Wednesday, 2015-11-18

*** sacharya_ has quit IRC00:01
*** mc_nair has joined #openstack-nova00:02
*** aginwal__ has quit IRC00:02
openstackgerritZhongcheng Lao proposed openstack/nova: VMware: Omit exceptions during snapshot cleanup
*** jaypipes has quit IRC00:04
*** akshai_ has quit IRC00:05
*** pushkaru has quit IRC00:07
*** pixelbeat has quit IRC00:09
*** pixelbeat has joined #openstack-nova00:10
*** aginwala has joined #openstack-nova00:10
*** derekh has joined #openstack-nova00:13
*** josecastroleon has quit IRC00:14
*** aginwala has quit IRC00:14
*** aginwala_ has joined #openstack-nova00:15
*** mc_nair has quit IRC00:15
*** aginwala_ has quit IRC00:15
*** aginwala has joined #openstack-nova00:18
*** stackdump has quit IRC00:18
*** rushil has quit IRC00:19
*** penick has quit IRC00:20
*** otter768 has joined #openstack-nova00:20
*** derekh has quit IRC00:20
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements
*** otter768 has quit IRC00:25
*** penick has joined #openstack-nova00:27
*** aginwala has quit IRC00:29
*** shaohe_feng has joined #openstack-nova00:39
openstackgerritAndrew Laski proposed openstack/nova: Add persistence to the RequestSpec object
*** shaohe_feng1 has joined #openstack-nova00:43
*** chenli has joined #openstack-nova00:44
*** mhorban__ has quit IRC00:44
*** mhorban__ has joined #openstack-nova00:44
*** thorst has joined #openstack-nova00:44
*** shaohe_feng has quit IRC00:45
*** leseb has quit IRC00:47
*** pratikmallya has quit IRC00:48
*** rushil has joined #openstack-nova00:48
*** thorst has quit IRC00:51
*** thorst has joined #openstack-nova00:51
*** leseb has joined #openstack-nova00:52
*** RichardRaseley has quit IRC00:53
*** jerrygb has quit IRC00:55
*** thorst has quit IRC00:56
*** aginwala has joined #openstack-nova00:56
*** electrocucaracha has quit IRC00:56
*** armax has joined #openstack-nova00:57
openstackgerritli,chen proposed openstack/nova: Add unit tests for nova.virt.firewall.IpTablesFirewallDriver
*** electrocucaracha has joined #openstack-nova00:59
*** Kennan has quit IRC00:59
*** rushil has quit IRC01:01
*** Kennan has joined #openstack-nova01:03
*** liusheng has joined #openstack-nova01:03
*** hemna is now known as hemnafk01:04
*** mriedem has joined #openstack-nova01:05
*** thumpba has joined #openstack-nova01:06
*** zhangjn has joined #openstack-nova01:06
*** EinstCrazy has joined #openstack-nova01:07
openstackgerritZhenyu Zheng proposed openstack/nova: Add project-id and user-id when list server-groups
*** ljxiash has joined #openstack-nova01:07
*** jamielennox is now known as jamielennox|away01:10
openstackgerritZhenyu Zheng proposed openstack/nova: Split-network-plane-for-live-migration
*** thumpba has quit IRC01:12
*** electrocucaracha has quit IRC01:12
*** ildikov has quit IRC01:15
*** ildikov has joined #openstack-nova01:16
*** shaohe_feng1 has quit IRC01:16
*** penick has quit IRC01:19
*** thorst has joined #openstack-nova01:23
*** thorst has quit IRC01:23
*** aginwala has quit IRC01:23
*** stevemar_ has quit IRC01:23
*** thorst has joined #openstack-nova01:23
*** shaohe_feng has joined #openstack-nova01:24
*** thorst has quit IRC01:24
*** Sree has joined #openstack-nova01:24
*** Sree has quit IRC01:24
*** thorst has joined #openstack-nova01:25
*** Sree has joined #openstack-nova01:25
*** oomichi has joined #openstack-nova01:25
*** aginwala has joined #openstack-nova01:27
*** aginwala has quit IRC01:27
*** thorst has quit IRC01:29
*** Sree has quit IRC01:29
*** oomichi has quit IRC01:31
mriedemgaryk: fyi, looks like intermittent vmware ci failure on image create
*** jaypipes has joined #openstack-nova01:33
mriedemooo glance_store bug
*** shaohe_feng has quit IRC01:34
mriedemgaryk: fyi
openstackLaunchpad bug 1517275 in glance_store "VMware NSX CI failing on image create with "AttributeError: HTTPResponse instance has no attribute 'body'"" [Undecided,New]01:36
*** eharney has quit IRC01:38
*** nic has quit IRC01:38
*** tpeoples has quit IRC01:40
*** tpeoples_ is now known as tpeoples01:40
*** ssurana has quit IRC01:40
*** mhorban__ has quit IRC01:41
*** zhenguo has joined #openstack-nova01:42
*** mhorban__ has joined #openstack-nova01:42
*** annegentle has quit IRC01:47
*** pratikmallya has joined #openstack-nova01:48
openstackgerritYingxin Cheng proposed openstack/nova: Use stevedore for scheduler host manager
*** boris-42 has joined #openstack-nova01:51
*** pratikmallya has quit IRC01:53
*** otter768 has joined #openstack-nova01:53
*** penick has joined #openstack-nova01:58
*** rushil has joined #openstack-nova02:03
*** baoli has joined #openstack-nova02:03
*** sacharya has joined #openstack-nova02:05
*** apoorvad has quit IRC02:07
*** rushil has quit IRC02:10
*** Marga_ has quit IRC02:11
*** vilobhmm has quit IRC02:12
*** stevemar_ has joined #openstack-nova02:14
*** apoorvad has joined #openstack-nova02:15
*** jwcroppe has quit IRC02:21
mriedemBobBall: seems we could exclude boto s3 tests for xen project CI?
*** shaohe_feng has joined #openstack-nova02:22
*** unicell1 has quit IRC02:22
mriedemactually that's not an s3 test failure, looks like snapshot fails02:24
*** baoli_ has joined #openstack-nova02:25
openstackLaunchpad bug 1517282 in OpenStack Compute (nova) "ImagesOneServerTestJSON.test_create_delete_image fails in dsvm-tempest-xen due to snapshot failure" [Undecided,New]02:27
*** penick has quit IRC02:27
*** baoli has quit IRC02:28
*** aginwala has joined #openstack-nova02:28
*** apoorvad has quit IRC02:28
*** davideagnello has quit IRC02:28
*** davideagnello has joined #openstack-nova02:29
*** jerrygb has joined #openstack-nova02:29
*** 7JTAA216T has joined #openstack-nova02:30
*** dims_ has quit IRC02:30
*** 7JTAA216T has quit IRC02:30
*** 21WAACTNP has joined #openstack-nova02:31
*** rubasov has quit IRC02:31
*** edleafe_ has quit IRC02:31
*** jyuso1 has joined #openstack-nova02:32
*** aginwala has quit IRC02:32
*** ccarmack has joined #openstack-nova02:33
*** rk4n has quit IRC02:34
mriedemmnestratov: your virtuozzo ci doesn't need to be running heat services, fyi02:34
*** rk4n has joined #openstack-nova02:34
*** annegentle has joined #openstack-nova02:38
*** rk4n has quit IRC02:39
*** mhorban__ has quit IRC02:39
*** mhorban__ has joined #openstack-nova02:40
*** stevemar_ has quit IRC02:42
*** fawadkhaliq has joined #openstack-nova02:42
*** yonglihe has quit IRC02:42
*** stevemar_ has joined #openstack-nova02:42
*** annegentle has quit IRC02:46
*** klkumar has joined #openstack-nova02:53
openstackgerritAugustina Ragwitz proposed openstack/nova: WIP: update functional tests to support optional project id
*** thorst has joined #openstack-nova02:57
*** thorst has quit IRC02:57
*** 21WAACTNP has quit IRC03:01
*** haomaiwang has joined #openstack-nova03:01
*** pixelbeat has quit IRC03:01
*** pixelbeat has joined #openstack-nova03:02
*** aginwala has joined #openstack-nova03:04
*** venkat_p has joined #openstack-nova03:04
*** venkat_p has left #openstack-nova03:06
*** rfolco has joined #openstack-nova03:08
*** aginwala_ has joined #openstack-nova03:09
*** browne has joined #openstack-nova03:09
*** aginwala_ has quit IRC03:10
*** ijw has quit IRC03:10
*** aginwala has quit IRC03:12
*** aginwala has joined #openstack-nova03:12
*** aginwala has quit IRC03:16
*** Sree has joined #openstack-nova03:20
*** shaohe_robot has quit IRC03:20
*** Sree has quit IRC03:21
*** rfolco has quit IRC03:21
*** Sree has joined #openstack-nova03:21
*** Sree has quit IRC03:22
*** guilh has quit IRC03:22
*** klkumar has quit IRC03:22
*** Sree has joined #openstack-nova03:24
*** achanda has joined #openstack-nova03:24
*** tpeoples is now known as tpeoples_away03:27
*** tjones has joined #openstack-nova03:28
*** tjones has left #openstack-nova03:28
*** liusheng has quit IRC03:35
*** klkumar has joined #openstack-nova03:35
*** gjayavelu has joined #openstack-nova03:35
*** mhorban__ has quit IRC03:37
*** chenli_ has joined #openstack-nova03:38
openstackgerritJianghua Wang proposed openstack/nova: xenapi: OVS agent updates the wrong port when using XenServer + Neutron
*** mhorban__ has joined #openstack-nova03:40
*** pixelbeat has quit IRC03:41
openstackgerritMatt Riedemann proposed openstack/nova: xen: mask passwords in volume connection_data dict
*** chenli has quit IRC03:42
*** akshai has joined #openstack-nova03:44
*** sacharya has quit IRC03:46
*** liusheng has joined #openstack-nova03:48
*** achanda has quit IRC03:48
*** Sree has quit IRC03:49
*** Sree has joined #openstack-nova03:49
mriedemjaypipes: you don't want to keep instance actions around forever, like in sandlot forever03:52
*** jhesketh has quit IRC03:52
*** jhesketh has joined #openstack-nova03:53
*** mriedem has quit IRC03:55
*** achanda has joined #openstack-nova03:55
*** haomaiwang has quit IRC04:01
*** haomaiwang has joined #openstack-nova04:01
*** Sree has quit IRC04:03
*** Sree has joined #openstack-nova04:04
*** fawadkhaliq has quit IRC04:06
*** leseb has quit IRC04:16
*** leseb has joined #openstack-nova04:18
*** hparekh2 has quit IRC04:20
*** leseb has quit IRC04:22
*** leseb has joined #openstack-nova04:25
*** gjayavelu has quit IRC04:27
*** edtubill has joined #openstack-nova04:27
*** fawadkhaliq has joined #openstack-nova04:27
*** hparekh has joined #openstack-nova04:27
*** abhishekk has joined #openstack-nova04:28
*** amotoki has joined #openstack-nova04:34
*** Sree_ has joined #openstack-nova04:34
*** mhorban__ has quit IRC04:34
*** Sree_ is now known as Guest5330304:35
*** mhorban__ has joined #openstack-nova04:35
*** chenli_ has quit IRC04:36
*** Guest53303 has quit IRC04:37
*** zhangjn has quit IRC04:37
*** Sree has quit IRC04:37
*** Sree has joined #openstack-nova04:38
*** zhangjn has joined #openstack-nova04:38
*** edtubill has quit IRC04:38
*** leseb has quit IRC04:38
*** markvoelker_ has quit IRC04:40
*** leseb has joined #openstack-nova04:41
*** annegentle has joined #openstack-nova04:47
*** otter768 has quit IRC04:47
*** amotoki has quit IRC04:49
*** chhavi has joined #openstack-nova04:51
*** leseb has quit IRC04:52
*** vilobhmm has joined #openstack-nova04:52
*** annegentle has quit IRC04:52
*** leseb has joined #openstack-nova04:53
*** jamielennox|away is now known as jamielennox04:54
*** mdrabe has joined #openstack-nova04:55
*** haomaiwang has quit IRC05:01
*** amotoki has joined #openstack-nova05:01
*** haomaiwang has joined #openstack-nova05:01
*** gcb has joined #openstack-nova05:01
*** Sree has quit IRC05:02
*** jwcroppe has joined #openstack-nova05:03
*** Sree has joined #openstack-nova05:04
*** akshai has quit IRC05:05
*** vilobhmm has quit IRC05:07
*** mdrabe has quit IRC05:09
*** mtanino has quit IRC05:09
*** amotoki has quit IRC05:10
*** achanda has quit IRC05:12
*** amotoki has joined #openstack-nova05:14
*** Marga_ has joined #openstack-nova05:20
*** jaypipes has quit IRC05:20
*** Marga_ has quit IRC05:20
*** Marga_ has joined #openstack-nova05:21
*** vilobhmm has joined #openstack-nova05:21
*** jerrygb has quit IRC05:25
*** achanda has joined #openstack-nova05:25
*** zhangjn_ has joined #openstack-nova05:26
*** zhangjn has quit IRC05:27
*** stevemar_ has quit IRC05:27
*** stevemar_ has joined #openstack-nova05:28
*** krtaylor has quit IRC05:29
*** achanda has quit IRC05:29
*** whenry has quit IRC05:30
*** stevemar_ has quit IRC05:30
*** mhorban__ has quit IRC05:32
*** mhorban__ has joined #openstack-nova05:32
*** rpodolyaka1 has joined #openstack-nova05:33
*** achanda has joined #openstack-nova05:35
*** zhangjn_ has quit IRC05:37
*** rpodolyaka1 has quit IRC05:37
*** zhangjn has joined #openstack-nova05:38
*** rpodolyaka1 has joined #openstack-nova05:38
*** amotoki has quit IRC05:40
*** rpodolyaka1 has quit IRC05:46
*** jwcroppe has quit IRC05:50
*** jwcroppe has joined #openstack-nova05:50
*** salv-orl_ has joined #openstack-nova05:50
*** rpodolyaka1 has joined #openstack-nova05:51
*** deepakcs has joined #openstack-nova05:52
*** achanda has quit IRC05:53
*** salv-orlando has quit IRC05:53
*** jwcroppe has quit IRC05:54
*** salv-orl_ has quit IRC05:55
*** achanda has joined #openstack-nova05:58
*** haomaiwang has quit IRC06:01
*** haomaiwang has joined #openstack-nova06:01
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Imported Translations from Zanata
*** rpodolyaka1 has quit IRC06:06
*** suro-patz has joined #openstack-nova06:18
*** suro-patz has quit IRC06:21
*** changbl has quit IRC06:22
*** changbl has joined #openstack-nova06:23
*** Sree_ has joined #openstack-nova06:24
*** Sree_ is now known as Guest4066706:24
*** rpodolyaka1 has joined #openstack-nova06:25
*** tpeoples has joined #openstack-nova06:26
*** unicell has joined #openstack-nova06:26
*** vilobhmm1 has joined #openstack-nova06:26
*** Sree has quit IRC06:27
*** rpodolyaka1 has quit IRC06:28
*** tyagiprince has joined #openstack-nova06:28
*** vilobhmm has quit IRC06:29
*** mhorban__ has quit IRC06:30
*** rpodolyaka1 has joined #openstack-nova06:30
*** mhorban__ has joined #openstack-nova06:30
*** angdraug has joined #openstack-nova06:30
*** klkumar has quit IRC06:31
*** lpetrut has joined #openstack-nova06:33
openstackgerritJianghua Wang proposed openstack/nova: xenapi: OVS agent updates the wrong port when using XenServer + Neutron
*** ryandahp has joined #openstack-nova06:34
*** jinxing has joined #openstack-nova06:35
*** liusheng has quit IRC06:36
*** rcernin has joined #openstack-nova06:37
*** liusheng has joined #openstack-nova06:37
*** irenab has joined #openstack-nova06:38
*** tpeoples has quit IRC06:39
*** markvoelker has joined #openstack-nova06:40
*** ankit_ag has joined #openstack-nova06:40
*** rajesht has joined #openstack-nova06:41
*** haomaiwang has quit IRC06:42
*** markvoelker has quit IRC06:45
openstackgerritAbhishek Kekane proposed openstack/nova-specs: Improve performance of UnShelve API
*** korzen has joined #openstack-nova06:48
*** annegentle has joined #openstack-nova06:48
*** klkumar has joined #openstack-nova06:48
*** otter768 has joined #openstack-nova06:48
*** haomaiwang has joined #openstack-nova06:49
*** salv-orlando has joined #openstack-nova06:51
*** otter768 has quit IRC06:53
*** annegentle has quit IRC06:53
openstackgerritAbhishek Kekane proposed openstack/nova-specs: Improve performance of UnShelve API
*** markmc has quit IRC06:59
*** markmc has joined #openstack-nova06:59
*** rpodolyaka1 has quit IRC06:59
*** haomaiwang has quit IRC07:01
*** alex_klimov has joined #openstack-nova07:01
*** rpodolyaka1 has joined #openstack-nova07:01
*** haomaiwang has joined #openstack-nova07:01
*** Marga_ has quit IRC07:03
*** angdraug has quit IRC07:03
*** Marga_ has joined #openstack-nova07:05
openstackgerritMoshe Levi proposed openstack/nova: libvirt: mlnx_direct vif type removal
*** cfriesen__ has quit IRC07:08
*** tdurakov has joined #openstack-nova07:08
*** unicell1 has joined #openstack-nova07:08
*** unicell has quit IRC07:09
*** josecastroleon has joined #openstack-nova07:10
*** mleroy has quit IRC07:11
*** salv-orlando has quit IRC07:12
*** nkrinner has joined #openstack-nova07:14
*** armax has quit IRC07:16
*** garyk1 has joined #openstack-nova07:16
*** tdurakov has quit IRC07:18
*** mjura has joined #openstack-nova07:18
openstackgerritZhenyu Zheng proposed openstack/nova: Add project-id and user-id when list server-groups
*** tdurakov has joined #openstack-nova07:22
ankit_agHi All, A new nova subteam "Operators Bug Triage subteam" was announced to keep a curated list of critical OPS bugs that have reviews that need attention.
