Tuesday, 2016-08-16

*** markvoelker has joined #openstack-nova00:00
*** shaohe_feng has quit IRC00:01
*** shaohe_feng has joined #openstack-nova00:02
mriedemoomichi: yeah, so looking at what calls that method, it looks like the instance we get here (or don't), doesn't even matter00:05
mriedemhttps://github.com/openstack/nova/blob/master/nova/api/openstack/compute/floating_ips.py#L19500:05
mriedembecause the instance passed in for that method in neutron isn't used.00:05
mriedemdoesn't look like the instance is used in the nova-network case either00:06
*** baoli has joined #openstack-nova00:07
oomichimriedem: oops, I misunderstood the instance is used in nova-network case..00:07
mriedemit's not used00:08
mriedemfor DELETE os-floatingips00:08
mriedemwe get the instance, then we pass it to the network API but it's not used by either network API00:08
*** diogogmt has joined #openstack-nova00:11
*** shaohe_feng has quit IRC00:12
*** shaohe_feng has joined #openstack-nova00:12
oomichimriedem: humm, then current API layer is doing unnecessary operation for the delete00:14
mriedemyup00:14
mriedemdisassociate_and_release_floating_ip and disassociate_floating_ip both take an instance but never use it00:14
mriedemso, i'm going to put up a WIP patch to see what removing instance from those network APIs does00:15
oomichimriedem: ok, I see that on the code now. How about fixing the bug on current one and removing that lately?00:16
mriedembecause none of them are versioned either so might be fine to just clean this up00:16
oomichimriedem: cool00:16
mriedemyeah i already +2d your change00:16
mriedemthere are comments inline00:16
mriedemthe logged message is redundant00:16
mriedemif you fix i'd +2 again00:16
oomichimriedem: ok, I changed the code localy and waiting pep8 check00:21
*** shaohe_feng has quit IRC00:22
*** shaohe_feng has joined #openstack-nova00:22
openstackgerritTakashi NATSUME proposed openstack/nova: Remove nova/api/validator.py  https://review.openstack.org/35539200:23
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Return None in get_instance_id_by_floating_address  https://review.openstack.org/35196000:24
oomichimriedem: ^^^ done00:24
*** Swami has quit IRC00:24
mriedemok00:25
*** thorst_ has joined #openstack-nova00:28
*** woodster_ has quit IRC00:29
*** shaohe_feng has quit IRC00:32
cfriesenCould someone take a look at this and see if I'm smoking crack?  I think I'm hitting an issue that would affect Liberty/Mitaka upgrades if you're using PCI device tracking.  https://bugs.launchpad.net/nova/+bug/161348800:34
openstackLaunchpad bug 1613488 in OpenStack Compute (nova) "changed fields of versionedobjects not tracked properly when down-versioning object" [Undecided,New]00:34
cfriesenbasically it's tied to oslo_versionedobjects and how the "changed fields" are tracked during RPC calls when the object gets down-versioned00:35
*** shaohe_feng has joined #openstack-nova00:35
*** hieulq has joined #openstack-nova00:36
*** gongysh has joined #openstack-nova00:36
mriedemdamn, so disassociate_floating_ip is decorated with @refresh_cache which does take an instance, that's the point00:37
mriedemoomichi: ^00:37
*** smatzek has joined #openstack-nova00:38
mriedemand if we don't pass an instance, we might hit a value error in there00:38
mriedem        except ValueError:00:38
mriedem            msg = _('instance is a required argument to use @refresh_cache')00:38
mriedem            raise Exception(msg)00:38
mriedemdisassociate_and_release_floating_ip doesn't have the refresh_cache decorator, but for nova-network disassociate_and_release_floating_ip calls disassociate_floating_ip, which does refresh the cache00:40
mriedemlooks like we don't do that in the caes of neutron, which is probably a separate bug00:41
*** shaohe_feng has quit IRC00:42
*** shaohe_feng has joined #openstack-nova00:43
mriedemdansmith: you know what's awesome, network api methods with decorators forcing code in the rest api which is totally obfuscated ^00:44
mriedemrest api: pass this down to the network api, even if the parameter isn't directly used by the method, but note the decorator...00:44
dansmithmriedem: I don't even know what you just said00:45
dansmithsounds bad tho00:45
mriedemha00:45
mriedemwhen you disassociate a floating IP from a server, we're passing the server down to the network API but the server parameter is only used by the refresh_cache decorator,00:46
mriedemand only for nova-network00:46
mriedem:(00:46
mriedems/disassociate/delete/ a floating ip00:46
