*** shaohe_feng has quit IRC21:23
wznoinskmriedem: I'm looking it more in terms of provisioning containers (as being between baremetal and VMs) but without all the clutter of magnum, coe21:23
wznoinskin term of NFVI if you (for some reason) like NFV terms21:24
dims_mriedem : still works with liberty :)21:24
*** vladikr has quit IRC21:25
*** kfarr has quit IRC21:25
dims_wznoinsk : there's a nova-docker channel if we want to avoid ruffling feathers :)21:26
melwittmy feathers remain unruffled21:27
wznoinskdims_: didn't know, I'm not sure I have anything interesting to say/ask as of now tho, I'll catch you when I do21:28
mriedemdims_: it does?21:28
dims_wznoinsk ack21:29
jaypipesmriedem: sorry, what's up?21:29
*** aginwala has quit IRC21:29
openstackgerritAndrew Laski proposed openstack/nova: Add persistence to the RequestSpec object  https://review.openstack.org/21175321:29
*** eglynn has joined #openstack-nova21:29
mriedemdims_: yeah i guess it must http://logs.openstack.org/54/245954/1/check/gate-tempest-dsvm-docker/993481b/21:29
mriedemjaypipes: nothing21:29
dims_mriedem i cut a branch too https://github.com/openstack/nova-docker/tree/stable/liberty21:29
mriedemdims_: heh, stable/icehouse https://review.openstack.org/#/c/236508/21:30
mriedemtime to put that one out to pasture21:30
dims_y gotta get rid of that one21:30
dims_mriedem may be if wznoinsk comes back :)21:31
wznoinskdims_: on nova-docker most likely not, I'm afraid on touching something in nova cores see limitations on ;-)21:31
dims_wznoinsk was just kidding, no worries21:32
*** krtaylor has quit IRC21:32
jroll"limitations" is a fun word there :)21:32
* jroll sees node/host discussion21:32
*** rpodolyaka1 has quit IRC21:32
dansmithmelwitt: where is instance 1.23 requested21:33
melwittdansmith: forgive the ignorance that will follow21:34
dansmithmelwitt: nope, I must be missing something21:34
*** sneti_ has joined #openstack-nova21:36
dansmithalaski: rs236.mailgun.us is on a blacklist21:36
melwittdansmith: but I thought if you do objects.InstanceV1.get_by_uuid() where it has VERSION = '1.23', and eventually object_class_action_versions is called to de/serialize, how does object_versions contain 2.0 and give that back21:36
*** sneti has quit IRC21:36
dansmithmelwitt: well, first off, objects.InstanceV1 shouldn't be a thing21:36
alaskidansmith: do you know which one?21:36
*** tonytan4ever has joined #openstack-nova21:37
melwittdansmith: I thought '1.23' would get passed to object_class_action_versions and it would in turn return the right version back21:37
dansmithalaski: http://mxtoolbox.com/SuperTool.aspx?action=blacklist%3ars236.mailgun.us&run=toolpage21:37
dansmithmelwitt: nope, when we go to call object_class_action_versions() we'll get the current process's manifest21:38
dansmithmelwitt: which will be whatever the highest version is that is supported21:38
melwittdansmith: okay, so that explains that. so next I wonder, how did it get 1.x back in o.vo 0.12.0?21:38
alaskidansmith: there are a lot more blacklists there than I remember.  I'll send out a signal, though they're probably already aware.21:39
dansmithmelwitt: well, there was no ovo support for object_class_action_versions() in 0.12.021:39
dansmithmelwitt: I had this all worked out yesterday, but it'll take me a minute to figure out all the mocking21:40
dansmithalaski: just FYI.. it's delaying a confirmation email for me, which is annoying me.. and I was already annoyed :)21:40
*** aysyd has quit IRC21:40
melwittdansmith: sorry, didn't want to cause more work. I guess before the object_class_action_versions there was kind of a no-op21:40
dansmithmelwitt: I think that what was happening here is that we were falling back to the non-versioned case21:41
dansmither, the non-manifest case21:41
alaskidansmith: sure.  But I still maintain a slight interest in our email.  Though I started on OpenStack right after mailgun joined.21:41
*** aginwala has joined #openstack-nova21:42
dansmithalaski: yeah, hence the ping :D21:42
*** pushkaru has quit IRC21:42
*** pushkaru has joined #openstack-nova21:42
*** mrkz has quit IRC21:43
*** rpodolyaka1 has joined #openstack-nova21:44
dansmithmelwitt: ah, right, so,21:44
dansmithmelwitt: because o.vo wasn't going to support this before liberty,21:44