*** subscope has joined #openstack-nova07:24
ankit_agIs there any weekly meeting scheduled for this subteam "Operators Bug Triage subteam"07:24
*** Guest40667 has quit IRC07:24
*** Sree has joined #openstack-nova07:25
*** vilobhmm1 has quit IRC07:25
*** jerrygb has joined #openstack-nova07:25
*** mhorban__ has quit IRC07:27
*** mhorban__ has joined #openstack-nova07:28
ankit_agIs there any weekly meeting scheduled today for Nova Bugs ?07:29
*** Sree has quit IRC07:29
*** jerrygb has quit IRC07:30
*** tdurakov has quit IRC07:31
*** eglynn has quit IRC07:32
*** asselin__ has joined #openstack-nova07:32
ankit_agCan someone please point me the right place where I should ask to review pending Nova bugs. e.g
*** achanda has quit IRC07:32
*** asselin_ has quit IRC07:33
*** hparekh has quit IRC07:33
garyk1johnthetubaguy: mikal: the subteam has looked at
*** alexschm has joined #openstack-nova07:34
*** korzen has quit IRC07:34
*** mleroy has joined #openstack-nova07:38
*** mhorban__ has quit IRC07:39
*** hparekh has joined #openstack-nova07:40
openstackgerritSergey Nikitin proposed openstack/nova: Replaced private field in get_session/engine with public method
*** markvoelker has joined #openstack-nova07:41
*** mhorban__ has joined #openstack-nova07:42
*** sahid has joined #openstack-nova07:44
*** markvoelker has quit IRC07:46
*** angdraug has joined #openstack-nova07:46
*** jmccrory has quit IRC07:49
*** scheuran has joined #openstack-nova07:49
*** abhishekk has quit IRC07:49
*** ankit_ag has quit IRC07:50
*** abhishekk has joined #openstack-nova07:50
*** ankit_ag has joined #openstack-nova07:50
openstackgerritlyanchih proposed openstack/nova: NUMATopologyFilter raise exception and not continue filter next node
*** browne has quit IRC07:52
*** rotbeard has joined #openstack-nova07:53
*** jmccrory has joined #openstack-nova07:54
*** abhishekk has quit IRC07:54
*** belmoreira has joined #openstack-nova07:57
*** breitz has quit IRC07:58
*** breitz1 has joined #openstack-nova07:58
openstackgerrithuan proposed openstack/nova: XenAPI:Resolve Nova/Neutron race condition
*** liusheng has quit IRC07:59
*** liusheng has joined #openstack-nova08:00
*** haomaiwang has quit IRC08:01
*** haomaiwang has joined #openstack-nova08:01
*** meh_di has joined #openstack-nova08:02
meh_dihi all dears. in nova.compute.api _create_instance how determine the host that  run  instance(s)08:02
*** matrohon has joined #openstack-nova08:05
*** lpetrut has quit IRC08:07
*** lpetrut has joined #openstack-nova08:07
*** liverpooler has joined #openstack-nova08:07
ndipanov hi meh_di - it's actually not determined there08:08
ndipanovit's done by the scheduler service08:08
*** Sree has joined #openstack-nova08:09
ndipanovthis is where we call the conductor service
ndipanovwhich then calls the scheduler here:
*** rmart04 has joined #openstack-nova08:11
ndipanovthat should allow you to figure it out if you read the code meh_di :) hope it helps!08:11
*** eglynn has joined #openstack-nova08:12
*** mhorban__ has quit IRC08:12
*** stevemar_ has joined #openstack-nova08:12
*** mkoderer has quit IRC08:15
meh_dindipanov: ok, I trace the  _create_instance in compute.api and I reached to and follow  _create_db_for_new_instance , in this method mentioned  "This is called by the scheduler after a location for the        instance has been determined." but where the scheduler determine the location of instance ??08:16
*** stevemar_ has quit IRC08:16
*** mkoderer has joined #openstack-nova08:17
ndipanovmeh_di, so _create_db_entry_for_new_instance would just add the record to the Nova DB08:18
*** salv-orlando has joined #openstack-nova08:19
ndipanovthen it goes on to call the conductor service which in turn valls the scheduler service which executes this method:
openstackgerrithuan proposed openstack/nova: XenAPI:Resolve Nova/Neutron race condition
ndipanovmeh_di, than based on the return value of that - the conductor would call the chosen compute node to finish actually provisioning the instance08:20
*** alex_klimov has quit IRC08:21
bauzasgood morning Nova08:25
*** salv-orlando has quit IRC08:27
*** mhorban__ has joined #openstack-nova08:29
meh_dindipanov: could you tell me on which part of this process scheduler service intervene ?08:29
openstackgerritJianghua Wang proposed openstack/nova: xenapi: OVS agent updates the wrong port when using XenServer + Neutron
meh_dindipanov : my mean is in the create_instance08:30
ndipanovmeh_di, so the last thing create_instance does is call the conductor - that's the first link I sent you08:30
alexschmdansmith: Hi, are you around?08:31
ndipanov(if you trace the calls in compute/ you will see that that link is the last line to get executed)08:31
*** tyagiprince has quit IRC08:31
ndipanovalexschm, unlikely he's on the west coast meaning it's past midnight there08:32
*** alex_klimov has joined #openstack-nova08:32
alexschmndipanov: okay thank you!08:32
*** tyagiprince has joined #openstack-nova08:33
*** jichen has joined #openstack-nova08:33
*** mpavone has joined #openstack-nova08:33
meh_dindipanov: your mean is
*** mhorban__ has quit IRC08:35
*** sudipto has joined #openstack-nova08:36
*** mhorban__ has joined #openstack-nova08:36
ndipanovno no08:36
ndipanovthat's what the scheduler does08:36
ndipanovbut that does not get called immediately08:36
bauzasmeh_di: nova boot is an async action08:37
ndipanovfirst the nova API service calls the conductor service08:37
bauzasmeh_di: the API just returns you an instance UUID directly, and then the boot itself is async'd by calling the conductor like ndipanov explained08:38
ndipanovyeah that is a good point meh_di, what bauzas is saying08:38
bauzasmeh_di: we never do API long-running actions synchronisously08:39
ndipanovasync'd means - the API call returns, but it has posted a message that the conductor service will receive and then send relevant messages to the scheduler (to decide on the destination) and then to the destination compute host to do the work08:39
bauzasmeh_di: because you wouldn't want to wait for 10 or more secs the answer to your HTTP request, right?08:39
*** fawadkhaliq has quit IRC08:41
*** fawadkhaliq has joined #openstack-nova08:42
*** abhishekk has joined #openstack-nova08:45
meh_dindipanov and bauzas : that really helped me to understand how it works, thank you both08:45
*** lpetrut has quit IRC08:45
*** jlanoux has joined #openstack-nova08:45
*** ferest has joined #openstack-nova08:48
*** annegentle has joined #openstack-nova08:49
*** sacharya has joined #openstack-nova08:49
*** otter768 has joined #openstack-nova08:49
ndipanovmeh_di, you're welcome :)08:52
*** ferest has quit IRC08:53
*** sacharya has quit IRC08:53
*** annegentle has quit IRC08:53
*** otter768 has quit IRC08:54
*** fawadkhaliq has quit IRC08:57
sudiptondipanov, have a very small review out here: jenkins failure is not dependent on the patch, wanted to get your thoughts on the same, when you have some time...08:59
*** ihrachys has joined #openstack-nova09:00
ndipanovsudipto, cool - will look at it09:00
sudiptondipanov, thanks!09:00
*** haomaiwang has quit IRC09:01
alogajohnthetubaguy: could you please raise your -2 from these patch series?,n,z09:01
alogajohnthetubaguy: thank you in advance :)09:01
*** haomaiwang has joined #openstack-nova09:01
rajeshtndipanov: could you please review
*** rubasov has joined #openstack-nova09:04
*** tyagiprince has quit IRC09:04
ndipanovrajesht, interesting - will look at it09:04
rajeshtndipanov: thanks :)09:05
*** achanda has joined #openstack-nova09:05
*** rpodolyaka1 has quit IRC09:08
openstackgerritSudipta Biswas proposed openstack/nova: Remove version hardcoding for 'show' method
ankit_agndipanov: Could you please help me to understand your comment on how migration will cause issues in fetching bdms09:10
ndipanovankit_ag, ah yes09:10
ndipanovwell it's quite simple though it's not up to that patch - I should probably remove the -109:11
*** romainh has joined #openstack-nova09:11
ndipanovbasically if we have a migration in progress09:11
ndipanovwe need to account for disk usage on both hosts09:11
*** lucas-afk is now known as lucasagomes09:12
*** yassine__ has joined #openstack-nova09:12
*** danpb has joined #openstack-nova09:12
ndipanovbecause in case the migration fails we need to be able to revert to the old host, and also we want to make sure that any other boot requests that happen while the instance is being migrated know that the resources on the destination are (soon to be) gone09:13
ndipanovankit_ag, ^ does that make sense?09:13
*** stevemar_ has joined #openstack-nova09:13
ndipanovthe reason this is broken currently is that we only ask libvirt for instances it can see, which does not account for instances that are currently being migrated09:14
ndipanov(both live and cold)09:14
ankit_agndipanov: I am planning to add following comment09:15
ankit_agIt should be based on instances and migrations to report more accurate info in edge cases with migrations.09:15
ankit_agIf it helps to understand the issue with migration09:15
ankit_agndipanov: Should I also add this in note "we only ask libvirt for instances it can see, which does not account for instances that are currently being migrated"09:16
ndipanovankit_ag, that would be awesome - I will raise a bug for this and we can fix it later then09:17
ndipanovand get you patch merged asap09:17
*** stevemar_ has quit IRC09:17
ankit_agndipanov: Sure, Thank you !09:18
*** mhorban__ has quit IRC09:20
*** mhorban__ has joined #openstack-nova09:20
*** Piet has quit IRC09:21
*** achanda has quit IRC09:22
johnthetubaguyaloga: I can't remove the -2 until the blueprint is approved for mitaka09:23
alogajohnthetubaguy: arg09:23
alogajohnthetubaguy: I missed that09:23
alogajohnthetubaguy: thanks anyway09:23
johnthetubaguyaloga: so I can help you get that approved09:23
* aloga sometimes is dazed with the processes09:24
rajeshtjohnthetubaguy: could you please give your feedback on
johnthetubaguyaloga: agreed, tried to demystify it here:
*** tdurakov_ has joined #openstack-nova09:24
johnthetubaguyaloga: I will put it on the etherpad of requests for spec-less blueprints here:
*** salv-orlando has joined #openstack-nova09:26
johnthetubaguyaloga: the next step is it gets raised in tomorrows nova meeting, if no objects, it is usually approved by the end of Friday09:26
alogajohnthetubaguy: I was going trough the doc09:26
alogajohnthetubaguy: thanks, is it needed that I am at the meeting?09:27
johnthetubaguyaloga: cools, let me know any questions, I can add claifications09:27
johnthetubaguyaloga: no, you don't have to, I think your blueprint describes its-self well, so they shouldn't be too much of a problem09:27
johnthetubaguyaloga: PS, I love your blueprints, its been on my TODO list for about a year, but never got around to it, so great to see that happening09:28
*** rk4n has joined #openstack-nova09:28
alogajohnthetubaguy: thanks :-)09:29
alogajohnthetubaguy: I leave them somehow abandoned from time to time, as some other urget matters are raised here :(09:30
johnthetubaguyaloga: understood09:30
alogajohnthetubaguy: but I try to not leave them die09:30
*** rk4n has quit IRC09:31
openstackgerritPawel Koniszewski proposed openstack/nova-specs: Pause VM during live migration
johnthetubaguyaloga: appreciate you pushing on them, if you get them all passing jenkins and rebased now, it should be all good when I remove the -209:31
*** e0ne has joined #openstack-nova09:32
alogajohnthetubaguy: I'm on it :)09:32
*** jinxing has quit IRC09:32
johnthetubaguyaloga: thank you!09:33
alogajohnthetubaguy: you're welcme, thank you as well09:33
*** jinxing has joined #openstack-nova09:34
*** jinxing has quit IRC09:35
*** mhorban__ has quit IRC09:35
*** salv-orlando has quit IRC09:36
*** salv-orlando has joined #openstack-nova09:37
*** xek has joined #openstack-nova09:37
*** mpavone has quit IRC09:37
*** tdurakov_ has quit IRC09:37
*** Sree__ has joined #openstack-nova09:38
*** tdurakov_ has joined #openstack-nova09:38
*** pixelbeat has joined #openstack-nova09:38
*** rdopiera has joined #openstack-nova09:38
*** mhorban__ has joined #openstack-nova09:39
*** tdurakov_ has quit IRC09:39
*** Sree has quit IRC09:41
*** rpodolyaka1 has joined #openstack-nova09:41
*** tdurakov_ has joined #openstack-nova09:42
*** markvoelker has joined #openstack-nova09:42
*** jinxing has joined #openstack-nova09:42
*** moshele has joined #openstack-nova09:42
*** jistr has joined #openstack-nova09:42
*** ZZelle has quit IRC09:44
*** ZZelle has joined #openstack-nova09:44
*** markvoelker has quit IRC09:46
*** rk4n has joined #openstack-nova09:49
*** ljxiash has quit IRC09:50
*** lpetrut has joined #openstack-nova09:52
*** tyagiprince has joined #openstack-nova09:57
*** ociuhandu has quit IRC10:00
*** haomaiwang has quit IRC10:01
*** haomaiwang has joined #openstack-nova10:01
*** aix has joined #openstack-nova10:08
*** rk4n has quit IRC10:09
*** Sree__ has quit IRC10:10
*** Sree has joined #openstack-nova10:10
*** alex_klimov has quit IRC10:11
*** rk4n has joined #openstack-nova10:15
*** eglynn has quit IRC10:15
*** jinxing has quit IRC10:16
*** prometheanfire has joined #openstack-nova10:19
*** rk4n has quit IRC10:19
prometheanfireso, can't boot images in liberty because of glanceclient-1.0 defaults to the --os-image-api-version 210:21
prometheanfireany workaround?10:21
*** Sree has quit IRC10:21
*** Sree has joined #openstack-nova10:22
*** rk4n has joined #openstack-nova10:22
sudipto prometheanfire nova is still using glance 1.0 - can you explain how you are trying this?10:24
*** jerrygb has joined #openstack-nova10:26
*** Sree has quit IRC10:26
prometheanfirewell, when doing a nova boot I get the 'AttributeError: id' backtrace, so it seems to be calling with verions 210:26
*** tdurakov_ has quit IRC10:27
*** zhangjn has quit IRC10:28
*** daemontool has joined #openstack-nova10:28
*** jwcroppe has joined #openstack-nova10:30
*** daemontool has quit IRC10:30
*** daemontool has joined #openstack-nova10:30
*** jerrygb has quit IRC10:31
*** daemontool has quit IRC10:31
*** ZZelle has quit IRC10:31
*** ZZelle has joined #openstack-nova10:32
moshelendipanov: hi10:33
*** jwcroppe has quit IRC10:34
mdboothHow do I type an i with dots on?10:34
*** fawadkhaliq has joined #openstack-nova10:36
danpbmdbooth: ï10:38
*** ihrachys has quit IRC10:38
mdboothdanpb: How did you get that?10:38
danpbcut+paste ;-P10:38
mdboothI got one by cut/pasting it from a website :)10:38
danpbmdbooth: Ctrl+Alt+u00ef+Enter    should do it10:40
danpbmdbooth: doh, i mean  Ctrl+Shift+u00ef+Enter10:42
mdboothI feel like I'm too stupid to operate a keyboard10:43
*** e0ne has quit IRC10:43
danpbmdbooth: or if you still have a arsdigita kinesis.... :-)10:44
mdboothI never got on with those things.10:44
*** naggappan has joined #openstack-nova10:45
*** prometheanfire has left #openstack-nova10:48
*** annegentle has joined #openstack-nova10:49
*** prometheanfire has joined #openstack-nova10:49
prometheanfirelooks like it was fixed in glanceclient 1.1.010:50
*** otter768 has joined #openstack-nova10:50
ndipanovmoshele, hello10:50
*** deepakcs has quit IRC10:50
naggappanHI i am trying to extract official ubuntu image for openstack using command "sudo kpartx -va trusty-server-cloudimg-amd64-disk1.img "  but it doesn't list me any /dev10:50
*** zhangjn has joined #openstack-nova10:51
*** rpodolyaka1 has quit IRC10:52
*** rpodolyaka1 has joined #openstack-nova10:53
mhorban_Hi guys, there is really old review Could you please look at that one please?10:54
*** sacharya has joined #openstack-nova10:54
*** annegentle has quit IRC10:54
*** otter768 has quit IRC10:54
moshelendipanov: regarding you comment in when VF =0 I see that dev_type is type-PCI and not type-PF ?10:56
mhorban@johnthetubaguy Could you please review trivial fix ?10:56