*** amotoki has joined #openstack-nova00:47
*** jamielennox is now known as jamielennox|away00:48
*** smatzek has quit IRC00:49
*** jamielennox|away is now known as jamielennox00:49
*** tonytan4ever has joined #openstack-nova00:50
*** csmart has quit IRC00:52
*** shaohe_feng has quit IRC00:53
*** csmart has joined #openstack-nova00:53
*** shaohe_feng has joined #openstack-nova00:53
*** cheneydc has joined #openstack-nova00:55
*** hoangcx has joined #openstack-nova00:56
*** phuongnh has joined #openstack-nova00:58
*** nic has quit IRC00:59
*** cheneydc has quit IRC01:01
*** sameo has quit IRC01:03
*** shaohe_feng has quit IRC01:03
*** shaohe_feng has joined #openstack-nova01:04
*** gongysh has quit IRC01:05
dansmithmriedem: this passed ironic and the gate is clear, so ready for +W https://review.openstack.org/#/c/355659/01:07
*** baoli has quit IRC01:10
*** baoli has joined #openstack-nova01:12
*** shaohe_feng has quit IRC01:13
mriedemdansmith: what about this https://review.openstack.org/#/c/355602/ ?01:13
mriedemalso need that or not?01:13
mriedemor that's just the revert of the revert01:13
*** BjoernT has joined #openstack-nova01:14
*** shaohe_feng has joined #openstack-nova01:14
dansmithmriedem: right, I'll finish that tomorrow01:15
*** BjoernT has quit IRC01:18
*** Apoorva_ has joined #openstack-nova01:19
mriedemso https://review.openstack.org/#/c/355659/ doesn't fix the cellsv1 job failure does it? looks like https://review.openstack.org/#/c/355602/ fixes that since the job was failing setup with "No hosts found to map to cell, exiting."01:21
mriedembut i guess even if we got past that, you'd get a duplicate entry exception01:21
mriedemwhich is what https://review.openstack.org/#/c/355659/ is for01:21
mriedembecause of this https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api_models.py#L14401:22
*** Apoorva has quit IRC01:23
*** shaohe_feng has quit IRC01:23
*** shaohe_feng has joined #openstack-nova01:24
*** Apoorva_ has quit IRC01:24
*** haplo37__ has joined #openstack-nova01:24
*** sameo has joined #openstack-nova01:24
*** kevinz has joined #openstack-nova01:25
*** ijw has quit IRC01:32
*** shaohe_feng has quit IRC01:34
*** EinstCrazy has joined #openstack-nova01:34
*** shaohe_feng has joined #openstack-nova01:35
*** wangqun has joined #openstack-nova01:37
*** ijw has joined #openstack-nova01:38
*** thorst_ has quit IRC01:38
*** thorst_ has joined #openstack-nova01:39
*** rfolco has quit IRC01:39
*** baoli has quit IRC01:43
*** amotoki has quit IRC01:43
*** vishwanathj has quit IRC01:44
*** shaohe_feng has quit IRC01:44
*** shaohe_feng has joined #openstack-nova01:44
*** Sukhdev has quit IRC01:44
*** gongysh has joined #openstack-nova01:45
*** thorst_ has quit IRC01:48
*** ijw has quit IRC01:49
*** baoli has joined #openstack-nova01:51
openstackgerritxhzhf proposed openstack/nova: [vncproxy] log for closing web is misleading  https://review.openstack.org/35380701:52
*** haplo37__ has quit IRC01:53
*** su_zhang has quit IRC01:54
*** shaohe_feng has quit IRC01:54
*** shaohe_feng has joined #openstack-nova01:55
*** gongysh has quit IRC01:55
*** EinstCrazy has quit IRC01:59
*** ijw has joined #openstack-nova02:01
*** thorst_ has joined #openstack-nova02:02
*** tiendc has joined #openstack-nova02:02
*** thorst_ has quit IRC02:03
*** EinstCrazy has joined #openstack-nova02:03
*** gongysh has joined #openstack-nova02:04
*** shaohe_feng has quit IRC02:04
*** dimtruck is now known as zz_dimtruck02:05
*** shaohe_feng has joined #openstack-nova02:05
*** zz_dimtruck is now known as dimtruck02:05
*** ijw has quit IRC02:06
*** jamielennox is now known as jamielennox|away02:07
*** xinliang has quit IRC02:09
*** xinliang has joined #openstack-nova02:09
*** xinliang has quit IRC02:09
*** xinliang has joined #openstack-nova02:10
*** EinstCra_ has joined #openstack-nova02:11
*** pradk has quit IRC02:12
*** EinstCrazy has quit IRC02:14
*** shaohe_feng has quit IRC02:15
*** shaohe_feng has joined #openstack-nova02:15
*** gongysh has quit IRC02:20
*** baoli has quit IRC02:21
*** shaohe_feng has quit IRC02:25
*** raunak has quit IRC02:25
*** shaohe_feng has joined #openstack-nova02:26
*** jamielennox|away is now known as jamielennox02:26
*** mriedem is now known as mriedem_away02:27
*** zerda2 has joined #openstack-nova02:27
*** gcb_ has joined #openstack-nova02:31
*** shaohe_feng has quit IRC02:35
*** rama_y has joined #openstack-nova02:35
*** houming has joined #openstack-nova02:36
*** gongysh has joined #openstack-nova02:36
*** shaohe_feng has joined #openstack-nova02:36
*** sameo has quit IRC02:37
*** mdrabe has joined #openstack-nova02:39
*** ijw has joined #openstack-nova02:40
*** fragatin_ has joined #openstack-nova02:42
*** fragatina has quit IRC02:45
*** shaohe_feng has quit IRC02:45
*** zhenguo has joined #openstack-nova02:46
*** shaohe_feng has joined #openstack-nova02:46
*** fragatin_ has quit IRC02:46
*** gongysh_ has joined #openstack-nova02:48
*** code-R has joined #openstack-nova02:48
openstackgerritMichael Still proposed openstack/nova: Documentation for the vendordata reboot.  https://review.openstack.org/35379302:50