dansmithmelwitt: I made the conductor rpcapi redirect the non-manifest call to the manifest call, first looking up the manifest21:45
*** mrkz has joined #openstack-nova21:45
*** krtaylor has joined #openstack-nova21:45
dansmithmelwitt: https://github.com/openstack/nova/blob/master/nova/conductor/rpcapi.py#L226-L23321:45
dansmithmelwitt: so now that 0.13 is out, we're hitting o.vo's fixture for class_action_versions() directly21:46
dansmithwhich I think means we're looking up the manifest not right at the call time, but earlier21:47
*** electrocucaracha has quit IRC21:47
dansmithbefore we do the indirection swap21:47
*** electrocucaracha has joined #openstack-nova21:48
*** dustins has quit IRC21:48
melwittdansmith: and the pre 0.13 manifest had 1.23 as the Instance version and post 0.13 has 2.0, and that's because of when the manifest is looked up?21:49
melwittor you're saying it's before the objects.Instance gets set to objects.InstanceV121:49
dansmithI think it's because of when we generate the manifest21:50
*** tongli has quit IRC21:50
melwittokay, I'll look at that. for my own curiosity21:51
*** mtanino has quit IRC21:51
*** dims_ has quit IRC21:51
*** electrocucaracha has quit IRC21:51
*** rpodolyaka1 has quit IRC21:52
dansmithmelwitt: print out the manifest in conductor's rpc and in the new fixture21:53
*** Marga_ has quit IRC21:53
dansmithmelwitt: you think that there's something actually wrong or what?21:53
melwittdansmith: no, I understand it has to do with how the test runs now. I meant I'm going to look at the manifest generation timing for my own knowledge21:56
*** dims has joined #openstack-nova21:57
*** sneti_ has quit IRC21:57
*** thangp has quit IRC21:59
*** w_verdugo has quit IRC22:01
*** penick has quit IRC22:01
*** electrocucaracha has joined #openstack-nova22:02
*** angdraug has quit IRC22:04
*** baoli has joined #openstack-nova22:06
*** rpodolyaka1 has joined #openstack-nova22:06
*** penick has joined #openstack-nova22:06
*** rushil has joined #openstack-nova22:06
*** doffm_ is now known as doffm22:06
*** rushil_ has joined #openstack-nova22:07
*** mtanino has joined #openstack-nova22:08
*** eliqiao has quit IRC22:09
*** eliqiao has joined #openstack-nova22:10
*** sneti has joined #openstack-nova22:10
*** sacharya has quit IRC22:11
*** baoli has quit IRC22:11
*** rushil has quit IRC22:11
*** akshai_ has joined #openstack-nova22:12
*** pkholkin has quit IRC22:13
*** rlrossit has left #openstack-nova22:13
akshai_has anyone tried nova network with liberty?22:13
*** sacharya has joined #openstack-nova22:14
*** vilobhmm has quit IRC22:14
*** vilobhmm has joined #openstack-nova22:14
*** rpodolyaka1 has quit IRC22:15
*** vilobhmm has quit IRC22:15
*** vilobhmm has joined #openstack-nova22:16
*** pkholkin has joined #openstack-nova22:16
*** rushil_ has quit IRC22:16
*** vilobhmm has quit IRC22:17
*** vilobhmm has joined #openstack-nova22:17
*** thorst has quit IRC22:19
*** otter768 has joined #openstack-nova22:19
*** thorst has joined #openstack-nova22:19
*** rcernin has quit IRC22:21
*** edleafe_ has quit IRC22:21
*** thorst has quit IRC22:23
*** otter768 has quit IRC22:24
*** edleafe_ has joined #openstack-nova22:25
*** browne has quit IRC22:27
*** rushil has joined #openstack-nova22:27
*** lykinsbd has quit IRC22:30
*** stackdump has joined #openstack-nova22:30
*** sneti has quit IRC22:31
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE  https://review.openstack.org/24505022:33
*** Marga_ has joined #openstack-nova22:33
*** mc_nair has quit IRC22:34
*** tdurakov has quit IRC22:35
*** tdurakov has joined #openstack-nova22:36
mriedemhmm, so it looks like when we delete instances, we don't delete instance_actions22:36
mriedemand that breaks the archive stuff22:36
bauzaswill look at that tomorrow22:38
*** mlavalle has quit IRC22:40
openstackgerritMatt Riedemann proposed openstack/nova: WIP: Reverse sort tables before archiving  https://review.openstack.org/24663522:40
mriedembauzas: yeah, found it with ^22:41
*** mlavalle has joined #openstack-nova22:41
dimsmriedem is this worth merging? https://review.openstack.org/#/c/244827/22:42
mriedemdims: yeah, the dict_value kwarg was something i moved to o-i from novaclient actually :)22:43
mriedemso we could replace novaclient's print_dict util with the one from o-i22:43