*** prometheanfire has left #openstack-nova10:57
moshelendipanov: what should be the value in such case ?10:58
*** sacharya has quit IRC10:58
ndipanovmoshele, hmmm that is very interesting11:00
moshelejohnthetubaguy: can you +W this commit  again I had to rebase it11:00
sdaguemoshele: +A11:00
*** haomaiwang has quit IRC11:01
moshelesdague: thanks11:01
ndipanovmoshele, let me think about it a bit...11:01
sdaguedanpb: you good with cburgess_'s ebtables patch -  ?11:01
*** haomaiwang has joined #openstack-nova11:01
moshelendipanov: sure :)11:02
ndipanovmoshele, let me try and see what happens with intel cards in that case11:02
*** aysyd has joined #openstack-nova11:03
*** ptm_away is now known as PaulMurray11:03
moshelendipanov: ok anyway we can add in to skip type-PCI as well11:04
*** gszasz has joined #openstack-nova11:04
*** e0ne has joined #openstack-nova11:04
*** jeblair has quit IRC11:04
*** atuvenie has joined #openstack-nova11:05
*** ildikov has quit IRC11:05
*** exploreshaifali has joined #openstack-nova11:05
ndipanovmoshele, well not sure that's desireable as there might be genuine cases when  you want to pass it thorough11:05
*** jeblair has joined #openstack-nova11:05
ndipanovmoshele, anyway - will comment soon11:05
*** moshele has quit IRC11:06
*** dims has joined #openstack-nova11:07
*** moshele has joined #openstack-nova11:07
PaulMurrayalex_xu, looking at your comment on about distinction between 'failed' and 'error'11:08
*** alex_klimov has joined #openstack-nova11:08
PaulMurrayalex_xu, is that used consistently at all - makes sense but I didn't know11:08
*** moshele has quit IRC11:09
ndipanovhey PaulMurray - have you had a chance to look at
*** rk4n has quit IRC11:10
*** moshele has joined #openstack-nova11:10
*** zhangjn has quit IRC11:11
PaulMurrayndipanov, actually I haven't - I will take a look11:11
*** dims has quit IRC11:12
ndipanovI was thinking that we could add the whole series to the subteam reviews list, but would be good to get more reviews frrom the subteam folks first11:12
PaulMurrayndipanov, did you see alex_xu's comment about error vs failed - is that a distinction that is used in general as far as you know?11:12
ndipanovPaulMurray, I have that patch marked to look at this morning but still on something else11:13
*** EinstCrazy has quit IRC11:13
ndipanovwill read in a bit and comment11:13
PaulMurrayndipanov, ok - won't distract you11:13
*** naggappan has quit IRC11:13
*** naggappan has joined #openstack-nova11:13
ndipanovfwiw - tdurakov looked at a bunch of the patches in that branch already11:14
PaulMurrayndipanov, for your patch - that's part of the improve resource tracking work item right?11:14
*** stevemar_ has joined #openstack-nova11:14
PaulMurrayndipanov, as listed here:
*** daemontool has joined #openstack-nova11:17
ndipanovPaulMurray, it is yes11:17
ndipanovPaulMurray, should we list it there?11:17
*** stevemar_ has quit IRC11:18
PaulMurrayndipanov, yes we should11:18
*** klkumar has quit IRC11:18
*** dims has joined #openstack-nova11:18
*** daemontool has quit IRC11:18
*** klkumar has joined #openstack-nova11:18
PaulMurrayndipanov, the list has "for subteam" and a "ready for cores" sections, so put them in as you see fit11:19
*** daemontool has joined #openstack-nova11:19
*** romainh has quit IRC11:19
ndipanovPaulMurray, will do - thanks!!11:19
*** tdurakov_ has joined #openstack-nova11:20
*** rk4n has joined #openstack-nova11:20
openstackgerritMadhu Mohan proposed openstack/nova: Corrected repeated text in doc
ndipanovPaulMurray, erm - where is that exactly ?11:21
*** daemontool has quit IRC11:21
*** tdurakov_ has quit IRC11:21
lxsliPaulMurray: live migration subteam should list reviews on but I don't see a section11:22
lxslioh wait, it is there, just not under subteams heading11:22
openstackgerritMadhu Mohan proposed openstack/nova: Corrected repeated text and another minor correction
*** daemontool has joined #openstack-nova11:24
*** tdurakov_ has joined #openstack-nova11:24
*** Okihita has joined #openstack-nova11:24
*** Okihita has left #openstack-nova11:25
*** zhenguo has quit IRC11:27
*** klkumar has quit IRC11:29
*** klkumar has joined #openstack-nova11:30
johnthetubaguyndipanov: have you looked at the instance actions stuff at all? does feel like we need to somehow merge migration objects and instance actions at some point11:31
*** aix has quit IRC11:31
*** aswadr has joined #openstack-nova11:32
*** tdurakov_ has quit IRC11:32
*** aix has joined #openstack-nova11:32
*** mwagner has quit IRC11:34
*** tdurakov_ has joined #openstack-nova11:35
*** rfolco has joined #openstack-nova11:35
openstackgerritZhihai Song proposed openstack/python-novaclient: Disable the client to retrive instance by ID in the DB
*** smatzek has joined #openstack-nova11:39
*** mhorban__ has quit IRC11:39
garyk1johnthetubaguy: mikal: can you guys please look at - it has been reveiwed by the subteam11:39
*** ihrachys has joined #openstack-nova11:40
ndipanovjohnthetubaguy, well yes I thought the same some time ago11:40
ndipanovbut tbh not sure what we get by that other than "purity"11:40
ndipanovas long as we use the migration record properly we should be good11:41
*** mhorban__ has joined #openstack-nova11:41
*** subscope has quit IRC11:41
ndipanovthat is to say - I am not convinced that the major refactor is worth the gain11:41
ndipanovsame thing with request_spec, vs instance, vs ...11:42
ndipanovand which bits go where..11:42
*** subscope has joined #openstack-nova11:42
johnthetubaguyndipanov: it might just be some API "sugar" we can add, honestly11:42
ndipanovas long as we keep the data and use it consistently and in a non-racy manner11:42
johnthetubaguyndipanov: I am thinking about us ensuring you can look up any action request-id in a single place, at some point11:42
ndipanovit's not critical whether we keep the flavor on the instance or request_spec etc11:42
*** markvoelker has joined #openstack-nova11:43
johnthetubaguyits tempting to add an instance action record in the DB, thats updating on the migration state changes, but just reading both lists in the API might be a smarter move.11:44
PaulMurraylxsli, thanks for posting the link - ndipanov sorry - I turned away for a while11:44
*** claudiub has joined #openstack-nova11:44
johnthetubaguygaryk1: cools, thanks for the ping, looks good now11:44
garyk1johnthetubaguy: thanks11:46
*** daemontool has quit IRC11:47
PaulMurraylxsli, ndipanov it should go in the live migration section under Priority Features - not subteams section11:47
*** markvoelker has quit IRC11:47
*** daemontool has joined #openstack-nova11:47
*** rdopiera has quit IRC11:48
*** exploreshaifali has quit IRC11:49
*** mhorban__ has quit IRC11:51
*** haomaiwang has quit IRC11:51
*** mhorban__ has joined #openstack-nova11:51
openstackgerritSean Dague proposed openstack/nova: Add more 'actions' for server concepts doc
lxsliPaulMurray: yep figured that out, maybe "subteams" section should be "Other features" but w/e11:52
*** rook has quit IRC11:52
*** exploreshaifali has joined #openstack-nova11:53
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: Enable options for oslo.reports
ndipanovjohnthetubaguy, so that's one of the reasons I'd like for us to have a programming API (objects? not sure it's the right place)11:53
*** ildikov has joined #openstack-nova11:54
bauzasjohnthetubaguy: so, I'm about republishing a new Tasks spec11:55
*** jhesketh has quit IRC11:55
*** sudipto has quit IRC11:55
bauzasjohnthetubaguy: and I'll also upload a new manifesto11:56
bauzasin case people wonder about instance actions11:56
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: Enable options for oslo.reports
*** romainh has joined #openstack-nova11:57
johnthetubaguyndipanov: we do have the compute/, but I guess thats not what you mean for that?11:58
johnthetubaguybauzas: OK, sounds promising, I am thinking we should upload a very small spec around making instance actions consistent everywhere? is that the spec?11:58
bauzasjohnthetubaguy: yup, agreed11:59
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: Enable options for oslo.reports
ndipanovjohnthetubaguy, no I mean a place where we insulate how we evolve this stuff from the business logic11:59
bauzasjohnthetubaguy: since the manifesto is there for saying which specs we should provide, I'm +111:59
bauzasthat way, we can also have one blueprint for every change related to that12:00
*** jinxing has joined #openstack-nova12:00
bauzasand providing to the low-hanging-fruit etherpad12:00
*** Sree has joined #openstack-nova12:00
ndipanovjohnthetubaguy, currently we either add a decorator (for example actions/faults) and miss a ton of cases, or we have to do a messy data migration which has it's consequences12:00
*** jhesketh has joined #openstack-nova12:00
ndipanovjohnthetubaguy, not sure if you get what I mean :)12:01
johnthetubaguyndipanov: oh sorry, I get what you mean now12:01
ndipanovso to me it feels that having something like12:02
johnthetubaguyndipanov: yeah, I would be happy with a consistent pattern that makes sense, its very messy right now12:02
*** doug-fish has joined #openstack-nova12:02
openstackgerritClaudiu Belu proposed openstack/nova: Hyper-V: adds os-win library
openstackgerritClaudiu Belu proposed openstack/nova: Hyper-V: removes *Utils modules and unit tests
ndipanovor even relocation = Relocation.intialize_for(instance)12:02
ndipanovor something12:02
* bauzas goes for an appointment12:02
*** exploreshaifali has quit IRC12:03
ndipanovwhere there is an API as opposed to just listing stuff that needs to happen in rpc callbacks12:03
johnthetubaguyyeah, I quite like the attached to instance idea12:03
johnthetubaguyfor instance actions and the migration stuff12:03
johnthetubaguy(and instance faults, I guess)12:03
ndipanovjohnthetubaguy, I worry that it will make the instance object that now does also sever additional things unweildy12:03
ndipanovjust because it's the central model of the business logic12:04
ndipanovit "looks" like the natural place to put things12:04
johnthetubaguyyeah, agreed, its tempting and concerning12:04
ndipanovbut we might end up with a "manager" class :)12:04
*** rk4n has quit IRC12:04
ndipanovthough that is just a guess12:04
ndipanovit may be fine12:05
johnthetubaguyyeah, feels like a "suck it and see kind" of thing12:05
ndipanovjohn5223, so that's what I tried to do with "apply_migration_context" stuff12:05
ndipanovbut that might be too low level12:05
johnthetubaguywe have those always lazyloaded things already, it feels like one of those12:05
*** EinstCrazy has joined #openstack-nova12:06
openstackgerritClaudiu Belu proposed openstack/nova: Converting nova.virt.hyperv to py3
*** atuvenie has quit IRC12:07
*** zhangjn has joined #openstack-nova12:09
*** zhangjn has quit IRC12:09
*** zhangjn has joined #openstack-nova12:10
*** zhangjn has quit IRC12:10
*** zhangjn has joined #openstack-nova12:10
*** rpodolyaka1 has quit IRC12:12
*** tojuvone has quit IRC12:13
*** tdurakov_ has quit IRC12:14
*** aswadr has quit IRC12:15
*** tdurakov_ has joined #openstack-nova12:16
*** rk4n has joined #openstack-nova12:19
openstackgerritClaudiu Belu proposed openstack/nova: Compute: Adds driver disk_gb instance overhead estimation
openstackgerritClaudiu Belu proposed openstack/nova: Hyper-V: Fixes disk overhead claim issue
*** ociuhandu has joined #openstack-nova12:19
*** shaohe_feng has quit IRC12:20
*** rk4n has quit IRC12:20
*** subscope has quit IRC12:20
*** wolsen has quit IRC12:20
*** subscope has joined #openstack-nova12:21
*** subscope has quit IRC12:21
ndipanovso what you are seeing is to be expected actually12:22
openstackgerritClaudiu Belu proposed openstack/nova: Replaces longs with ints
*** ociuhandu_ has joined #openstack-nova12:22
ndipanovthe problem is the "PF address whitelists all of VFs" is just a bad decision that opens us up to edge cases and ambiguities like this12:23
*** ociuhandu has quit IRC12:23
*** ociuhandu_ is now known as ociuhandu12:23
*** alejandrito has joined #openstack-nova12:23
*** aarefiev has quit IRC12:25
openstackgerritClaudiu Belu proposed openstack/nova: Fixes dict keys and items references for Python 3
openstackgerritAnkit Agrawal proposed openstack/nova: Pass bdm info to _get_instance_disk_info method
openstackgerritClaudiu Belu proposed openstack/nova: Replaces izip_longest with six.moves.zip_longest
*** aarefiev has joined #openstack-nova12:27
alex_xuPaulMurray: at least from this patch, _cleanup_incomplete_migrations clean up all the error migration, and set them to failed.12:27
openstackgerritPavel Kholkin proposed openstack/nova: fixed warn logging in sqlalchemy/
*** jkraj has joined #openstack-nova12:28
*** thorst has joined #openstack-nova12:28
openstackgerritClaudiu Belu proposed openstack/nova: Fixes Python 3 compatibility for filter results
*** CORTEX\krfa has joined #openstack-nova12:30
CORTEX\krfaHello. I've configured cinder to use our NFS NetApp storage solution for the shared storage. How should I proceed to use the NFS shared storage for the compute node using nova? I didn't find any exemple about it.12:32
openstackgerritbharaththiruveedula proposed openstack/nova-specs: Returning Compute Node IP address in hypervisor list
*** rook has joined #openstack-nova12:33
*** ihrachys_ has joined #openstack-nova12:34
*** ihrachys has quit IRC12:35
*** tojuvone has joined #openstack-nova12:37
*** lucasagomes is now known as lucas-hungry12:37
openstackgerritMadhu Mohan proposed openstack/nova: Corrected few grammatical nitpics
*** fawadkhaliq has quit IRC12:42
*** abhishekk has quit IRC12:43
*** markvoelker has joined #openstack-nova12:43
alex_xusdague: looks like there still can have some improvement for, or you want to merge first version, continue improve later?12:43
*** ziggypup has joined #openstack-nova12:43
*** ziggypup has left #openstack-nova12:44
sdaguehonestly, I'd rather merge and add more patches later12:44
sdaguebecause a merged document goes straight to the web, so is better for folks immediately12:44
alex_xusdague: ok, cool, let me +w12:45
sdagueI think with docs we're never going to be perfect, so the important criteria is "does this make things better", I find we can go around and around on docs patches and not land anything some times12:46
openstackgerritPawel Koniszewski proposed openstack/nova: Live migration failure in API leaves VM in MIGRATING state
*** romainh has quit IRC12:47
*** markvoelker has quit IRC12:48
alex_xusdague: ok, got it12:48
*** annegentle has joined #openstack-nova12:50
*** otter768 has joined #openstack-nova12:51
*** atuvenie has joined #openstack-nova12:52
alex_xusdague: I'm thinking the different between concept doc and api ref, for example, the concept already explain what is evacuate, then the api ref is more focus what effect based on each parameter?12:52
*** naggappan has quit IRC12:53
*** atuvenie_ has joined #openstack-nova12:53
*** klkumar has quit IRC12:53
*** otter768 has quit IRC12:55
*** annegentle has quit IRC12:55
PaulMurrayalex_xu, I also noticed migration.status being set to failed a lot when I was going through patches ndipanov is doing at the moment12:56
*** atuvenie has quit IRC12:57
*** tdurakov_ has quit IRC12:57
openstackgerritOleg Bondarev proposed openstack/nova: Live migration: wait for vif-plugged event on pre live migration
*** e0ne has quit IRC12:57
*** mpavone has joined #openstack-nova12:57
jichenalex_xu: I think api ref should tell user what's the input and what's the expect output... that's why I think  is fine , we can't have too much detail on the api-ref12:58
*** tdurakov_ has joined #openstack-nova12:58
*** moshele has quit IRC12:58
alex_xuPaulMurray: cool, will try to take a look at those patches also. It will be good we have consistent on those status, also a good chance doc them12:59