*** yamahata has quit IRC02:54
*** ijw has quit IRC02:54
*** shaohe_feng has quit IRC02:56
*** shaohe_feng has joined #openstack-nova02:56
*** sudipto has joined #openstack-nova02:57
*** sudipto_ has joined #openstack-nova02:57
*** gongysh_ has quit IRC02:57
*** baoli has joined #openstack-nova03:01
*** code-R has quit IRC03:01
*** krtaylor has joined #openstack-nova03:02
*** yamahata has joined #openstack-nova03:03
*** thorst_ has joined #openstack-nova03:03
*** shaohe_feng has quit IRC03:06
*** shaohe_feng has joined #openstack-nova03:07
*** dimtruck is now known as zz_dimtruck03:07
*** raunak has joined #openstack-nova03:08
*** thorst_ has quit IRC03:11
*** shaohe_feng has quit IRC03:16
*** shaohe_feng has joined #openstack-nova03:16
*** raunak has quit IRC03:20
*** daidv has joined #openstack-nova03:20
*** raunak has joined #openstack-nova03:21
*** haplo37__ has joined #openstack-nova03:22
*** su_zhang has joined #openstack-nova03:24
*** vishwanathj has joined #openstack-nova03:25
*** baoli has quit IRC03:26
*** gongysh has quit IRC03:26
*** shaohe_feng has quit IRC03:26
*** shaohe_feng has joined #openstack-nova03:27
*** gongysh has joined #openstack-nova03:28
*** tiendc has quit IRC03:31
*** diogogmt has quit IRC03:34
*** shaohe_feng has quit IRC03:37
*** shaohe_feng has joined #openstack-nova03:37
*** tiendc has joined #openstack-nova03:37
*** yamahata has quit IRC03:43
*** daidv has quit IRC03:45
*** shaohe_feng has quit IRC03:47
*** shaohe_feng has joined #openstack-nova03:48
openstackgerritMaciej Szankin proposed openstack/nova: Fix check_config_option_in_central_place  https://review.openstack.org/35559703:48
*** tbachman has quit IRC03:49
*** gongysh has quit IRC03:52
*** vishwanathj has quit IRC03:54
hieulqmikal: can you take a look at this patch-set https://review.openstack.org/#/c/316539/ ?03:55
*** cfriesen has quit IRC03:56
*** gongysh has joined #openstack-nova03:56
*** hshiina has joined #openstack-nova03:56
*** cfriesen has joined #openstack-nova03:56
*** code-R has joined #openstack-nova03:56
*** shaohe_feng has quit IRC03:57
*** shaohe_feng has joined #openstack-nova03:58
*** code-R_ has joined #openstack-nova03:59
*** GB21 has joined #openstack-nova04:01
*** code-R has quit IRC04:03
*** shaohe_feng has quit IRC04:07
*** sudipto_ has quit IRC04:08
*** sudipto has quit IRC04:08
*** shaohe_feng has joined #openstack-nova04:08
*** gongysh has quit IRC04:09
*** thorst_ has joined #openstack-nova04:10
*** code-R_ has quit IRC04:12
*** Dinesh_Bhor has joined #openstack-nova04:12
*** code-R has joined #openstack-nova04:12
*** daidv has joined #openstack-nova04:14
*** GB21 has quit IRC04:17
*** vishwanathj has joined #openstack-nova04:17
*** thorst_ has quit IRC04:17
*** shaohe_feng has quit IRC04:18
*** shaohe_feng has joined #openstack-nova04:18
*** links has joined #openstack-nova04:20
*** raunak has quit IRC04:22
*** code-R has quit IRC04:22
*** gongysh has joined #openstack-nova04:23
*** raunak has joined #openstack-nova04:24
*** fragatina has joined #openstack-nova04:24
*** bkopilov has joined #openstack-nova04:25
*** bkopilov is now known as bkopilov_wfh04:25
*** tbachman has joined #openstack-nova04:25
*** tonytan4ever has quit IRC04:27
*** shaohe_feng has quit IRC04:28
*** fragatina has quit IRC04:29
*** chandankumar has joined #openstack-nova04:30
*** shaohe_feng has joined #openstack-nova04:30
openstackgerritTakashi NATSUME proposed openstack/nova: Fix 'No data to report' error  https://review.openstack.org/35571604:34