mriedemthat's kind of a lost cause now22:43
*** ccarmack has quit IRC22:44
mriedemdims: you know you need to provide the git hashes in the commit message right? :P22:44
mriedemdon't let the last o-i sync break all of the pleasure of doing those22:44
dimsit's going away dude :)22:44
*** apoorvad has quit IRC22:44
*** burgerk has joined #openstack-nova22:45
openstackgerritDavanum Srinivas (dims) proposed openstack/python-novaclient: Last sync from oslo-incubator  https://review.openstack.org/24482722:45
*** burgerk_ has quit IRC22:46
dimsmriedem luckily i had that handy :)22:46
mriedemwait a minute22:46
mriedemthat's not the git log of the actual changes in cliutils22:46
*** annegentle has quit IRC22:47
dimsmriedem that was the SHA when i ran the sync script22:47
*** annegentle has joined #openstack-nova22:47
*** claudiub has quit IRC22:47
*** lucas-dinner is now known as lucas-afk22:48
*** edtubill has quit IRC22:48
openstackgerritMiguel Lavalle proposed openstack/nova-specs: Integrate DNS resolution with Neutron using Nova hostname  https://review.openstack.org/9015022:49
*** aginwala has quit IRC22:49
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Last sync from oslo-incubator  https://review.openstack.org/24482722:49
mriedemdims: now +222:49
dimsmriedem thanks!22:50
mriedemtime to head out22:50
dimsgood night22:50
*** liverpooler has quit IRC22:53
*** mriedem has quit IRC22:53
*** thumpba has joined #openstack-nova22:54
*** liverpooler has joined #openstack-nova22:54
*** aginwala has joined #openstack-nova22:54
*** mwagner has quit IRC22:55
*** stackdump has quit IRC22:56
cburgess_Is the citrix xenserver CI job voting?22:56
cburgess_Sorry I meant the XenProject CI job22:56
*** stackdump has joined #openstack-nova22:57
*** burgerk has quit IRC22:58
*** liverpooler has quit IRC23:00
*** dims has quit IRC23:00
*** dims has joined #openstack-nova23:01
*** jhesketh has quit IRC23:01
*** mdrabe has quit IRC23:01
*** mhorban__ has joined #openstack-nova23:02
*** vilobhmm has quit IRC23:04
*** vilobhmm has joined #openstack-nova23:04
*** dims_ has joined #openstack-nova23:04
cfriesen__anyone left on here?  I'm curious why we don't allow resize-to-same-node by default?23:04
*** jhesketh has joined #openstack-nova23:04
*** aginwala_ has joined #openstack-nova23:05
*** sacharya_ has joined #openstack-nova23:05
*** sacharya has quit IRC23:05
*** apoorvad has joined #openstack-nova23:05
*** thumpba has quit IRC23:07
*** dims has quit IRC23:07
*** aginwala has quit IRC23:08
*** thumpba has joined #openstack-nova23:08
*** ctrath has quit IRC23:08
*** aginwala_ has quit IRC23:09
*** aginwala has joined #openstack-nova23:09
*** exploreshaifali has quit IRC23:10
*** gokrokve has joined #openstack-nova23:11
*** burt has quit IRC23:14
*** tonytan4ever has quit IRC23:14
*** ctrath has joined #openstack-nova23:14
*** gokrokve has quit IRC23:15
*** pixelbeat has joined #openstack-nova23:19
*** gokrokve has joined #openstack-nova23:19
*** rook has joined #openstack-nova23:20
*** wuhg has joined #openstack-nova23:26
*** salv-orl_ has joined #openstack-nova23:28
*** tdurakov has quit IRC23:28
*** electrocucaracha has quit IRC23:28
melwittcfriesen__: my guess is because resize == cold migrate and you don't want it to migrate to the same host it's on, by default23:29
*** doug-fish has quit IRC23:30
*** jdurgin has joined #openstack-nova23:30
*** salv-orlando has quit IRC23:30
cfriesen__melwitt: why not?23:31
*** suro-patz has quit IRC23:31
cfriesen__i.e. how is it any worse23:31
*** aginwala_ has joined #openstack-nova23:31
*** mwagner has joined #openstack-nova23:31
*** alejandrito has quit IRC23:31
cfriesen__it seems like it can even offer opportunities for optimization...like if you're just adding memory/cpus you could skip all the disk stuff23:32
*** aginwal__ has joined #openstack-nova23:32
melwittcfriesen__: again guessing, because "migrate" implies you want the instance to move, not have it stay in the same place23:32
*** ctrath has quit IRC23:32
cfriesen__melwitt: but the end-user asked for a resize, not a migration.  so the fact that it shares code with cold migration is purely an implementation detail23:32
melwittcfriesen__: I agree if you're on the "resize" flow it would seem fine to let it stay on the same host23:32