*** rotbeard has quit IRC13:00
*** moshele has joined #openstack-nova13:02
sdaguealex_xu: yeh, I honestly also think the concept doc should explain why you'd use a particular action13:03
sdaguebecause snapshot vs. backup13:03
sdagueor why anyone would want to use pause13:03
sdaguebut, honestly, almost all the time more explanation is better than less13:04
*** obondarev_afk is now known as obondarev13:04
alex_xusdague: yea13:04
*** rk4n has joined #openstack-nova13:04
*** jyuso1 has quit IRC13:04
*** angdraug has quit IRC13:05
*** jerrygb has joined #openstack-nova13:06
johnthetubaguysdague: alex_xu: +1 those comments about the concept guide.13:06
alex_xusdague: I'm think we should have some patch to show what the api ref is we expected, then can guide people to improve it13:06
alex_xuwe already have example for concept doc13:06
sdagueyeh, probably good point13:06
johnthetubaguyalex_xu: +1 for getting good worked example folks can copy, I would do the API complete ref and concept doc, linking to each other maybe?13:07
sdagueok, I have to drop for a bit and will be intermitent over the next couple of hours13:07
alex_xujohnthetubaguy: yea, I think we can linking each other13:07
alex_xusdague: no problem13:07
sdaguejohnthetubaguy: - completes our parity with libvirt on ebtables. Would like to make sure that lands13:07
PaulMurrayalex_xu, at the summit we talked about better doc for live-migration - I see that in that concept guide patch13:08
PaulMurraybut we also talked about giving an example of where and how it might be used13:08
PaulMurraye.g. rolling through a site to implement kernel patch upgrades that require reboot13:08
PaulMurraydoes that fit in there somehow?13:09
alex_xuPaulMurray: yea, I think that is good in concept doc13:09
*** markvoelker has joined #openstack-nova13:09
johnthetubaguyPaulMurray: totally should be in the concept guide I think, explains the why you use it that way, and the patterns that work well.13:09
* johnthetubaguy needs to go for some food13:09
PaulMurrayalex_xu, I can add something like that in another patch if you like13:09
alex_xuPaulMurray: cool, appreciate the doc patch :)13:10
PaulMurrayIt was in the back of my mind alread13:10
*** jwcroppe has joined #openstack-nova13:10
*** jerrygb has quit IRC13:10
*** e0ne has joined #openstack-nova13:10
*** klkumar has joined #openstack-nova13:13
*** vladikr has joined #openstack-nova13:14
*** jwcroppe has quit IRC13:15
*** stevemar_ has joined #openstack-nova13:15
*** jwcroppe has joined #openstack-nova13:15
*** jwcroppe_ has joined #openstack-nova13:17
*** stevemar_ has quit IRC13:18
*** ihrachys_ has quit IRC13:18
*** atuvenie_ has quit IRC13:18
*** ihrachys has joined #openstack-nova13:19
*** sacharya has joined #openstack-nova13:19
*** jwcroppe has quit IRC13:20
*** rotbeard has joined #openstack-nova13:21
*** angdraug has joined #openstack-nova13:21
*** sacharya has quit IRC13:24
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE
*** rpodolyaka1 has joined #openstack-nova13:26
*** irenab_ has joined #openstack-nova13:26
*** annegentle has joined #openstack-nova13:26
*** irenab has quit IRC13:27
*** irenab_ is now known as irenab13:27
ndipanovalex_xu, PaulMurray - I am -1 on this whole approach tbh
PaulMurrayndipanov, ack13:32
*** alex_klimov has quit IRC13:33
*** alex_klimov has joined #openstack-nova13:33
PaulMurrayndipanov, what is the security issue?13:33
*** rubasov has quit IRC13:34
PaulMurraysome kind of DOS?13:34
ndipanovPaulMurray, well check the commit message and the linked bug13:34
*** atuvenie_ has joined #openstack-nova13:34
*** atuvenie has joined #openstack-nova13:34
ndipanovallegedly since all migrations are admin actions13:35
*** kbyrne has quit IRC13:36
openstackgerritClaudiu Belu proposed openstack/nova: Transforms requests and responses into bytes
*** ndipanov has quit IRC13:38
*** ndipanov has joined #openstack-nova13:39
*** atuvenie has quit IRC13:39
*** atuvenie has joined #openstack-nova13:40
*** meh_di has left #openstack-nova13:42
*** mwagner has joined #openstack-nova13:42
alex_xundipanov: miss the rollback set status to error also. really mess on status error and failed13:43
*** kbyrne has joined #openstack-nova13:43
*** achanda has joined #openstack-nova13:44
ndipanovalex_xu, consider that before the migration object was introduced - there was no way to tell what was happening13:44
ndipanovunless you trace the code and logs13:44
alex_xundipanov: yea, indeed :)13:45
alex_xundipanov: is it worth there is propose to clear the usage of failed and error?13:45
*** krtaylor has joined #openstack-nova13:45
*** achanda has quit IRC13:46
*** lucas-hungry is now known as lucasagomes13:48
ndipanovalex_xu, well sure I'd like to see it cleaned up but on their own, not tied to that half-assed security fix13:49
*** achanda has joined #openstack-nova13:49
ndipanov<airquote> security fix </airquote>13:49
alex_xundipanov: yea13:50
*** tpeoples has joined #openstack-nova13:50
ndipanovalex_xu, so that's why I am -1 on that change really (as I tried to explain on the commit message comment)13:50
*** rubasov has joined #openstack-nova13:51
alex_xundipanov: yea, agree13:51
*** sudipto has joined #openstack-nova13:53
openstackgerritLudovic Beliveau proposed openstack/nova: Support configuration of multiple pci_alias within an array
*** achanda has quit IRC13:53
*** eharney has joined #openstack-nova13:53
openstackgerritAlexis Lee proposed openstack/nova: Move Process and Mentoring pages to devref
openstackgerritAlexis Lee proposed openstack/nova: Move Process and Mentoring pages to devref
*** gjayavelu has joined #openstack-nova13:57
*** doug-fish has quit IRC13:58
*** carolynvs has joined #openstack-nova13:59
*** rook has quit IRC13:59
*** tdurakov_ has quit IRC13:59
*** rlrossit has joined #openstack-nova14:00
*** carolynvs has quit IRC14:00
*** tpeoples has quit IRC14:00
*** tpeoples_away is now known as tpeoples14:01
*** changbl has quit IRC14:01
*** rook has joined #openstack-nova14:01
*** burgerk has joined #openstack-nova14:02
*** akshai has joined #openstack-nova14:04
*** exploreshaifali has joined #openstack-nova14:04
*** mdrabe has joined #openstack-nova14:05
*** lykinsbd has joined #openstack-nova14:05
*** ijuwang has joined #openstack-nova14:06
*** LukeHinds has joined #openstack-nova14:06
*** tdurakov_ has joined #openstack-nova14:06
*** bapalm has joined #openstack-nova14:08
*** boris-42 has quit IRC14:08
*** vlaza has joined #openstack-nova14:09
*** ctrath has joined #openstack-nova14:09
*** burgerk has quit IRC14:11
*** angdraug has quit IRC14:12
*** ankit_ag has quit IRC14:12
*** achanda has joined #openstack-nova14:12
garyk1sdague: you want me to update the commit message here?
sdaguegaryk: yeh14:15
sdaguethen we can flush it back through14:15
sdagueit just needed to get pulled because ever nova change was exploding14:15
*** dane-fichter has joined #openstack-nova14:15
openstackgerritgaryk proposed openstack/nova: Use flavor instead of flavour
garyk1sdague: ^14:16
openstackgerritTimofey Durakov proposed openstack/nova: Multinode job for live-migration
*** mgoddard__ has joined #openstack-nova14:18
garyk1sdague: gracias14:19
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo FlexibleBooleanField
*** romainh has joined #openstack-nova14:20
tdurakov_sdague, johnthetubaguy hi, could  you assist with merging this one:
*** mgoddard_ has quit IRC14:20
*** doug-fish has joined #openstack-nova14:21
*** jichen has quit IRC14:22
*** jaypipes has joined #openstack-nova14:23
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo VersionPredicateField
*** daemontool_ has joined #openstack-nova14:25
*** daemontool_ has quit IRC14:26
*** lsmola has quit IRC14:26
*** daemontool_ has joined #openstack-nova14:27
*** Sree has quit IRC14:27
*** daemontool has quit IRC14:27
*** Sree has joined #openstack-nova14:27
*** angdraug has joined #openstack-nova14:27
openstackgerritRyan Rossiter proposed openstack/nova: Use o.vo DictOfListOfStringsField
*** daemontool_ has quit IRC14:28
*** mgoddard__ has quit IRC14:28
*** shaohe_feng has joined #openstack-nova14:29
*** mgoddard has joined #openstack-nova14:29
*** daemontool_ has joined #openstack-nova14:29
*** lsmola has joined #openstack-nova14:29
*** dustins has joined #openstack-nova14:30
*** lsmola has quit IRC14:31
*** smatzek has quit IRC14:31
*** openstackgerrit has quit IRC14:31
*** Sree has quit IRC14:32
*** openstackgerrit has joined #openstack-nova14:32
*** aspiers has quit IRC14:34
*** tojuvone has quit IRC14:35
*** jkraj has quit IRC14:37
tdurakov_jaypipes, morning14:38
lxslianyone to look at please?14:39
*** bapalm has quit IRC14:43
*** angdraug has quit IRC14:43
*** angdraug has joined #openstack-nova14:43
*** bapalm has joined #openstack-nova14:43
*** burt has joined #openstack-nova14:44
*** fawadkhaliq has joined #openstack-nova14:44
*** mriedem has joined #openstack-nova14:45
lxslijohnthetubaguy: FYI
*** raildo is now known as raildo-afk14:46
*** cfriesen__ has joined #openstack-nova14:48
tdurakov_bauzas, alex_xu, folks, are you around?)14:49
bauzastdurakov_: yep14:49
bauzastdurakov_: wassup ?14:50
*** vlaza has left #openstack-nova14:50
tdurakov_bauzas, - could you review this?14:50
bauzastdurakov_: I'm honestly not that seasoned14:51
bauzastdurakov_: so I could +1, but I'm a bit torn with providing a +W if I agree14:51
*** daemontool_ has quit IRC14:51
tdurakov_bauzas, heh, it's actually hook for job that is in experimental pipeline, this one blocks me from implementing nfs14:52
*** otter768 has joined #openstack-nova14:52
tdurakov_bauzas, trust me:)14:52
*** garyk1 has quit IRC14:52
mriedemtdurakov_: come on14:52
bauzassdague: ^14:52
bauzastdurakov_: that's not that I don't trust you14:53
bauzastdurakov_: but you see, I don't have a context for that14:53
mriedemi don't trust anyone that says trust me14:53
mriedemtdurakov_: more details in the commit message would be helpful, or as a reply in the comments14:53
sdagueyeh, probably another check experimental14:54
sdagueso we can be confident it's working14:54
jaypipestdurakov_: Утро добрым не бывает14:54
tdurakov_sdague, it won't work14:54
tdurakov_there is no changes in code, so experimental will ignore this14:55
mriedemso i guess it's just a hook called from the gate-tempest-dsvm-multinode-live-migration job in project-config14:55
*** pedroalvarez has left #openstack-nova14:55
bauzastdurakov_: so, see, like mriedem explained, if you want to have cores reviewing, you should maybe explain more what's the problem and what can be fixed14:55
mriedemand the hook is here
tdurakov_bauzas, mriedem added comments14:56
*** smatzek has joined #openstack-nova14:56
bauzastdurakov_: okay, I need to ramp up, see14:56
*** otter768 has quit IRC14:56
mriedemand i see it's an experimental queue job here
bauzastdurakov_: that's why I think I could +1 but not +W14:57
tdurakov_jaypipes, heh, why so?14:57
mriedemi also see that we're running this on stable/kilo and stable/liberty
mriedemwhich won't work14:57
bauzasje ne comprend pas le russe :p14:57
*** dustins has quit IRC14:57
bauzasmriedem: oh that would break the experimental pipeline then ?14:58
mriedembauzas: experimental isn't voting14:58
*** ctrath has quit IRC14:58
bauzasmriedem: that I know14:58
mriedemso that doesn't really matter, but it's a waste of resources to run a job we know won't work14:58
bauzasmriedem: that's my point14:58
mriedemso i can update that and filter it from kilo/liberty14:58
bauzasmriedem: even if your dog doesn't vote, will you accept that some bad guy hurts your dog ?14:59
mriedemsince it uses nova/tests/live_migration/hooks/ which isn't in kilo and liberty14:59
mriedemi would, personally, but i don't like our dog14:59
*** moshele has quit IRC14:59
mriedemsdague: so what does this buy us, having this hook in tree? since it's just a pass through to running tempest15:00
mriedemall of the config still has to happen elsewhere15:00
*** tyagiprince has quit IRC15:00
*** dustins has joined #openstack-nova15:00
mriedemi guess so we don't have another DEVSTACK_GATE_TEMPEST_LIVE_MIGRATION flag in d-g15:00
mriedems/another/yet a new/15:00
tdurakov_mriedem, well, it fair enough to drop it from kilo and liberty until it will be back ported, but what about my hook?15:00
mriedemtdurakov_: i don't know what it's fixing, your commit message doesn't explain that15:01
mriedemi'm assuming it's just fixing the way the args are passed15:01
mriedemand that you're actually running tempest (hence the cd)15:01
tdurakov_not only, ok will update commit message15:01
*** daemontool has joined #openstack-nova15:01
mriedembut you can't just be like, here is this change i have, new core people, please +W it15:01
mriedem"trust me"15:02
openstackgerritTimofey Durakov proposed openstack/nova: Multinode job for live-migration
*** angdraug has quit IRC15:03
tdurakov_mriedem, updated commit message15:04
jaypipestdurakov_: because it's way too early for me :)15:04
openstackgerritAlexis Lee proposed openstack/nova: Remove SoftDeleteMixin from BandwidthUsage
*** signed8bit has joined #openstack-nova15:04
tdurakov_mriedem, it wasn't about hey, new cores, please merge my patch, btw it was a joke15:05
johnthetubaguylxsli: thanks for getting that review up there, will try look at that15:05
dansmithtdurakov_: not funny :/15:05
*** klkumar has quit IRC15:05
lxslijohnthetubaguy: cheers :)15:06
tdurakov_dansmith, sorry then, there was a smile. don't expect such reaction:(15:06
*** subscope has joined #openstack-nova15:08
*** thangp has joined #openstack-nova15:08
openstackgerritPaul Murray proposed openstack/nova: docs: update the concept guide shelve actions
lxsliCores: I have a bunch of patches (in merge conflict) which convert mox tests to mock. If I fix them, might you look at them, or should I just abandon?15:09
openstackgerritMatt Riedemann proposed openstack/nova: Multinode job for live-migration
mriedemsdague: you can +W
bauzastdurakov_: tbh, you should identify who in the core team is more seasoned re: your change, and ping him directly15:10
bauzastdurakov_: because while I'm able to review that change, it will also take more time for me to ramp up and get context15:10
bauzasin particular since the commit message was pretty brief15:11
mriedemtdurakov_: this filters out the stable branches
bauzas^that I can +1 because I understand it15:12
dansmithbauzas: alaski: This release note is passing jenkins now:
*** sahid has quit IRC15:12
bauzasmriedem: maybe a Depend-On ?15:12
* mriedem feels left out15:12
mriedembauzas: not necessary15:12
mriedemthe hook script is already in nova and the job is already defined15:12
mriedemi.e. it's already busted15:12
tdurakov_well, this job already here, so no need for deps15:12
*** exploreshaifali has quit IRC15:12
bauzasoh snap15:13
bauzasokay, I see15:13
sdagueright, so lets get check experimental run15:13
bauzasmriedem: dansmith: so the reno CI job is still blocked because of stable in the weeds15:14
dansmithwait, what?15:14
tdurakov_sdague, check experimental will ignore my job15:14
sdaguetdurakov_: ?15:14
dansmithbauzas: I thought that landed, sorry15:14
tdurakov_sdague, no change in code15:14
lxsliok going to abandon those mox patches15:14
tdurakov_sdague, only in tests folder15:14
sdagueoh, because it's the same as in PS 16?15:14
dansmithbauzas: wanna give me the change-id and I'll make this depends-on?15:15
bauzasdansmith: is still pending because of stable a bit unstable15:15
bauzasdansmith: and is depend-on the above15:15
tdurakov_sdague, i've provided tests results explicitly15:15
openstackgerritAndrew Laski proposed openstack/nova: Add persistence to the RequestSpec object
bauzasdansmith: I don't feel it's necessary15:15
bauzasdansmith: so my take is15:15
sdaguetdurakov_: oh, we're going to need to fix that eventually15:15
bauzaseither we hold any +W for reno patches15:15
*** mrsoul has quit IRC15:15
bauzasuntil lands15:15