*** Dinesh_Bhor has quit IRC04:35
*** shaohe_feng has quit IRC04:38
*** shaohe_feng has joined #openstack-nova04:38
*** hieulq has quit IRC04:39
*** irenab has quit IRC04:42
*** marst has quit IRC04:42
*** Sukhdev has joined #openstack-nova04:43
*** sudipto has joined #openstack-nova04:43
*** sudipto_ has joined #openstack-nova04:43
*** pbandark has joined #openstack-nova04:44
*** hieulq has joined #openstack-nova04:45
*** shaohe_feng has quit IRC04:48
*** moshele has joined #openstack-nova04:49
*** shaohe_feng has joined #openstack-nova04:49
*** mdrabe has quit IRC04:54
*** shaohe_feng has quit IRC04:59
*** marst has joined #openstack-nova04:59
*** shaohe_feng has joined #openstack-nova04:59
*** tonytan4ever has joined #openstack-nova05:03
*** GB21 has joined #openstack-nova05:03
*** raunak has quit IRC05:05
*** raunak has joined #openstack-nova05:06
*** shaohe_feng has quit IRC05:09
*** takedakn has joined #openstack-nova05:09
*** shaohe_feng has joined #openstack-nova05:10
*** rama_y has quit IRC05:10
*** rama_y has joined #openstack-nova05:10
*** irenab has joined #openstack-nova05:11
*** code-R has joined #openstack-nova05:15
*** ecelik has joined #openstack-nova05:15
*** thorst_ has joined #openstack-nova05:15
*** su_zhang has quit IRC05:16
*** shaohe_feng has quit IRC05:19
*** shaohe_feng has joined #openstack-nova05:20
*** raunak has quit IRC05:20
*** ekuris has joined #openstack-nova05:21
*** raunak has joined #openstack-nova05:21
*** thorst_ has quit IRC05:22
*** houming has quit IRC05:24
*** tiendc has quit IRC05:25
*** lpetrut has joined #openstack-nova05:26
*** Sukhdev has quit IRC05:26
*** shaohe_feng has quit IRC05:29
*** jaosorior has joined #openstack-nova05:30
*** Wenzhi has quit IRC05:30
*** shaohe_feng has joined #openstack-nova05:30
*** Wenzhi_ has joined #openstack-nova05:31
*** Wenzhi_ has quit IRC05:32
*** prateek has joined #openstack-nova05:33
*** raunak has quit IRC05:35
*** baoli has joined #openstack-nova05:38
*** code-R has quit IRC05:39
*** shaohe_feng has quit IRC05:40
*** fragatina has joined #openstack-nova05:40
*** shaohe_feng has joined #openstack-nova05:40
*** haplo37__ has quit IRC05:41
*** baoli has quit IRC05:42
*** raunak has joined #openstack-nova05:43
*** Dinesh_Bhor has joined #openstack-nova05:43
*** fragatina has quit IRC05:45
*** sridharg has joined #openstack-nova05:46
*** rama_y has quit IRC05:46
*** tonytan4ever has quit IRC05:46
*** shaohe_feng has quit IRC05:50
*** raunak has quit IRC05:50
*** shaohe_feng has joined #openstack-nova05:51
*** tiendc has joined #openstack-nova05:51
*** tonytan4ever has joined #openstack-nova05:54
*** raunak has joined #openstack-nova05:55
*** oanson has joined #openstack-nova05:58
*** markvoelker has quit IRC05:58
*** shaohe_feng has quit IRC06:00
*** shaohe_feng has joined #openstack-nova06:01
*** tonytan4ever has quit IRC06:04
*** paul-carlton2 has joined #openstack-nova06:07
*** raunak has quit IRC06:09
*** shaohe_feng has quit IRC06:10
*** shaohe_feng has joined #openstack-nova06:11
*** rcernin has joined #openstack-nova06:14
*** Hosam_ has joined #openstack-nova06:16
*** mitchel_jhonson has joined #openstack-nova06:16
*** raunak has joined #openstack-nova06:17
mitchel_jhonsonhi all, can I backport a patch which has object version bump?06:17
*** wangqun_ has joined #openstack-nova06:18
mitchel_jhonsonsuppose I have changed a version from 1.6 to 1.7 in objects/request_specs then can I backport it to stable/mitaka branch?06:18
*** rama_y has joined #openstack-nova06:19
*** thorst_ has joined #openstack-nova06:20
*** shaohe_feng has quit IRC06:21
*** wangqun has quit IRC06:21
*** raunak has quit IRC06:21
*** shaohe_feng has joined #openstack-nova06:22
*** diga_ has joined #openstack-nova06:22
*** gongysh has quit IRC06:22
*** raunak has joined #openstack-nova06:25
*** tbachman has quit IRC06:25
*** gongysh has joined #openstack-nova06:25
*** xinliang has quit IRC06:26
*** thorst_ has quit IRC06:27
*** maestropandy has joined #openstack-nova06:28
*** lpetrut has quit IRC06:29
*** Jeffrey4l has joined #openstack-nova06:30
*** unicell has joined #openstack-nova06:30
*** liusheng has quit IRC06:30
*** shaohe_feng has quit IRC06:31
*** liusheng has joined #openstack-nova06:31
*** shaohe_feng has joined #openstack-nova06:32
*** maestropandy has quit IRC06:32
*** maestropandy has joined #openstack-nova06:32
*** zhangjn has quit IRC06:33
*** ljxiash has joined #openstack-nova06:33
*** raunak has quit IRC06:35
*** zhangjn has joined #openstack-nova06:36
*** ljxiash has quit IRC06:37
*** EinstCra_ has quit IRC06:38
*** bkopilov_wfh is now known as bkopilov06:39
*** shaohe_feng has quit IRC06:41
*** EinstCrazy has joined #openstack-nova06:42
*** shaohe_feng has joined #openstack-nova06:42
*** xinliang has joined #openstack-nova06:43
openstackgerritDinesh Bhor proposed openstack/nova: Remove code duplication in enums  https://review.openstack.org/35574506:44
*** raunak has joined #openstack-nova06:47
*** ihrachys has joined #openstack-nova06:47
*** ljxiash has joined #openstack-nova06:49
*** raunak has quit IRC06:50
*** markvoelker has joined #openstack-nova06:51
*** shaohe_feng has quit IRC06:51
*** shaohe_feng has joined #openstack-nova06:52
*** takedakn has quit IRC06:52
*** yamahata has joined #openstack-nova06:53
*** liverpooler has joined #openstack-nova06:53
*** jojden has joined #openstack-nova06:56
jojdenquestion regarding the evacuate06:56