*** electrocucaracha has joined #openstack-nova23:32
melwittso I'm not sure other than the fact that the share the code path23:33
cburgess_This issue comes up from time to time. I think the reason at this point is, because its always been that way and changing it would be a problem.23:33
cfriesen__cburgess: :)23:33
cburgess_As in no one wants to document/deal with the change.23:33
*** mlavalle has quit IRC23:34
melwittheh, I hope not. I don't know the history of why resize == cold migrate but decoupling at least some part of it for the initial checks makes sense to me23:34
*** thumpba has quit IRC23:34
*** aginwala has quit IRC23:34
*** w_verdugo has joined #openstack-nova23:34
cburgess_I could be wrong.23:35
cburgess_Seems like this issue comes up from time to time and we never change it.23:35
cfriesen__melwitt: they do share a lot of the code...resize is basically a cold migration to a different flavor (but optionally to the same host)23:35
cfriesen__melwitt: sort of similar to how evacuate and rebuild share a lot of code23:35
*** aginwala_ has quit IRC23:35
dansmithcfriesen__: resize is not a resize in place23:36
dansmithcfriesen__: it's a two-phase commit,23:36
dansmithwhich means you do the disk part anyway23:36
dansmithwe could do it a little more efficiently if it was on the same box,23:36
melwittcfriesen__: ah, okay23:36
dansmithbut since lots of people pack smalls with smalls and larges with larges, it means we have to consult the scheduler23:36
cfriesen__dansmith: but we still consult the scheduler with the option set to true23:37
dansmithso it's largely the reason that cburgess_ said, but also because it was designed to be one thing23:37
*** mrkz has quit IRC23:37
*** ijuwang has quit IRC23:37
dansmithcfriesen__: right, I understand, I meant the reason why it's a migrate under the covers23:37
dansmithcfriesen__: the same-host thing was added purely so we could test it in devstack23:38
*** EinstCrazy has joined #openstack-nova23:39
cfriesen__dansmith: yeah, I got that.  I just don't see the downside of allowing it (and even removing the option and setting it permanently to true).  Aside from the fact that it's a change, I guess.23:39
melwittI wonder if it could be as simple as "if user requested resize then treat it as allow same host" for the later check23:39
*** signed8__ has quit IRC23:42
dansmithcfriesen__: I don't see any reason why one default is better than another23:43
dansmithcfriesen__: because as you say, people who assume the current default will have a change anyway23:44
*** josecastroleon has joined #openstack-nova23:45
*** mhorban__ has quit IRC23:46
*** jgjhgsonchn88523 has quit IRC23:46
*** mhorban__ has joined #openstack-nova23:46
cfriesen__dansmith: I'm actually not seeing why anyone would ever want to prevent a resize to the same host.  (Whereas cold migration to same host only makes sense for development)23:47
cburgess_cfriesen__ As dansmith said, scheduling related reasons might block it.23:47
dansmiththere was some other reason a while ago, and I'm blanking on it23:47
cfriesen__cburgess: then that's fine, it'll go to another host23:47
cburgess_cfriesen__ Some people setup aggregates for running only a specific flavor and target every flavor to a specific aggregate.23:47
cburgess_So if you resize you need to move.23:48
cfriesen__cburgess: which is fine, since we'd still go through the scheduler23:48
dansmithI want to say it also screws with quotas or reservations or something23:48
cburgess_dansmith At one point it was broken locally. But that was fixed long ago.23:48
*** EinstCrazy has quit IRC23:48
cburgess_dansmith unrelated, is xenproject CI voting?23:49
dansmithcburgess_: running or voting?23:49
cburgess_dansmith voting? As in do I need to care it gave me a -1?23:49
dansmithcburgess_: no23:49
*** salv-orl_ has quit IRC23:49
dansmithcburgess_: it giving you a -1 means it's voting, but it's not gating23:50
dansmithcburgess_: and also, it's been failing a lot recently, IIRC23:50
cburgess_dansmith yeah the error looked totally unrelated to me but I didn't know if that -1 would be a bad thing or not.23:50
*** salv-orlando has joined #openstack-nova23:50
*** gokrokve has quit IRC23:51
*** ZZelle_ has quit IRC23:58

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