*** stevemar_ has joined #openstack-nova15:16
bauzaswe agree to test locally a reno run before +W'ing15:16
*** liverpooler has quit IRC15:16
tdurakov_sdague, should we? it's functional job, or it will prevent from breaking it?15:16
*** andrewbogott has quit IRC15:16
dansmithbauzas: yeah, I was just going to document the linkage so I don't forget to see that the other lands15:16
*** sneti has joined #openstack-nova15:16
sdaguetdurakov_: we need to run live migration tests when we change stuff here15:16
openstackgerritDan Smith proposed openstack/nova: Add release note for API DB migration requirements
bauzasdansmith: I'm fine with the latter, ie. testing on my local venv before +W'ing15:16
*** shaohe_feng has quit IRC15:16
sdaguewe don't want it getting skipped15:16
sdagueso can you figure out what we need in project config to make that happen15:17
*** sballe has quit IRC15:17
tdurakov_sdague, ok, will add this to project-config15:17
bauzasdansmith: the bonus point being that will skip a full Tempest pass for reno patches15:17
dansmithbauzas: well, I just added the link for my own records :)15:17
*** d34dh0r53 has quit IRC15:17
*** angdraug has joined #openstack-nova15:17
*** Nakato has quit IRC15:17
bauzasdansmith: that could save a bit of resources15:17
*** mc_nair has joined #openstack-nova15:17
*** d34dh0r53 has joined #openstack-nova15:17
bauzasdansmith: hence the +W being okay but should be very very with caution15:17
*** andrewbogott has joined #openstack-nova15:18
*** sballe has joined #openstack-nova15:18
bauzasdansmith: like accepting a reno-only patch wasting all CI resources is a bit bad IMHO15:18
*** ndipanov has quit IRC15:18
*** Nakato has joined #openstack-nova15:18
*** pumaranikar has joined #openstack-nova15:18
*** stevemar_ has quit IRC15:18
*** ndipanov has joined #openstack-nova15:19
*** markmc has quit IRC15:19
*** mrsoul has joined #openstack-nova15:19
*** markmc has joined #openstack-nova15:19
*** jinxing has quit IRC15:19
*** rubasov has quit IRC15:20
*** sudipto has quit IRC15:20
*** rpodolyaka1 has quit IRC15:20
*** tonytan4ever has joined #openstack-nova15:21
*** rushil has joined #openstack-nova15:22
*** rpodolyaka1 has joined #openstack-nova15:22
*** sahid has joined #openstack-nova15:23
*** sacharya has joined #openstack-nova15:23
*** andrewbogott has quit IRC15:23
*** andrewbogott has joined #openstack-nova15:23
lxslithanks for +2 jaypipes! Anyone else for please?15:25
jaypipesbauzas: all the python job failures in the base patch in your requestspec series.. are those due to a known issue?15:25
*** pratikmallya has joined #openstack-nova15:25
bauzasjaypipes: sec15:25
bauzasjaypipes: checking15:25
*** raildo-afk is now known as raildo15:25
bauzasjaypipes: the gate was pretty unhappy when I uploaded my series, so I would bet yes, but I'll doublecheck15:25
*** pratikma_ has joined #openstack-nova15:25
*** rushil has quit IRC15:26
*** mrkz has joined #openstack-nova15:26
bauzasjaypipes: oh yeah, the failing unittests, yup it's a known issue15:27
*** pradk has joined #openstack-nova15:28
*** sacharya has quit IRC15:28
bauzasjaypipes: but sec, I'm going to -2 because of the reno file15:28
*** ildikov has quit IRC15:28
*** yonglihe has joined #openstack-nova15:29
*** pratikmallya has quit IRC15:30
*** salv-orlando has quit IRC15:30
*** salv-orlando has joined #openstack-nova15:31
bauzasmriedem: so I litteraly had zero time to work further on
mriedembauzas: i know what hte issue is, it's in the -dev list15:34
mriedemwith options to resolve15:34
bauzasoh ack15:34
mriedemjaypipes wants to hard delete instance_actions when we archive instnaces15:34
mriedemthere might be other tables after we get past those though15:35
*** tojuvone has joined #openstack-nova15:35
*** edtubill has joined #openstack-nova15:35
mriedemi'm interested in getting alaski's thoughts on it15:35
alaskimriedem: I responded, though I didn't address hard vs soft delete15:36
mriedemyou neatly sidestepped that issue15:36
mriedemwhich is why i'm calling you out :)15:36
*** haomaiwang has joined #openstack-nova15:37
mriedem"However there is a use case  for holding on to the deleted action so that someone could query when or  by whom their instance was deleted"15:37
mriedemso we don't have a query API for this15:37
mriedemso if they really wanted those, they could query them out of shadow tables after archive15:37
mriedemwhich sucks, but...15:37
alaskithat was a bit of a tangent, I don't think it should affect anything here15:38
mriedemso are you leaning toward soft delete when we delete an instance?15:38
*** atuvenie has quit IRC15:38
alaskiwell, I'm not sure I see a reason for it really.  What's important is what we do when we hard delete or archive an instance15:38
bauzasmriedem: alaski: honestly, I tend to agree with jaypipes15:39
bauzasmriedem: alaski: because I wonder why we should keep the actions if the instance is deleted15:39
mriedembauzas: well the instance is soft deleted, and then archive puts it in shadow tables, all of which you can still query if needed,15:40
bauzasif that's for tracking purposes, notifications are far better IMHO15:40
alaskiwhen we soft delete an instance I'm not sure I see why it matters if we soft delete the instance-actions or not15:40
mriedemalthough it's just more pack rat stuff15:40
*** stevemar_ has joined #openstack-nova15:40
alaskithough I don't see a reason not to do it15:40
mriedemalaski: heh15:40
mriedempick a side!15:41
bauzasalaski: mriedem: so I'm okay with soft-deleting an instance, not soft-deleting the actions related to it15:41
mriedemteam edward or team jacob!15:41
alaskiI would soft-delete15:42
bauzasI'm actually even wondering why we're soft-deleting instances, but that's another question :p15:42
*** electrocucaracha has joined #openstack-nova15:42
bauzasif that's for tracking purpose15:42
alaskithe advantage I see for soft delete is being able to easily query for what would be cleaned up on an archive/purge15:43
bauzasalaski: if we soft-delete an instance but hard-delete all the actions associated to it, it's still the same stuff for getting what could be archived15:45
bauzasalaski: because when an instance is deleted, you cannot resurrect it, right?15:46
bauzasso getting all the actions you had for that specific instance when it was alive sounds a bit weird to me15:46
*** rmart04 has quit IRC15:46
alaskibauzas: what about soft-delete?  though I'm not sure how that's implemented15:47
bauzasalaski: that would dramatically increase the archive size, nope ?15:47
*** xyang1 has joined #openstack-nova15:48
mriedemalaski: "the advantage I see for soft delete is being able to easily query for what would be cleaned up on an archive/purge" - yeah, it makes archive/purge easier, *but* i'd also have to add some logic in there for handling non-soft deleted instance_actions that are tied to soft deleted instances15:49
mriedemfrom before the time that we start soft deleting instance_actions - if that's what we do15:49
*** mhorban__ has quit IRC15:49
*** mhorban__ has joined #openstack-nova15:50
alaskibauzas: okay, API soft delete doesn't set the deleted flag on an instance.  so yeah, you can't resurrect a db deleted instance.15:51
*** ccarmack has left #openstack-nova15:51
alaskibauzas: you're in favor of hard deleting actions when an instance is set to deleted in the db?15:51
bauzasalaski: wait, what?15:52
bauzasalaski: you said 'API soft-delete doesn't do deleted=instance_id in the DB ?15:52
alaskibauzas: not that I saw15:52
alaskiinstance.destroy is called in terminate_instance on compute, which soft-delete doesn't call15:53
bauzasalaski: so it keeps the virt files, but still puts the deleted flag in the DB right?15:53
bauzasalaski: so soft-delete is just a sync call to modify the state, right?15:54
*** annasort_ has quit IRC15:54
alaskimriedem: this is why I don't have a strong opinion on soft-deleting on instance delete.  I don't see a strong advantage either way.15:54
mriedemalaski: ok, want to reply with those details to the ML?15:56
mriedemi have a meeting then can get back into this15:56
*** mriedem is now known as mriedem_meeting15:56
mriedem_meetingthe purge stuff that ctrath is working on is going to have to deal with this also
*** jerrygb has joined #openstack-nova15:57
alaskimriedem_meeting: sure15:57
alaskibauzas: soft-delete seems to just set the vm_state to soft-delete and shut down the instance.  But it's not something we offer so I haven't spent much time looking at it before so I could be missing something15:57
*** daemontool has quit IRC15:58
*** dustins has quit IRC15:58
*** dustins has joined #openstack-nova15:58
*** daemontool has joined #openstack-nova15:59
bauzasalaski: oh fun
*** yamahata has joined #openstack-nova15:59
*** mgoddard has quit IRC15:59
*** moshele has joined #openstack-nova16:00
*** gjayavelu has quit IRC16:00
*** haomaiwang has quit IRC16:01
*** tdurakov_ has quit IRC16:01
*** tdurakov_ has joined #openstack-nova16:01
*** tdurakov_ has quit IRC16:01
*** haomaiwang has joined #openstack-nova16:01
*** pratikma_ is now known as pratikmallya16:02
*** alex_klimov has quit IRC16:02
*** mdrabe has quit IRC16:03
*** mhorban__ has quit IRC16:04
*** mjura has quit IRC16:05
*** paul-carlton1 has quit IRC16:06
*** hemnafk is now known as hemna16:06
*** rushil has joined #openstack-nova16:08
bauzasmmm, doesn't seem to match right?16:08
*** erlon has quit IRC16:08
*** READ10 has joined #openstack-nova16:09
*** rcernin has quit IRC16:09
mdboothbauzas: Are you aware of a doc describing the purpose of soft delete, btw? Don't worry if not. I've been wondering about this lately. Seems like it's mostly just a source of pain.16:09
bauzasmdbooth: like I understand, it's just a way to mark an instance deleted in the Nova DB without notifying the driver - but I'm just checking that16:10
mdboothAh... I'm getting confused about my soft deletes16:10
mdboothI was thinking of db records.16:11
mdboothI get the purpose of the other type.16:11
*** wuhg has quit IRC16:11
sdaguemdbooth: mostly it's because the database was very normalized16:12
sdagueand there are fk between object types16:12
sdagueso if you delete a flavor, for instance, how do you display the flavor info in an instance16:13
bauzasmdbooth: oh, sec, that's more complex than what I said16:13
sdaguewhich, granted is now fixed16:13
*** jaypipes has quit IRC16:13
*** sayali has quit IRC16:13
cfriesen__question for people...why does the os-migrateLive API have the caller specifying "block_migration"?  Shouldn't nova be able to figure that out on its own?16:13
*** sahid has quit IRC16:13
sdaguebut the entire object model has some implicit keys that aren't enforced16:13
mdbooth'some' :/16:14
mdboothsdague: Yeah, I guessed it would be something like that, thanks.16:14
ndipanovcfriesen__, it should and we should remove it imho16:14
*** shaohe_feng has joined #openstack-nova16:15
bauzasmdbooth: so that's a full nova delete16:16
dansmithbauzas: that's not the soft delete that alaski is talking about right?16:17
dansmithoh, it is16:17
dansmithnevermind me16:17
*** moshele has quit IRC16:17
bauzasdansmith: well, there is no soft-delete API method AFAICT16:17
bauzasdansmith: there is only a delete that can be a soft delete if the operator asks for a delayed delete16:17
mdboothdansmith: Glad it's not just me :)16:18
dansmithyou guys went from db purged to soft deleted16:18
dansmithand I didn't make the transition with you :)16:18
*** burgerk has joined #openstack-nova16:18
bauzaswell, I'm caring of soft-delete because of the restore action16:18
bauzasbecause it *is* possible to restore a soft-deleted instance16:18
bauzaswhich means that it should be idempotent in terms of instance actions16:19
bauzasthat leads me to :16:19
*** wolsen has joined #openstack-nova16:19
bauzasmriedem_meeting: alaski: nevermind my position, we need to soft-delete instance actions too16:19
*** edmondsw has quit IRC16:19
*** mpavone has quit IRC16:20
*** mdrabe has joined #openstack-nova16:20
*** fawadkhaliq has quit IRC16:20
*** dims_ has joined #openstack-nova16:22
*** flip214 is now known as flip21116:23
*** jwcroppe_ has quit IRC16:24
*** dims has quit IRC16:24
*** belmoreira has quit IRC16:24
*** rpodolyaka1 has quit IRC16:25
*** exploreshaifali has joined #openstack-nova16:25
*** alexschm has quit IRC16:25
*** sayali has joined #openstack-nova16:26
*** rpodolyaka1 has joined #openstack-nova16:26
openstackgerritPaul Murray proposed openstack/nova: docs: describe migration and other movement concepts
bauzasmriedem_meeting: alaski: oh snap, nevermind, do_soft_delete() never calls instance.destroy() which means that when it's called, there is no way to revert its state => now I understand your point, alaski16:26
mdboothHmm, that's a wart. Was just looking to see why a resize operation required converting the disk image to raw and back. Turns out it's because we run resize2fs directly on the image file. We could do that with libguestfs and save the double copy.16:26
alaskibauzas: yeah.  I'm in a meeting as well now so I'm not following here much16:27
bauzasalaski: np, I went thru code to understand the soft delete action16:28
*** mhorban__ has joined #openstack-nova16:28
*** edmondsw has joined #openstack-nova16:28
PaulMurrayjohnthetubaguy, alex_xu a couple of concept doc patches for shelve and migrations:
bauzasand at the end, that doesn't change my position of 'if an instance is API deleted, it's soft-deleted in DB but with no way to resurrect it, so we can drop the actions related to it'16:29
*** sneti has quit IRC16:29
PaulMurrayalaski, you might be interested in the second one ^^^16:29
mdboothDo we have a policy on libguestfs, btw? We use it elsewhere in Nova, but iirc it has a wrapper round it for if it's not available. Is libguestfs ever not available?16:29
*** jdurgin1 has joined #openstack-nova16:30
*** ssurana has joined #openstack-nova16:31
*** mgoddard has joined #openstack-nova16:31
danpbmdbooth: non-KVM hypervisors for example16:32
danpblxc, xen, vz16:32
*** annegentle has quit IRC16:33
*** sneti has joined #openstack-nova16:33
mdboothAll of those could use libguestfs, surely?16:33
mdboothThe VMware driver might have trouble, though...16:33
*** annegentle has joined #openstack-nova16:33
mdboothBut in that case, what would calling resize2fs on it do?16:33
mdboothThis is some weird code.16:34
ndipanovhey johnthetubaguy can you look at this
mdboothI mean, the VMware driver could still run libguestfs, but the compute node doesn't have access to the disks.16:34
ndipanovit's been stuck in the queue for very long for no reason :/16:34
*** sahid has joined #openstack-nova16:35
danpbmdbooth: if any of those tried to use libguestfs it'd at best be falling back to running qemu w/ tcg16:35
mdboothdanpb: Right. Is that a problem?16:35
danpbapart from the general slowness16:35
mdboothIt's not generally that awful, iirc, especially if you just doing a bunch of io.16:36
*** sahid has quit IRC16:36
mdboothLike resizing a disk partition.16:36
*** suro-patz has joined #openstack-nova16:37
mdboothIt's going to be less awful than converting to raw and back, though. Although, its weird that this code is calling qemu-img.16:38
openstackgerritBalazs Gibizer proposed openstack/nova: Add service status notification
openstackgerritBalazs Gibizer proposed openstack/nova: Add infra for versioned notifications
mdboothAs it's not in the libvirt driver16:38
*** Piet has joined #openstack-nova16:38
*** ijuwang has quit IRC16:38
*** kfarr has joined #openstack-nova16:39
*** Marga_ has quit IRC16:40
*** Marga_ has joined #openstack-nova16:42
*** sahid has joined #openstack-nova16:44
*** mgoddard_ has joined #openstack-nova16:45
*** sacharya has joined #openstack-nova16:45
*** ijuwang has joined #openstack-nova16:45
openstackgerritGábor Antal proposed openstack/nova: Replaced deprecated timeutils methods