jojdenIs there any chance to get the status of the evacuate.06:56
jojdenLike evacuation is complete or in progress like that06:57
jojden?06:57
*** lpetrut has joined #openstack-nova06:57
*** ljxiash has quit IRC07:00
*** shaohe_feng has quit IRC07:02
*** ljxiash has joined #openstack-nova07:02
*** yamahata has quit IRC07:02
_gryfjojden, afaik no. It is just rebuilding process, so the only change you'll see is the moment of rebuilding, which will become either active or error07:02
*** shaohe_feng has joined #openstack-nova07:03
jojdenoh07:03
jojdenbut if we set the password then after the completion it will return the admin password07:03
jojdenso we can say it is completed right ?07:04
jojden_gryf07:04
_gryfjojden, ttaa07:04
_gryfsorry07:04
_gryfjojden, you can check the destination host07:04
*** maestropandy1 has joined #openstack-nova07:05
*** ljxiash has quit IRC07:05
*** ljxiash has joined #openstack-nova07:05
*** mingyu has joined #openstack-nova07:05
*** maestropandy has quit IRC07:05
jojdencheck the destination host to verify that it contains the vms right ?07:05
_gryfactually07:05
_gryfyou can either select destination host07:05
jojdenyes07:06
_gryfor rely on the scheduler07:06
jojdenok07:06
jojdenWe are selecting the destination host07:06
jojdenit is on  shared storage07:07
_gryfeither way destination host should differ from the (possibly down) source host07:07
_gryfoh07:07
_gryfso you gave to be sure that source host is down07:07
_gryfs/gave/have07:07
jojdenyes07:07
jojdenso how we can say that evacuation is done with success/fail07:08
*** lpetrut has quit IRC07:08
jojdenmeans we have to check the destination is having the instances with same ID, name etc07:08
jojdenright ?07:08
jojden_gryf07:09
*** houming has joined #openstack-nova07:09
*** rstarmer_ has quit IRC07:09
_gryfI think so07:09
jojdenok07:09
jojdenthanks _gryf07:10
*** shaohe_feng has quit IRC07:12
*** shaohe_feng has joined #openstack-nova07:12
*** maestropandy1 has quit IRC07:17
*** maestropandy has joined #openstack-nova07:17
*** nicolasbock has quit IRC07:18
*** nicolasbock has joined #openstack-nova07:19
*** tonytan4ever has joined #openstack-nova07:22
*** shaohe_feng has quit IRC07:22
*** maestropandy has left #openstack-nova07:22
*** shaohe_feng has joined #openstack-nova07:23
openstackgerritLisaLi proposed openstack/nova: Use os-brick encryptors  https://review.openstack.org/32678807:23
*** sahid has joined #openstack-nova07:24
*** jichen has joined #openstack-nova07:26
*** tonytan4ever has quit IRC07:26
*** thorst_ has joined #openstack-nova07:28
*** markus_z has joined #openstack-nova07:28
*** GB21 has quit IRC07:28
*** raunak has joined #openstack-nova07:30
*** ljxiash has quit IRC07:31
*** nicolasbock has quit IRC07:31
*** nicolasbock has joined #openstack-nova07:32
markus_zFYI: nova-bugs-team meeting in 30 min in #openstack-meeting-4 https://wiki.openstack.org/wiki/Meetings/Nova/BugsTeam07:32
*** shaohe_feng has quit IRC07:32
*** thorst_ has quit IRC07:32
*** shaohe_feng has joined #openstack-nova07:33
prateekoomichi, can you please take a look, i have addressed your review comments07:34
*** liverpooler has quit IRC07:35
fricklermarkus_z: I won't have time for the meeting, but maybe someone can look at https://bugs.launchpad.net/nova/+bug/1608934 which is a regression, currently it is not clear to me whether it was introduced into mitaka only recently or whether it has been lurking for over a year07:36
openstackLaunchpad bug 1608934 in OpenStack Compute (nova) "ephemeral disk creation fails for local storage with image type raw/lvm" [Undecided,In progress] - Assigned to Dr. Jens Rosenboom (j-rosenboom-j)07:36
*** ljxiash has joined #openstack-nova07:36
*** liverpooler has joined #openstack-nova07:37
markus_zfrickler: OK, I'll check that one.07:37
*** raunak has quit IRC07:37
*** ljxiash_ has joined #openstack-nova07:38
*** daidv has quit IRC07:38
*** daidv has joined #openstack-nova07:38
openstackgerritAndrey Volkov proposed openstack/nova: Eliminate additional DB queries in nova lists  https://review.openstack.org/35555807:38
*** bauzas_off is now known as bauzas07:38
bauzasgood morning Nova07:38
*** fragatina has joined #openstack-nova07:41
*** ljxiash has quit IRC07:41
*** shaohe_feng has quit IRC07:43
*** rama_y has quit IRC07:44
*** gongysh has quit IRC07:45
*** raunak has joined #openstack-nova07:45
*** belmoreira has joined #openstack-nova07:46
*** fragatina has quit IRC07:46
moshelejohnthetubaguy: hi can you review some pci patches https://review.openstack.org/#/c/349060/ and https://review.openstack.org/#/c/347444/07:46
*** shaohe_feng has joined #openstack-nova07:48
*** RuiChen has joined #openstack-nova07:48
markus_zmoshele: You're leading the sriov subteam, right? I wanted to ask if you could check out some bug reports in that area.07:48
*** prateek has quit IRC07:48
*** prateek_ has joined #openstack-nova07:48
*** baoli has joined #openstack-nova07:50
moshelemarkus_z: yes, what bug?07:50
markus_zmoshele: for example https://bugs.launchpad.net/nova/+bug/157282607:50
openstackLaunchpad bug 1572826 in OpenStack Compute (nova) "dev name in the pci whitelist is not honored for SRIOV" [Undecided,New]07:50
*** GB21 has joined #openstack-nova07:50
markus_zhttps://bugs.launchpad.net/nova/+bug/158282207:51
openstackLaunchpad bug 1582822 in OpenStack Compute (nova) "PCI-PT : SRIOV enabled interface dev_name in pci whitelist does not give the product_id of PF for the direct-physical_network , it always take the VF's product_id" [Undecided,New] - Assigned to Chen Fan (fan-chen)07:51
markus_zhttps://bugs.launchpad.net/nova/+bug/160674707:51
openstackLaunchpad bug 1606747 in OpenStack Compute (nova) "[sriov] nova does't regard second nic after booting" [Undecided,New] - Assigned to xhzhf (guoyongxhzhf)07:51
*** gongysh has joined #openstack-nova07:51
*** gongysh has quit IRC07:51
markus_zmoshele: These are the 3 I currently know about.07:51
*** raunak has quit IRC07:52
*** code-R has joined #openstack-nova07:52
*** shaohe_feng has quit IRC07:53
moshelemarkus_z: sure, I will add the to the agenda, thanks07:53
markus_zmoshele: Cool, thanks a lot!07:53
*** shaohe_feng has joined #openstack-nova07:54
*** baoli has quit IRC07:54
*** Hosam has joined #openstack-nova07:54
*** code-R has quit IRC07:57
*** Hosam_ has quit IRC07:57
*** josecastroleon has joined #openstack-nova07:58
*** zzzeek has quit IRC08:00
*** zzzeek has joined #openstack-nova08:00
*** pilgrimstack has joined #openstack-nova08:01
*** markvoelker has quit IRC08:01
*** shaohe_feng has quit IRC08:03
*** shaohe_feng has joined #openstack-nova08:04
*** paul-carlton2 has quit IRC08:04
*** pilgrimstack has quit IRC08:05
*** raunak has joined #openstack-nova08:06
*** pilgrimstack has joined #openstack-nova08:07
*** roeyc has joined #openstack-nova08:07
*** pkoniszewski_ has joined #openstack-nova08:08
*** danpb has joined #openstack-nova08:09
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: [WIP] Testing latest u-c  https://review.openstack.org/30334508:10
*** raunak has quit IRC08:11
*** shaohe_feng has quit IRC08:13
*** shaohe_feng has joined #openstack-nova08:14
*** matrohon has joined #openstack-nova08:19
openstackgerritPawel Koniszewski proposed openstack/nova: Change default value of live_migration_tunnelled to False  https://review.openstack.org/35048008:19
*** lucas-dinner is now known as lucasagomes08:21
*** tonytan4ever has joined #openstack-nova08:23
*** shaohe_feng has quit IRC08:24
*** paul-carlton2 has joined #openstack-nova08:24
*** shaohe_feng has joined #openstack-nova08:24
openstackgerritAndrey Volkov proposed openstack/nova: Add create cell command  https://review.openstack.org/33271308:26
openstackgerritAndrey Volkov proposed openstack/nova: Trivial: add ability to define action description  https://review.openstack.org/35578108:26
*** wangqun_ has quit IRC08:27
*** tonytan4ever has quit IRC08:28
openstackgerritJohn Hua proposed openstack/nova: Use physical utilisation for cached images  https://review.openstack.org/35386208:29
*** mvk has quit IRC08:29
*** thorst_ has joined #openstack-nova08:30
openstackgerritPawel Koniszewski proposed openstack/nova: [libvirt] Remove live_migration_flag & block_migration_flag  https://review.openstack.org/33486008:31
*** shaohe_feng has quit IRC08:34
*** shaohe_feng has joined #openstack-nova08:35
*** dikonoor has joined #openstack-nova08:35
*** maestropandy has joined #openstack-nova08:35
*** maestropandy1 has joined #openstack-nova08:36
openstackgerritAndrey Volkov proposed openstack/nova: Do not reserve device before reserving the volume  https://review.openstack.org/29079308:36
*** thorst_ has quit IRC08:37
*** cdent has joined #openstack-nova08:39
*** maestropandy has quit IRC08:40
*** zenoway has joined #openstack-nova08:41
mitchel_jhonsonhi all, can I backport a patch which has object version bump?08:41
mitchel_jhonsonsuppose in nova I have changed a version from 1.6 to 1.7 in objects/request_specs then can I backport it to stable/mitaka branch?08:41