*** rpodolyaka1 has quit IRC16:45
*** EinstCrazy has quit IRC16:46
*** jwcroppe has joined #openstack-nova16:46
*** mgoddard has quit IRC16:48
*** scheuran has quit IRC16:49
*** mhorban__ has quit IRC16:49
*** pumaranikar has quit IRC16:50
*** pumaranikar has joined #openstack-nova16:50
*** haomaiwang has quit IRC16:51
*** sudipto has joined #openstack-nova16:52
*** rpodolyaka1 has joined #openstack-nova16:52
*** otter768 has joined #openstack-nova16:53
*** jaypipes has joined #openstack-nova16:53
*** rpodolyaka1 has quit IRC16:53
*** irenab has quit IRC16:54
openstackgerritLudovic Beliveau proposed openstack/nova: Allow saving empty pci_device_pools in ComputeNode object
*** fawadkhaliq has joined #openstack-nova16:54
*** fawadkhaliq has quit IRC16:55
*** fawadkhaliq has joined #openstack-nova16:55
*** rpodolyaka1 has joined #openstack-nova16:56
*** otter768 has quit IRC16:57
*** belmoreira has joined #openstack-nova16:58
*** stevemar_ has quit IRC16:59
*** tongli has joined #openstack-nova17:00
*** tdurakov_ has joined #openstack-nova17:00
*** tdurakov_ has quit IRC17:00
*** e0ne has quit IRC17:01
*** penick has joined #openstack-nova17:02
*** Sree has joined #openstack-nova17:04
*** diana_clarke has quit IRC17:05
*** jlanoux has quit IRC17:06
*** unicell has joined #openstack-nova17:07
*** matrohon has quit IRC17:07
*** unicell1 has quit IRC17:07
*** moshele has joined #openstack-nova17:07
*** mriedem_meeting is now known as mriedem17:08
*** sahid has quit IRC17:10
mriedemcfriesen__: there is already a spec to remove block_migration from the internal live migration stuff17:10
*** sileht has quit IRC17:10
*** ctrath has joined #openstack-nova17:10
*** electrocucaracha has quit IRC17:11
*** sileht has joined #openstack-nova17:11
*** _diana_ has joined #openstack-nova17:11
*** jistr has quit IRC17:11
*** _diana_ is now known as diana_clarke17:11
*** suro-patz has quit IRC17:12
*** romainh has quit IRC17:15
*** ociuhandu has quit IRC17:17
*** daemontool has quit IRC17:20
*** browne has joined #openstack-nova17:20
mriedemgibi_: btw, someone in the ops list did ask about deprecating the nova-manage service commands
openstackgerritPavel Kholkin proposed openstack/nova: Add log warning hints in sqlalchemy/
*** exploreshaifali has quit IRC17:22
mriedembauzas: so did you make up your mind on soft deleting instance_actions?17:22
mriedembtw, by soft delete i'm not talking about the api17:22
bauzasmriedem: yup17:22
mriedemi'm talking about instance_actions.deleted != 017:22
bauzasmriedem: yeah, but I wanted to see what was the stuff with restore17:22
bauzasmriedem: hence me soliloquing in the channel17:23
*** dguitarbite has quit IRC17:23
bauzashah, bad translation17:23
mriedemi get it17:23
mriedemtalking to yourself?17:23
*** moshele has quit IRC17:23
*** nkrinner has quit IRC17:23
*** tyagiprince has joined #openstack-nova17:23
*** exploreshaifali has joined #openstack-nova17:23
mriedemi took high school english class...17:23
*** dguitarbite has joined #openstack-nova17:24
*** nkrinner has joined #openstack-nova17:24
melwittme too, I remember that, soliloquy17:24
bauzasmriedem: so the outcome is, there is no way to resurrect a soft-deleted instance in DB17:24
bauzasmriedem: which means that when it's deleted, it's deleted17:25
*** angdraug has quit IRC17:25
*** lpetrut has quit IRC17:25
bauzasmriedem: in that case, I don't see the point of getting the list of actions17:25
*** belmoreira has quit IRC17:25
bauzasbecause you deleted that instance, it's not like if it was in error state and you needed to investigate17:25
melwittbauzas mriedem: on that I was thinking the instance_actions could be some historical data on a deleted instance however IIRC if you actually try to show instance_actions for a deleted instance it will show you nothing, ironically17:25
melwittbecause it tries to do an instance lookup first in the api17:25
bauzasthe contract is "go f*ck, I'm gonna recreate a fresh new instance, I don't care about that one particularly"17:26
bauzasmelwitt: that's the usecase I don't understand17:26
mriedemyou care if it was your pet17:26
mriedemright, but you don't care about people that care about pets :)17:27
*** gszasz has quit IRC17:27
bauzasmriedem: if it's a pet, you have something like notifications and AODH ?17:27
alaskimelwitt: that could be useful.  I'm not sure if it's worth adding the semantics to the API to do that since there's talk of deprecating instance-actions with tasks17:27
*** Sree has quit IRC17:27
mriedembauzas: if you're using those, yeah17:27
melwittI'm thinking of things like, someone deleted an instance in a tenant and someone else in the tenant asks ops, who deleted that instance? instance_actions would have that, I *think*17:27
*** tyagiprince has quit IRC17:27
mriedemmelwitt: it owuld17:27
mriedemstores the action and user that performed the action17:28
bauzasmelwitt: again, that's a notification query I feel17:28
melwittyeah, I realize that notifications is supposed to cover all of the instance_actions usecases for forensic data. I am not sure if it's in a great state by now that we can sort of get rid of instance_actions use in that way?17:28
*** changbl has joined #openstack-nova17:29
bauzasmelwitt: well, that's the point17:29
mriedemalso, where are people storing notifications and for how long17:29
bauzasmriedem: that's certainly something we can explain to operatoes17:30
*** ildikov has joined #openstack-nova17:30
bauzasmriedem: on the other hand, archiving the full list of actions for a deleted instance increases dramatically the size of the archive, nope ?17:30
bauzasif we consider that *all* instance actions are covered - which is not the case atm17:31
*** stevemar_ has joined #openstack-nova17:31
*** tyagiprince has joined #openstack-nova17:32
bauzasbut I feel you're considering me opiniated while honestly I defer that point to operatoes17:32
mriedembauzas: yes it increases the size of the archive, but it's not a change in the actual DB since the rows are moved to shadow tables17:32
mriedemyeah, so i was going to bring this up in ops in a bit17:32
mriedemwas looking at stacktach first, which i've never used17:32
bauzasmriedem: well, the telemetry team delivered a new backend for storing timestamp-related info, called aodh17:33
*** unicell has quit IRC17:33
bauzasI heard some ops feedback that it was filling up the hole17:33
alaskimriedem: we rely on stacktach for storage, but I don't know our retention policy.  I don't think it goes beyond 90 days though.17:33
*** tonytan4ever has quit IRC17:33
bauzasnot aodh, gnocchi - snap17:34
bauzasmeh, the retention is a fair point17:34
bauzasthat said, when I was ops, I was usually doing forensics with logs not more than 90 days ago17:35
bauzasI mean, apache logs and other stuff17:35
mriedemi thought gnocchi was metrics17:36
mriedemhonestly, it'd be nice if these projects could have a name that remotely resembes wtf they do17:36
bauzasmriedem: like nova ? :D17:36
mriedemi know17:36
openstackgerritEd Leafe proposed openstack/nova: Config options: centralize section "scheduler"
openstackgerritEd Leafe proposed openstack/nova: config options: centralize section "serial_console"
bauzasmeh, neither gordc nor jd are around17:37
mriedemso you store an 'instance' resource in gnocchi17:38
*** mwagner has quit IRC17:38
*** yassine__ has quit IRC17:38
*** mgoddard has joined #openstack-nova17:38
*** mgoddard_ has quit IRC17:38
bauzasokay, I don't want to overthink about notifications17:39
dansmithalaski: question for you on
mriedemok, so i guess we punt and say we hard delete instance_actions during archive/purge of soft deleted instances, if you want instance_actions stored elsewhere, hook up stacktach or gnocchi or $custom_notification_listener_storage_thingy17:39
bauzasif operators think it's a good idea to keep the actions as longer as the instance in DB, then I'm fine17:39
PaulMurrayndipanov, I saw this spec on migration with PCI device and wondered if it overlaps with anything you have been doing?17:40
bauzasmriedem: do a x-post with -ops@ I guess17:40
*** tyagiprince1 has joined #openstack-nova17:40
mriedembauzas: yup, doing that now17:40
ndipanovPaulMurray, that came out of a patch discussion recently I think17:40
ndipanovPaulMurray, it's complementary17:40
*** tyagiprince has quit IRC17:41
*** tyagiprince1 is now known as tyagiprince17:41
*** armax has joined #openstack-nova17:41
* bauzas puts his parent hat on17:41
PaulMurrayndipanov, I was going to review it - is there anything related I should look at to understand how its related?17:42
* PaulMurray not sur eI make sense in the evening17:42
*** garyk1 has joined #openstack-nova17:45
*** garyk has quit IRC17:45
*** claudiub has quit IRC17:46
*** tdurakov_ has joined #openstack-nova17:47
*** exploreshaifali has quit IRC17:47
*** exploreshaifali has joined #openstack-nova17:48
mriedemwhat was that other M project that competed with ceilometer?17:48
dims_y that's the one17:48
*** electrocucaracha has joined #openstack-nova17:50
*** electrocucaracha has quit IRC17:50
mriedemlet's rename it17:50
*** aix has quit IRC17:51
*** pratikmallya has quit IRC17:52
*** shaohe_feng1 has joined #openstack-nova17:52
mriedemalaski: heh, we posted at the same time17:52
openstackgerritEd Leafe proposed openstack/nova: Config options: centralize section "scheduler"
openstackgerritEd Leafe proposed openstack/nova: config options: centralize section "serial_console"
*** electrocucaracha has joined #openstack-nova17:52
*** electrocucaracha has quit IRC17:53
*** shaohe_feng1 has quit IRC17:54
*** electrocucaracha has joined #openstack-nova17:54
*** shaohe_feng has quit IRC17:54
*** jerrygb_ has joined #openstack-nova17:54
*** stackdump has joined #openstack-nova17:54
*** jerrygb_ has quit IRC17:54
*** jerrygb_ has joined #openstack-nova17:55
*** apoorvad has joined #openstack-nova17:56
*** tyagiprince1 has joined #openstack-nova17:56
*** tdurakov_ has quit IRC17:57
*** jerrygb has quit IRC17:57
*** EinstCrazy has joined #openstack-nova17:58
*** tdurakov_ has joined #openstack-nova17:59
*** jerrygb_ has quit IRC17:59
*** tyagiprince has quit IRC17:59
*** tyagiprince1 is now known as tyagiprince17:59
*** achanda has quit IRC18:00
alaskidansmith: commenting on the db index, but testing showed that it does work.  I messed it up the first time and migration tests yelled at me and don't any longer18:00
alaskidansmith: on,cm I think I'm doing what you're saying on line 32718:01
*** jerrygb_ has joined #openstack-nova18:01
*** jerrygb_ has quit IRC18:01
*** jerrygb_ has joined #openstack-nova18:01
alaskimriedem: nice.  you at least cross-posted though18:02
*** achanda has joined #openstack-nova18:02
*** achanda has quit IRC18:03
raildoHey guys, I'm with a doubt here about default quotas... There is a API call for default quotas for tenant
dansmithalaski: hmm, okay18:04
raildoIs there a default quota for the service, and a default quota for every project?18:04
dansmithalaski: oh, right, so why are you iterating the fields again?18:04
*** belmoreira has joined #openstack-nova18:04
*** EinstCrazy has quit IRC18:05
*** daemontool has joined #openstack-nova18:05
*** belmoreira has quit IRC18:05
*** tyagiprince has quit IRC18:06
*** tyagiprince1 has joined #openstack-nova18:06
*** PaulMurray is now known as ptm_away18:07
*** tyagiprince1 is now known as tyagiprince18:08
*** READ10 has quit IRC18:10
*** tdurakov_ has quit IRC18:11
*** tdurakov_ has joined #openstack-nova18:11
*** gjayavelu has joined #openstack-nova18:12
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: imagecache: set base_dir as a property
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: imagecache: implement listing of cached images
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: virt: add method for obtaining the list of cached images
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: libvirt: implement get_cached_images method
*** Marga_ has quit IRC18:13
openstackgerritPaul Carlton proposed openstack/nova-specs: Cancel ongoing live migrations
alaskidansmith: because I used an intermediate object.  I see now, you're recommending not doing that18:13
dansmithyeah, just load-n-go (tm)18:14
alaskiheh.  just need to pull id and instance_uuid18:14
dansmiththey're on the object already no?18:14
dansmithI mean, id I guess18:14
dansmithalthough it will  be in the blob already if you've saved it more than once18:15
*** e0ne has joined #openstack-nova18:15
*** tonytan4ever has joined #openstack-nova18:15
alaskiyeah, that's right.  I should stop looking at this while distracted18:16
*** daemontool has quit IRC18:16
*** gjayavelu has quit IRC18:16
dansmithalaski: I thought his name was more conventional, like "luke" or something18:17
*** signed8b_ has joined #openstack-nova18:17
*** bkopilov has joined #openstack-nova18:17
*** suro-patz has joined #openstack-nova18:17
*** jerrygb_ has quit IRC18:17
bkopilovHi Experts,  need help18:17
alaskidansmith: lol18:18
*** pratikmallya has joined #openstack-nova18:18
*** danpb has quit IRC18:18
bkopilovi have test image deactivate / activate and boot instance , this test fails only on citrix xen only18:18
*** unicell has joined #openstack-nova18:18
*** signed8bit has quit IRC18:19
bkopilovcould you please check and let me know if its a bug -
*** njirap has joined #openstack-nova18:21
bkopilovhere is the bug
openstackLaunchpad bug 1517367 in OpenStack Compute (nova) "unable to boot instance after image reactivate with xenapi citrix" [Undecided,New]18:21
*** fawadkhaliq has quit IRC18:21
*** atuvenie has joined #openstack-nova18:21
*** tyagiprince has quit IRC18:22
*** tyagiprince has joined #openstack-nova18:23
*** ociuhandu has joined #openstack-nova18:24
*** ihrachys has quit IRC18:24
openstackgerritAugustina Ragwitz proposed openstack/nova: WIP: update functional tests to support optional project id
*** mwagner has joined #openstack-nova18:27
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE NFS setup for multinode job
openstackgerritAndrew Laski proposed openstack/nova: Add persistence to the RequestSpec object
*** vilobhmm has joined #openstack-nova18:28
*** rpodolyaka1 has quit IRC18:28
*** annegentle has quit IRC18:30
*** tyagiprince1 has joined #openstack-nova18:31
*** rpodolyaka1 has joined #openstack-nova18:31
*** rpodolyaka1 has quit IRC18:31
openstackgerritPaul Carlton proposed openstack/nova-specs: Cancel ongoing live migrations
tdurakov_sdague, ping, need for project-config tests enablement for live-migration job18:32
*** rpodolyaka1 has joined #openstack-nova18:32
tdurakov_*need advice18:32
*** tyagiprince has quit IRC18:33
*** tyagiprince1 is now known as tyagiprince18:33
*** jaypipes has quit IRC18:34
*** tonytan4ever has quit IRC18:37
andrewbogottbauzas: I’d appreciate another look at when you have the time.  (Also I could use guidance through the CI maze… currently can’t tell if my patch is buggy or the tests are buggy.)18:38
*** sudipto has quit IRC18:38
mriedemraildo: default quota per project18:38
mriedemraildo: not sure what you mean by service18:38
mriedemthe quotas are on projects/users and resources18:38
mriedemandrewbogott: recheck, that unit test thing was fixed late yesterday18:40
andrewbogottmriedem: ok!18:41
*** lucasagomes is now known as lucas-dinner18:41
*** thangp has quit IRC18:42
mriedemdansmith: sounds like cern is moving to kilo next week,18:44
*** signed8b_ is now known as signed8bit_ZZZzz18:44
mriedemdansmith: and don't have a ton of conductors either :)18:44
*** jerrygb has joined #openstack-nova18:44
mriedem16-48 conductor workers with a max of 200 computes18:46
*** obondarev has quit IRC18:46
*** signed8bit_ZZZzz is now known as signed8b_18:47
dansmithif they're not already concerned with cpu usage, then I expect they'll be okay18:47
openstackgerritCorey Wright proposed openstack/nova: Prevent redundant instance.update notifications