*** takashin has left #openstack-nova08:42
*** tiendc has quit IRC08:44
*** shaohe_feng has quit IRC08:44
*** hshiina has quit IRC08:45
*** shaohe_feng has joined #openstack-nova08:45
*** Hosam_ has joined #openstack-nova08:48
*** maestropandy1 has left #openstack-nova08:49
openstackgerritChen Fan proposed openstack/nova: log: delay the variables interpolation in logging code  https://review.openstack.org/35578908:49
*** Hosam has quit IRC08:52
*** cristicalin has joined #openstack-nova08:53
*** PaulMurray has joined #openstack-nova08:54
*** shaohe_feng has quit IRC08:54
*** daidv has quit IRC08:55
*** shaohe_feng has joined #openstack-nova08:55
*** mvk has joined #openstack-nova08:55
*** takashin has joined #openstack-nova08:58
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/35445908:58
*** efoley has joined #openstack-nova09:01
*** markvoelker has joined #openstack-nova09:02
*** derekh has joined #openstack-nova09:03
*** shaohe_feng has quit IRC09:05
*** shaohe_feng has joined #openstack-nova09:05
*** cristicalin has quit IRC09:06
openstackgerrithejiawei proposed openstack/nova: a patch to force detach volume  https://review.openstack.org/35579909:07
*** markvoelker has quit IRC09:07
openstackgerritEvgeny Antyshev proposed openstack/nova: WIP Image format naming conversion from qemu to libvirt  https://review.openstack.org/35580009:07
*** tiendc has joined #openstack-nova09:11
jamespdanyone able to give me a clue about getting the following after a Kilo -> Liberty upgrade?  "Unknown column 'compute_nodes.ram_allocation_ratio' in 'field list'"09:12
*** ljxiash_ has quit IRC09:12
*** shaohe_feng has quit IRC09:15
*** shaohe_feng has joined #openstack-nova09:15
openstackgerritPavel Gluschak proposed openstack/nova: Make live migration possible with libvirt vz driver  https://review.openstack.org/35580509:16
PaulMurrayjohnthetubaguy, do you need anything for neutron in lm meet today ?09:17
*** GB21 has quit IRC09:18
bauzasjamespd: did you run the db sync ?09:19
johnthetubaguyPaulMurray: good question, I have started a write up, it would be nice to get more eyes on that: https://review.openstack.org/#/c/353982/09:20
johnthetubaguyPaulMurray: I will probably try refresh/improve the live-migrate section in a little bit09:20
*** xavpaice has joined #openstack-nova09:20
jamespdbauzas: yeah, now we are getting a bunch of "Cannot call obj_load_attr on orphaned Instance object"09:20
PaulMurrayjohnthetubaguy, I'll put a mention on the agenda09:21
*** Hosam has joined #openstack-nova09:21
johnthetubaguyPaulMurray: great stuff, thanks09:21
openstackgerritJohn Garbutt proposed openstack/nova: Improve help text for service options.  https://review.openstack.org/33286109:21
PaulMurrayI see no the mid cycle page you are going on wednesday?09:21
openstackgerritRaghuveer Shenoy proposed openstack/nova-specs: Support SR-IOV and PCI statistics through Nova API  https://review.openstack.org/34200109:23
markus_zfrickler: I've put it as an agenda item on the nova meeting (see "open discussion"). https://wiki.openstack.org/wiki/Meetings/Nova09:23
johnthetubaguyPaulMurray: I should be there wed 10am (ish) through to the end of friday09:24
markus_zfrickler: I couldn't figure out the exact change(s) but left a suspicious one as comment in the bug report you gave me.09:24
*** Hosam_ has quit IRC09:25
PaulMurrayjohnthetubaguy, I doubt you'll need anything from me but let me know if I can be of help09:25
johnthetubaguyPaulMurray: armax was hoping to pencil in some Nova stuff on Thursday morning09:25
openstackgerritPavel Gluschak proposed openstack/nova: openvz: make live migration possible with libvirt driver  https://review.openstack.org/35580509:25
johnthetubaguyPaulMurray: appreciate it, more eyes on that patch is probably the best thing09:25
PaulMurraynp09:25
*** shaohe_feng has quit IRC09:25
*** rstarmer_ has joined #openstack-nova09:26
johnthetubaguyPaulMurray: I am thinking a join nova, cinder and neutron session at the summit is going to be the best way forward with these new live-migrate flows, discuss a concrete plan there, ideally with POCs already completed09:26
johnthetubaguys/join/joint/09:26
johnthetubaguypossibly including keystone for the token issues we are all having09:26
PaulMurraysounds a lot for a single session09:27
*** shaohe_feng has joined #openstack-nova09:28
PaulMurraydo you know how far the keystone part has got - I had the idea that its just implementation on our part left?09:28
johnthetubaguyPaulMurray: I haven't checked with keystone recently, this isn't the trusts thing, its something different (that is less crazy for deployers)09:29