dansmithalthough they may want a few extra nodes for the initial bump18:47
raildomriedem: when I create a project, nova will update the project quota using the default quota, right?18:48
dansmithsounds like it was about 4% normal overhead to have that code enabled for the makeup of the other one, so really not that big18:48
*** lpetrut has joined #openstack-nova18:48
mriedemraildo: create a project in keystone?18:48
raildomriedem: yes, and later nova put the quota information about this project in the quota table18:49
raildomaybe, I'm creating a confusion here...18:49
mriedemdefault quotas are created on first access yes18:50
*** erlon has joined #openstack-nova18:50
*** pratikma_ has joined #openstack-nova18:50
mriedemthe update API actually does the create
mriedemand falls back to update if create fails due to the quota record already existing18:51
mriedemsomewhere in that mess the default quota class is used18:51
raildomriedem: ok, got it... thanks!18:51
melwittI thought default quotas were in absence of a set quota. like if you haven't yet set quotas for a project, the defaults are returned18:52
*** thangp has joined #openstack-nova18:52
*** jwcroppe_ has joined #openstack-nova18:52
*** pratikmallya has quit IRC18:52
mriedemmelwitt: yeah i think that's right18:52
melwittI didn't think it actually sets the defaults18:52
melwittyour project will not have any quota entry until you set quotas18:53
mriedemthis gets down in there
mriedemget_settable_quotas gets the default quotas18:53
mriedemif you don't pass anything in, it just uses the defaults18:53
raildomelwitt: but if didn't update the quota before create a intance, for example, the default quota will be used as the quota limit for this project18:53
*** otter768 has joined #openstack-nova18:53
melwittraildo: right18:54
*** lykinsbd_ has joined #openstack-nova18:54
*** lykinsbd_ has quit IRC18:54
melwittI'm just saying I don't think your project will have a quota record in the table unless you call the quota update api first18:54
*** rpodolyaka1 has quit IRC18:54
mriedemor the nova-manage quota update command18:54
*** salv-orl_ has joined #openstack-nova18:54
mriedemor whatever it is18:54
*** salv-orl_ has quit IRC18:54
mriedemwhich we shoujd probably deprecate...18:54
raildomelwitt: right18:54
*** tyagiprince1 has joined #openstack-nova18:54
*** lykinsbd_ has joined #openstack-nova18:54
melwittcool, sorry if I needlessly said stuff :)18:55
mriedemno it's good,18:55
*** tonytan4ever has joined #openstack-nova18:55
mriedembecause like i said, somewhere in the mess,18:55
raildocan we have more than one "default quota" in nova?18:55
*** jwcroppe has quit IRC18:55
mriedemthat is a good question18:56
raildolike default quota flavors? :P18:56
mriedemi think there is only one default quota class18:56
*** tyagiprince has quit IRC18:57
*** tyagiprince1 is now known as tyagiprince18:57
*** lykinsbd has quit IRC18:57
mriedemquota class is global18:57
*** salv-orlando has quit IRC18:57
mriedemand there is only one, 'default'18:57
mriedemat least upstream18:57
mriedemrax might have some custom quota class mumbo jumbo out of tree18:57
melwittyeah, that's what I'm seeing in the db api18:58
raildoright :)18:58
mriedemeverything else for quotas is per project18:58
*** otter768 has quit IRC18:58
*** lykinsbd has joined #openstack-nova18:59
mriedemso if you do nova quota-class-update --instance 20 default18:59
mriedemthat changes the default quota for instances to 20 globally for any new project that's created18:59
*** rushil has quit IRC18:59
*** obondarev has joined #openstack-nova19:00
*** lykinsbd_ has quit IRC19:00
*** doug-fis_ has joined #openstack-nova19:00
raildomriedem: makes sense19:00
*** suro-patz has quit IRC19:00
*** rotbeard has quit IRC19:01
*** rook has quit IRC19:01
*** ctrath has quit IRC19:02
alaskimriedem: we don't use quota classes.  we have a hack for quotas per flavor class that's done via a different method19:02
*** suro-patz has joined #openstack-nova19:02
*** suro-patz has quit IRC19:03
*** suro-patz has joined #openstack-nova19:03
*** signed8bit has joined #openstack-nova19:03
*** doug-fi__ has joined #openstack-nova19:03
*** signed8b_ has quit IRC19:04
*** doug-fish has quit IRC19:04
*** doug-fi__ has quit IRC19:04
*** rk4n has quit IRC19:04
*** tyagiprince has quit IRC19:04
*** Guest12059 is now known as med_19:04
*** med_ has quit IRC19:04
*** med_ has joined #openstack-nova19:04
*** tyagiprince1 has joined #openstack-nova19:04
*** doug-fish has joined #openstack-nova19:04
*** doug-fish has quit IRC19:04
*** doug-fis_ has quit IRC19:05
*** doug-fish has joined #openstack-nova19:05
*** harlowja has quit IRC19:05
mriedemi wonder how much cruft we could remove from nova and the quotas flow if we can assume there is only one quota class and it's name is 'default'19:05
*** jwcroppe_ is now known as jwcroppe19:06
*** rlrossit has quit IRC19:06
*** doug-fish has quit IRC19:06
*** doug-fish has joined #openstack-nova19:07
*** tyagiprince1 is now known as tyagiprince19:07
*** moshele has joined #openstack-nova19:07
*** harlowja has joined #openstack-nova19:08
*** ctrath has joined #openstack-nova19:09
*** rlrossit has joined #openstack-nova19:09
*** moshele has quit IRC19:10
*** rushil has joined #openstack-nova19:15
*** tyagiprince1 has joined #openstack-nova19:17
*** chhavi has quit IRC19:19
*** Marga_ has joined #openstack-nova19:19
*** tyagiprince has quit IRC19:19
*** tyagiprince1 is now known as tyagiprince19:19
*** mriedem has quit IRC19:20
*** ndipanov has quit IRC19:20
*** penick has quit IRC19:20
*** penick has joined #openstack-nova19:20
*** doug-fish has quit IRC19:22
*** browne has quit IRC19:22
*** doug-fish has joined #openstack-nova19:23
rlrossitgibi_: do you mind if I pull down the NotificationBase patch and add unit tests to it?19:23
*** bapalm has quit IRC19:23
*** tdurakov_ has quit IRC19:24
*** tdurakov_ has joined #openstack-nova19:25
*** tdurakov_ has quit IRC19:25
*** bapalm has joined #openstack-nova19:26
cfriesen__what's the expected behaviour if we do a cold migration on a boot-from-volume instance with ephemeral storage backed by qcow2 using libvirt/kvm?  Would we properly copy the ephemeral storage over to the destination and not copy the cinder volume?19:28
*** tyagiprince has quit IRC19:29
cfriesen__oh, and the compute nodes are not on shared storage19:29
*** vishwanathj has quit IRC19:29
cburgess_sdague *ping*19:29
*** mriedem has joined #openstack-nova19:30
*** jichen has joined #openstack-nova19:31
sdaguecburgess_: pong19:33
cburgess_sdague Want to chat the workaround script19:34
cburgess_cfriesen__ What exactly do you mean by ephemeral storage in that BFV scenario?19:36
cburgess_sdague Or you know.. run away screaming from it is also a valid strategy.19:36
*** LukeHinds has quit IRC19:36
*** lykinsbd has quit IRC19:36
sdaguecburgess_: sure19:40
cfriesen__cburgess: I mean booting with "nova --ephemeral=X --block-device-mapping...."  where the rootfs is on a cinder volume, but there's a local scratch disk on the compute node.19:40
cburgess_cfriesen__ I didn't think we migrated any of the ephemeral disk. I thought those were always re-created on the destination. Maybe I'm wrong.19:41
cburgess_sdague You mentioned we should have a conversation in the channel. Do you want to do that now or table this for the nova meeting?19:41
cfriesen__cburgess: I thought ephemeral disappeared only when the instance was terminated, in which case it would make sense to migrate them.19:42
*** rushil has quit IRC19:42
*** dane-fichter has quit IRC19:43
*** rushil has joined #openstack-nova19:43
*** burt has quit IRC19:44
cfriesen__cburgess: a superficial scan of the libvirt code make it look like we'll migrate the ephemeral/swap disks.  I should probably actually try it though.19:46
*** Marga_ has quit IRC19:47
*** alex_klimov has joined #openstack-nova19:48
*** lykinsbd has joined #openstack-nova19:49
*** onder has quit IRC19:50
*** lykinsbd has quit IRC19:53
*** onder has joined #openstack-nova19:53
sdaguecburgess_: so, honestly, because I've got a few patches in flight that I'm trying to land today, lets postpone that till tomorrow if that's ok. mordred pointed out there is a dpkg trick we could use so it would survive package upgrade, which we should probably mention to people19:53
sdaguethe biggest issue is that you don't need to do all the stop / start things you put in the commit message19:53
openstackgerritAugustina Ragwitz proposed openstack/nova: WIP: update functional tests to support optional project id
*** nic has joined #openstack-nova19:55
*** rpodolyaka1 has joined #openstack-nova19:55
*** lpetrut has quit IRC19:57
*** nic1 has joined #openstack-nova19:57
cburgess_sdague OK fair enough.19:57
*** stackdump has quit IRC19:57
cburgess_sdague Lets discuss tomorrow, I'll ping mordred for the dpkg trick, doesn't help RHEL customers but whatever.19:57
sdagueI think the more important part is landing the first patch19:57
cburgess_sdague Right19:58
sdaguewell rhel 7 customers will have new enough libvirt19:58
sdagueso that's at least something19:58
*** suro-patz has quit IRC19:58
sdaguehonestly, I expect on low rates for spawns it won't be seen super often19:59
sdagueespecially if you don't use floating ips19:59
*** nic has quit IRC19:59
*** lpetrut has joined #openstack-nova20:00
*** suro-patz has joined #openstack-nova20:00
*** lykinsbd has joined #openstack-nova20:02
sdagueauggy: on, we also need a change to the scenarios -
sdagueto run a version of the tests without project_id20:04
auggyright yeah that's in the todo list20:04
auggyoh wait20:04
auggyso my plan was to try to get it working by forcing it to run without project id and get that working20:04
*** ZZelle_ has joined #openstack-nova20:04
*** atuvenie has quit IRC20:05
auggythen muck around with the scenarios20:05
*** subscope has quit IRC20:06
auggyi'll add a todo item to the patch for that though20:06
*** subscope has joined #openstack-nova20:06
auggyright now i'm just running to try to isolate what it's failing on and it looks like it's an issue with the template data not matching the response data20:09
sdagueok, let me take a look20:09
*** subscope has quit IRC20:09
auggyi think i need to create another template for the no project id case?20:09
*** subscope has joined #openstack-nova20:10
auggyor make the project id a variable20:10
*** salv-orlando has joined #openstack-nova20:11
auggysdague: so here's what i saw for the response vs template -
auggyi will pull up a link to the chunk of code20:11
auggythat's where it's starting the post command20:13
*** e0ne has quit IRC20:13
auggywhat i'm seeing is the links generated in the response don't have the project name "openstack" in them, but the template data expects them to have it20:14
*** Marga_ has joined #openstack-nova20:14
*** njirap has quit IRC20:16
sdagueoh, you mean the bookmarks?20:16
sdague  ?20:16
auggyyeah whatever those links are for20:16
*** subscope has quit IRC20:17
sdagueyeh, hmmm... I think we're going to need to make the infrastructure support inflating those20:17
auggyyeah i noticed some variables in the templating20:18
auggyeg %(uuid)s20:18
auggyso it would make sense if possible to do something like that for project id20:18
auggyif that's the way to go then that's what i was contemplating pursuing20:19
sdagueyeh, that seems sensible20:19
auggyi didn't -1 WIP my patch so the full functional tests would run on it20:19
auggyin case anything else pops up unrelated to this issue20:19
*** ccard has quit IRC20:19
sdaguethey'll run even on a WIP20:19
auggyoh they will! i did not know!20:20
auggyok i'll WIP it then20:20
auggyWIP it good20:20
* dansmith slow claps20:21
*** yamahata has quit IRC20:21
sdaguebut yeh, I think a way to make the project_id optional in those templates is going to be the right thing to do20:21
sdagueit will also make them a bit less magical with the openstack keyword20:22
dansmithalaski: so we decided to do all the sqlalchemy bits in the object for the API db going forward, right?20:22
*** lpetrut has quit IRC20:22
dansmithalaski: I think I was in favor of that, but just caught myself barfing over that as done in the flavor-api patch20:23
alaskidansmith: we did, and I think you were in favor as well20:23
alaskiI need to look at that patch again because I think they did it differently than in other patches20:24
openstackgerritjichenjc proposed openstack/nova: Change Invalid exception to a specified exception
dansmithalaski: it's going to make every piece of nova dependent on sqlalchemy, even though almost none of it will use it20:24
dansmithalaski: and thus, we'll have to install SA on compute nodes20:24
* auggy is trying to say sqlalchemy five times fast20:25
*** ccard has joined #openstack-nova20:25
alaskidansmith: hmm20:25
alaskiI'm not sure if that is actually a problem20:26
alaskibut I'm heading out for the day.  I'll be back to discuss though20:26
dansmithI'll comment20:29
*** penick has quit IRC20:34
*** harlowja has quit IRC20:36
*** tongli has quit IRC20:38
*** harlowja has joined #openstack-nova20:41
*** tonytan4ever has quit IRC20:45
*** Marga_ has quit IRC20:45
*** nic has joined #openstack-nova20:45
bauzasyay, reno change merged for liberty \o/20:47
*** ctrath has quit IRC20:47
*** thorst has quit IRC20:47
*** nic1 has quit IRC20:48
*** rk4n has joined #openstack-nova20:48
*** thorst has joined #openstack-nova20:48
bauzassince I'm not a stable core, could someone +W again,n,z20:49
bauzas(I would like to not recheck, just go the gate pipeline)20:49
bauzassdague: dansmith: ^20:49
*** thorst has quit IRC20:49
dansmithdid it20:49
bauzascool, ty20:49
*** thorst has joined #openstack-nova20:49
*** thorst has quit IRC20:49
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements
*** thorst has joined #openstack-nova20:50
*** ctrath has joined #openstack-nova20:51
openstackgerritOpenStack Proposal Bot proposed openstack/python-novaclient: Updated from global requirements
*** thorst_ has joined #openstack-nova20:53
*** rpodolyaka1 has quit IRC20:53
*** thorst has quit IRC20:54
*** FL1SK has quit IRC20:54
*** otter768 has joined #openstack-nova20:54
*** jogo has quit IRC20:55
*** jogo has joined #openstack-nova20:55
*** jogo has quit IRC20:55
*** jogo has joined #openstack-nova20:55
*** thorst_ has quit IRC20:57
*** EinstCrazy has joined #openstack-nova20:58
*** otter768 has quit IRC20:59
*** rk4n has quit IRC20:59
*** thangp has quit IRC21:01
*** openstack has joined #openstack-nova21:04
*** alejandrito has quit IRC21:04
*** e0ne has joined #openstack-nova21:09
*** pumaranikar has quit IRC21:09
*** pumaranikar has joined #openstack-nova21:10
*** penick has joined #openstack-nova21:10
*** rfolco has quit IRC21:10
*** jistr has joined #openstack-nova21:10
*** thorst has joined #openstack-nova21:11
openstackgerritmelanie witt proposed openstack/nova: xenapi: Add helper function and unit tests for client session
*** jerrygb_ has joined #openstack-nova21:12
openstackgerritPushkar Umaranikar proposed openstack/nova: Build timeouts task filters deleting instances
openstackgerritPushkar Umaranikar proposed openstack/nova: Modify VM's updated_at field on volume actions
mriedemour db archive unit tests are pretty hokey, they test archive on a table by table basis, or very limited set21:13
mriedemreally need a functional test for that...boot an instance, delete, then archive it21:13
* mriedem gets to work21:14
mriedemalso interesting, there are no-db functional tests in the nova.tests.functional.db21:15
mriedemclass InstanceMappingTestCase(test.NoDBTestCase):21:15
mriedemseems counter productive21:15
*** thorst has quit IRC21:15
*** penick has quit IRC21:15
*** jerrygb has quit IRC21:15
*** raildo is now known as raildo-afk21:15
*** rpodolyaka1 has joined #openstack-nova21:16
dansmithbauzas: sdague: do we need to recheck this?
bauzasdansmith: I was waiting for
*** signed8bit_ZZZzz is now known as signed8bit21:19
*** rpodolyaka1 has quit IRC21:19
*** changbl has joined #openstack-nova21:19
bauzasthe Depends-On is a bit wrong21:19
mriedemdansmith: yes on the recheck21:19
mriedemthe dependent changes are merged21:19
bauzasmriedem: sec21:19
*** penick has joined #openstack-nova21:19
openstackgerritEd Leafe proposed openstack/nova: Config options: centralize section "scheduler"
openstackgerritEd Leafe proposed openstack/nova: WIP: Add better help text to scheduler options
bauzasmriedem: dansmith:
mriedembauzas: you only need the first change in that series though right?21:20
mriedemfor the tox bits21:20
mriedemok, well, should be long for 244761 to land anyway21:21
*** jerrygb_ has quit IRC21:22
bauzasmriedem: well, I feel it should work, but I wonder if we could break the CI if not landing the unrelease patch first21:22
bauzaslemme think about that21:22
*** ijw has joined #openstack-nova21:22
bauzasno, it's okay21:22
*** ijw has quit IRC21:22
bauzasbecause that's only if we add some reno changes for liberty21:23
bauzasmriedem: ^21:23
*** ijw has joined #openstack-nova21:23
*** jerrygb_ has joined #openstack-nova21:23
bauzasso yeah, I'm okay +W'ing again
bauzasdansmith: mriedem: sdague: ^21:23
*** thorst has joined #openstack-nova21:24
sdaguebauzas: tickled the +A bit21:24
sdagueit's now in the gate21:24
*** breitz1 is now known as breitz21:25
bauzassdague: cool, thanks !21:25
*** thorst_ has joined #openstack-nova21:25
*** tonytan4ever has quit IRC21:25
*** smatzek has quit IRC21:26
*** jerrygb__ has joined #openstack-nova21:26
*** jerrygb__ has quit IRC21:27
*** tonytan4ever has joined #openstack-nova21:28
*** jerrygb has joined #openstack-nova21:28
*** jerrygb_ has quit IRC21:28
*** thorst has quit IRC21:28
mriedemsdague: looking at some of the in tree functional tests that create a server, they dont have addCleanup for deleting the server after the create, is that because we're just testing against a sqlite db that gets created for each test run?21:31
mriedemso we don't care about leaking resources21:31
sdaguemriedem: yep21:31
*** e0ne has quit IRC21:31
sdaguesqlite and fakevirt21:31
sdagueso everything is gone at the end of the test21:31
dansmithephemeral cloud21:31
sdagueit's the best ephemeral cloud21:31
sdaguewe even have multinode21:32
bauzasalaski: around ?21:32
mriedemas long as i have resources in the db when i run the archive i'm happy21:32
mriedemsdague: yeah i was going to basically copy except for the rename stuff21:33
bauzasalaski: nvm, will look into fixing
mriedemso create server, wait for active, delete it, wait for deleted, then archive it21:33
sdaguesure, though I don't know there is much wait for active21:34
mriedemcreating the server should give us at least some interesting foreign keys, like instance_actions21:34
mriedemit just waits for !BUILD21:34
sdagueright, sure21:34
*** Marga_ has joined #openstack-nova21:34
*** pratikma_ has quit IRC21:35
*** pratikmallya has joined #openstack-nova21:35
dansmithsdague: did you see my query during the cells meeting earlier/21:35
sdaguejust with fakevirt I don't think we even bother with a build step21:35
openstackgerritPushkar Umaranikar proposed openstack/nova: Modify VM's updated_at field on volume actions
sdaguedansmith: I guess I missed it21:39
*** rwmjones has quit IRC21:39
*** Marga_ has quit IRC21:39
*** Marga_ has joined #openstack-nova21:40
dansmithsdague: I was wondering if you'd be okay with merging that grenade cells api change before all this reno crap actually lands21:40
*** kfarr has quit IRC21:40
dansmithsdague: given it's approved and just waiting, and the PTL has acked21:41
*** jistr has quit IRC21:41
*** rwmjones has joined #openstack-nova21:41
sdagueok, how far away are we from getting reno to land?21:41
bauzasnot that far21:41
dansmithit's been soonish for a while21:42
bauzasbut there is still a chain of 2 patches to land before the grenade change can land21:42
bauzasie. merging the CI job + the reno patch itself21:43
sdaguedansmith: sure, pull the depends on21:43
dansmithsdague: okay21:43
openstackgerritDan Smith proposed openstack/nova: Add release note for API DB migration requirements
dansmiththat is not what I wanted21:43
bauzashah :)21:44
bauzasI can fast-verify that one^21:44
dansmithbauzas: yea, you wanted that change anyway, so.. :D21:44
sdaguebut if the reno doesn't land by M1 then you'll owe us all beer :)21:44
sdagueor maybe bauzas will21:44
*** dustins has quit IRC21:45
bauzasdansmith: I actually already verified that one because of the prelude section21:45
bauzasdansmith: so I'm +2 with it21:45
dansmithsdague: if it's the kind of beers that are oft-promised and ne'r-collected, then sure :)21:46
sdagueI'm collecting from someone :)21:46
dansmith"Currency Ale" I believe it is21:46
*** e0ne has joined #openstack-nova21:47
sdaguemriedem: anyway, if you build a test, build one like either test_server_groups.py21:47
sdagueor there is some stuff in wsgi/ dir there that also pokes directly into the db21:47
sdaguein case you want grey box testing21:48
*** rook has joined #openstack-nova21:48
mriedemsdague: this first rev is quick and dirty but re-uses parts of the other tests, like test_server_groups21:49
mriedemso minimal copy/paste21:49
mriedemjust want to see if i can make this hit the archive db api21:49
sdagueyeh, right, test_server_groups is a good starting point21:49
*** rlrossit has left #openstack-nova21:49
sdaguebecause it doesn't go through a huge indirection like the samples tests21:50
*** e0ne has quit IRC21:50
sdaguethis is how you hit the db api -
sdaguein a test like that21:51
*** yamahata has joined #openstack-nova21:51
sdaguebauzas: the release notes project-config bit merged21:52
mriedemsdague: yeah that's basically what i'm writing21:53
sdaguemriedem: cool21:53
sdagueand ++ on more tests like that21:53
bauzaslet's make it a success, nope?21:53
sdaguebauzas: go for it21:53
sdagueit was your patch :)21:53
bauzas#success reno in place for the nova release notes21:53
openstackstatusbauzas: Added success to Success page21:53
yongliheanyone running into this error:AttributeError: 'ConductorAPI' object has no attribute 'object_class_action_versions' ?21:54
bauzasokay, I can now check for all the merged UpgradeImpact patches21:54
dansmithyonglihe: where?21:54
yonglihedansmith: some time PCI CI will trigger this one, and seems not like a configration error, i'm not sure if it's a bug or not21:55
dansmithyonglihe: seems like a mismatch-of-versions sort of problem to me21:55
bauzasan old conductor21:55
yonglihetoday, one time hit, jenkins not  finish it's testing, i will check that to confirm , thanks dansmith21:56
dansmithyeah, could be old conductor, depending on where it's coming in21:56
*** salv-orl_ has joined #openstack-nova21:57
*** rwmjones has quit IRC21:57
yonglihedansmith:  n-cpu is die while boot up time21:57
*** salv-orlando has quit IRC21:57
dansmithyonglihe: too new of o.vo for the nova code you're running?21:58
yongliheit's pull the oslo based on requirements, let me check the pip freeze21:59
dansmithyonglihe: is this on master?22:00
*** jerrygb has quit IRC22:00
klindgrendansmith, I ran a cProfiler on conductor with a single worker.  Call graph is here:  f - unsure if this helps spans any additional ideas?22:00
yonglihedansmith: it's a new patch, but nothing related to that change. i'm trying to figure what's wrong22:00
*** penick has quit IRC22:01
*** pratikmallya has quit IRC22:01
dansmithklindgren: I dunno what the percentages are, so, not really22:02
*** changbl has quit IRC22:02
dansmithklindgren: you have lots of build/destroy traffic going on on your cloud right?22:03
klindgrenvery very very little22:03
dansmithoh I thought you said you did22:03
klindgrenmaybe 80 or so a day22:03
klindgrenactually I would say we might have ~500 creates a day22:03
klindgrensimilar number of deletes22:04
yongliherequired oslo.versionedobjects>=0.9.0, in the test it's oslo.versionedobjects==0.13.0, dansmith22:04
dansmithyonglihe: I'm not sure what you're saying22:04
dansmithyonglihe: 0.13 is the current release22:05
dansmithyonglihe: which should be fine for master, and clearly is because it works in our gate22:05
klindgrendansmith, the call graph was generated via: gprof2dot - do you have something that you would rather see used?22:05
*** gjayavelu has joined #openstack-nova22:05
dansmithklindgren: no, I just don't know what the numbers mean22:07
mriedemdansmith: klindgren: i might not be reading this correctly, but it looks like the deep stuff at the bottom is sqlalchemy ORM22:07
dansmithklindgren: but it looks to me like there are a bunch of timer things that are running a ton22:07
mriedemdoing a query, fetching results, iterating22:08
dansmitha lot more than anything else22:08
dansmithmriedem: yes, and those are called a ton as well22:08
mriedemare the timers/pollers rabbit heartbeat things?22:08
dansmithmriedem: I dunno22:09
yonglihedansmith, anyway, thanks , i'm trying to figure out if it's a mis-configuration or not.22:09
dansmithlooks to me like langhelpers.__get__ gets called like 40,000 times as a result of one service call22:09
mriedemoh that's sqlalchemy22:10
*** burt has joined #openstack-nova22:10
mriedemor mysql-python, no idea really22:10
*** doug-fish has quit IRC22:10
dansmithsqlalchemy I think22:10
dansmithklindgren: I assume you've done slow and large query analysis already?22:11
dansmithbecause if something is pulling a ton of items, it'll cause a lot of churn here22:11
klindgrenI haven't checked the slow_querry log in a while22:12
klindgrenbut last time I did anything that was logged was just over 1 second22:12
*** rwmjones has joined #openstack-nova22:12
dansmiththat's slow, right?22:12
klindgrenWe also pruned out database for  soft_deleted items about a month ago22:13
dansmithklindgren: are you running ssl talking to rabbit?22:15
dansmiththe first percentage in the box is total time22:15
dansmithssl:read is 65%22:15
dansmithdispatch and reply, which is where all the real work is done is 9%22:15
*** edtubill has quit IRC22:17
klindgrenwe are running ssl talking to rabbit22:17
dansmithcan you turn that off as a test?22:17
dansmithmaybe that's why others aren't seeing the same pressure22:17
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE NFS setup for multinode job
*** edtubill has joined #openstack-nova22:18
*** jichen has quit IRC22:18
*** claudiub has joined #openstack-nova22:19
yonglihedansmith: i'm going to open a bug tracking this problem, include figure out if it's a bug or  not.22:19
dansmithyonglihe: okay22:19
yonglihe2 times hit this today, it's worth to track.22:20
dansmithyonglihe: it should be pretty easy to look at whatever this is running on and see if conductor/rpcapi has that newer method22:20
dansmithyonglihe: are you guys properly rebasing every patch on top of master when you run it?22:20
dansmith(zuul does this I think, if you're running zuul)22:21
dansmithbut this would be easily hit if you're not and someone rechecked an old patch22:21
*** edtubill has quit IRC22:21
*** edtubill has joined #openstack-nova22:21
yonglihedansmith: yes we pull new nova, and pull the testing patch.22:22
yonglihedansmith: i got your point22:22
bauzasdoes someone know how to pdb a functional test?22:22
*** edtubill has quit IRC22:23
klindgrendansmith, soooo22:23
klindgrenCPU usage dropped like a mofo22:24
dansmithklindgren: ORLY22:24
klindgren7% -> 0.7%22:24
klindgrenwhen not processing a message22:24
yonglihedansmith: this is does happen if author not re-base his patch, got.  i will track  this in CI.22:24
dansmithyonglihe: sounds like you're not rebasing then22:24
*** penick has joined #openstack-nova22:24
*** tdurakov has joined #openstack-nova22:25
dhellmannbauzas : \o/22:25
yonglihedansmith: i remember not re-basing,  i better do that, thanks.22:25
klindgrenso yea - conductor is basically sleeping on cpu now22:27
mriedemklindgren: cool, time to update that etherpad :)22:27
mriedemedmondsw: ^ re: the ssl thing, i know you love you some ssl22:30
yonglihedansmith: tracking this in this link:
edmondswmriedem, reading back...22:30
*** rushil has quit IRC22:31
*** rwmjones has quit IRC22:31
claudiubmriedem: hello. If you have a teeny tiny amount of time available, could you take a look at the os-win in nova patch? Hyper-V CI has +1, subteam has +1, your comments have been addressed. Thanks!
openstackgerritMatt Riedemann proposed openstack/nova: WIP: Reverse sort tables before archiving
mriedemsdague: ^ i'm pretty proud of the test added there22:32
edmondswwe've been using SSL/TLS (really TLS) with rabbit for several releases now in the product I work on22:32
mriedemclaudiub: ok, probably tomorrow22:32
claudiubmriedem: cool. :)22:33
mriedemedmondsw: scaled out to 250 compute nodes though?22:33
mriedemand 3 control nodes22:33
edmondswno... we only have one control node, 30 hosts22:34
*** FL1SK has joined #openstack-nova22:35
*** hemna has quit IRC22:35
*** penick has quit IRC22:36
*** hemna has joined #openstack-nova22:36
*** pratikmallya has joined #openstack-nova22:37
*** lykinsbd has quit IRC22:37
*** sacharya has quit IRC22:38
*** browne has joined #openstack-nova22:39
*** tonytan4ever has quit IRC22:40
*** rwmjones has joined #openstack-nova22:41
*** browne has quit IRC22:41
edmondswmriedem, actually, I think we've gone up to 200 hosts with NovaLink (the 30 was KVM)22:44
dansmithoooh, what's novalink?22:44
mriedempowervm v222:44
*** jaypipes has joined #openstack-nova22:47
*** burgerk_ has joined #openstack-nova22:48
*** burgerk has quit IRC22:49
*** edmondsw has quit IRC22:49
mriedemawesome, so i now have a functional test showing why nova-manage db archive_deleted_rows is busted22:51
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements
mriedemforgot we had to flip a switch to tell sqlite to care about fkeys22:51
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE NFS setup for multinode job
openstackgerritMatt Riedemann proposed openstack/nova: WIP: Reverse sort tables before archiving
*** gjayavelu has quit IRC22:52
*** burgerk_ has quit IRC22:54
*** mriedem has quit IRC22:54
*** otter768 has joined #openstack-nova22:55
*** unicell has quit IRC22:55
*** otter768 has quit IRC23:01
*** mdrabe has quit IRC23:02
*** ijw has quit IRC23:03
openstackgerritTimofey Durakov proposed openstack/nova: DO NOT MERGE NFS setup for multinode job
*** ijuwang has quit IRC23:03
klindgrenso turns I got way ahead of myself23:05
*** jerrygb has joined #openstack-nova23:05
klindgren2015-11-18 16:02:13.374 16155 ERROR oslo_messaging._drivers.impl_rabbit [req-dfe0c216-9f36-4a57-970d-0475dc80a01c - - - - -] AMQP server closed the connection. Check login credentials: Socket closed23:05
*** Piet has quit IRC23:05
klindgrenit wasn't connecting to rabbitmq hence why hte cpu dropped so much23:05
*** penick has joined #openstack-nova23:06
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements
*** burt has quit IRC23:07
*** ijw has joined #openstack-nova23:07
*** jerrygb has quit IRC23:09
*** ctrath has quit IRC23:10
*** jerrygb has joined #openstack-nova23:14
*** pumaranikar has quit IRC23:16
*** ctrath has joined #openstack-nova23:16
*** sneti has quit IRC23:16
*** pradk has quit IRC23:19
*** signed8bit is now known as signed8bit_ZZZzz23:20
auggybauzas: did anyone ever answer your how to pdb functional tests question?23:21
bauzasauggy: nope, but I'm okay23:21
auggyidk if this is useful or relevant to what you were asking, but i just throw a unittest.main loop into the test and run it through the debugger23:21
bauzasauggy: I just used -m testtools.run23:22
bauzasin my functional venv23:22
auggyah cool, and you were able to interact with pdb through that?23:22
* auggy makes a note23:22
bauzaslike for unittests23:22
auggyyeah, i use a visual debugger in PyCharm for that23:22
auggybut it's good to know how to do it on the shell too, so thanks for that :)23:23
bauzasI don't like visual debuggers :)23:23
auggyi got hooked early23:23
bauzasand I also don't like IDEs :p23:24
bauzasbut it's not Friday yet...23:24
auggymy favorite is still gdb in emacs23:24
bauzaswell, it's post-midnight and I have to leave so...23:25
bauzasAtom + pdb !23:25
* auggy is still getting used to not thinking of pdb as Perl Debugger ;)23:26
*** alex_klimov has quit IRC23:26
*** xyang1 has quit IRC23:28
*** suro-patz has quit IRC23:30
*** aysyd has quit IRC23:30
*** rmart04 has joined #openstack-nova23:31
*** suro-patz has joined #openstack-nova23:32
*** signed8bit_ZZZzz is now known as signed8bit23:34
*** penick has quit IRC23:35
*** ZZelle_ has quit IRC23:35
*** tdurakov has quit IRC23:38
dansmithdims_: you're not in -operators?23:45
*** suro-patz has quit IRC23:46
*** signed8bit is now known as signed8bit_ZZZzz23:50
*** ctrath has quit IRC23:52
*** thorst_ has quit IRC23:55
*** rmart04 has quit IRC23:56
*** mc_nair has quit IRC23:56
*** stackdump has joined #openstack-nova23:57
*** suro-patz has joined #openstack-nova23:59

Generated by 2.14.0 by Marius Gedminas - find it at!