*** rstarmer_ has quit IRC09:29
johnthetubaguyPaulMurray: we went through some of it at the last summit, in a cross project session, I think thats the last I followed up with that09:29
PaulMurrayjohnthetubaguy, I was thinking about token refresh09:30
* PaulMurray has a meeting09:30
johnthetubaguyPaulMurray: this was more token delegates, for want of a better word09:30
*** thorst_ has joined #openstack-nova09:35
pkoniszewski_johnthetubaguy: hi, i just updated reno here (default value of live_migration_tunelled) - https://review.openstack.org/#/c/350480/ - please let me know whether it is enough, not sure if i should describe it with more details09:35
*** shaohe_feng has quit IRC09:35
johnthetubaguypkoniszewski_: thank you, that looks good now09:36
*** shaohe_feng has joined #openstack-nova09:36
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: WIP: live-migration test hook for serial console  https://review.openstack.org/34747109:36
johnthetubaguypkoniszewski: I was worried about not telling users about the security impact, that covers it nicely09:36
*** lpetrut has joined #openstack-nova09:37
pkoniszewski_johnthetubaguy: yeah, non-tunnelled might not mean non-secure09:39
pkoniszewski_johnthetubaguy: thanks!09:39
*** ociuhandu has joined #openstack-nova09:40
*** cheneydc has joined #openstack-nova09:41
*** thorst_ has quit IRC09:41
*** fragatina has joined #openstack-nova09:43
*** pkoniszewski_ has quit IRC09:44
*** shaohe_feng has quit IRC09:46
*** dmellado has quit IRC09:46
*** shaohe_feng has joined #openstack-nova09:46
*** fragatina has quit IRC09:48
*** cheneydc has quit IRC09:50
*** GB21 has joined #openstack-nova09:51
openstackgerritTakashi NATSUME proposed openstack/nova: Remove nova/api/validator.py  https://review.openstack.org/35539209:52
*** code-R has joined #openstack-nova09:53
*** shaohe_feng has quit IRC09:56
*** dmellado has joined #openstack-nova09:56
*** shaohe_feng has joined #openstack-nova09:56
*** ihrachys has quit IRC09:58
*** code-R has quit IRC09:58
*** roeyc has quit IRC10:00
*** EinstCrazy has quit IRC10:01
*** markvoelker has joined #openstack-nova10:03
*** jed56 has joined #openstack-nova10:03
*** EinstCrazy has joined #openstack-nova10:04
*** shaohe_feng has quit IRC10:06
*** Hosam_ has joined #openstack-nova10:06
*** shaohe_feng has joined #openstack-nova10:07
*** markvoelker has quit IRC10:08
*** Hosam has quit IRC10:09
Dinesh_Bhorjohnthetubaguy: Hi, I have a question regarding backporting. Can I backport a patch which has object version bump ?10:14
*** ihrachys has joined #openstack-nova10:14
*** shaohe_feng has quit IRC10:16
*** shaohe_feng has joined #openstack-nova10:17
*** hoangcx has quit IRC10:18
*** dikonoor has quit IRC10:20
*** lpetrut has quit IRC10:20
*** mingyu has quit IRC10:21
*** ecelik has quit IRC10:22
*** lpetrut has joined #openstack-nova10:22
*** mingyu has joined #openstack-nova10:22
*** ecelik has joined #openstack-nova10:22
*** sdague has joined #openstack-nova10:23
*** ihrachys has quit IRC10:23
*** Sravan has joined #openstack-nova10:24
*** tiendc has quit IRC10:24
*** tonytan4ever has joined #openstack-nova10:24
*** jdillaman has quit IRC10:24
*** mvk has quit IRC10:25
Sravan@cdent : Need some help10:25
cdenthi Sravan what's up?10:25
SravanI am new to micro versions and I gone through the document10:26
Sravanregarding gerrit change10:26
Sravanhttps://review.openstack.org/#/c/349952/210:26
*** mvk has joined #openstack-nova10:26
SravanNeed some suggestions, how to create micro versions and upload10:26
*** shaohe_feng has quit IRC10:27
*** shaohe_feng has joined #openstack-nova10:28
Sravan@cdent , If you have any sample change of micro versions.. That would be very helpful for me to understand10:28
cdentSravan: the important part of that document that I pointed at is the section with the title 'Other necessary changes'. Let me find another change that updated the microversions, just a sec10:28
*** tonytan4ever has quit IRC10:28
Sravanok10:29
cdentSravan: here's a recent change that bumped the microversion: https://review.openstack.org/#/c/284868/10:29
*** ftersin_ has joined #openstack-nova10:30
SravanThanks Chris.. I will go through with that change10:30
Sravanand let you know.. if I stuck at anywhere10:31
SravanThanks much for the help10:31
cdentyou're welcome10:32
*** ftersin_ is now known as ftersin10:32
*** jdillaman has joined #openstack-nova10:36
*** shaohe_feng has quit IRC10:37
johnthetubaguyDinesh_Bhor: I think the answer is no10:37
johnthetubaguyDinesh_Bhor: although the real answer might well be "it depends"10:37
*** shaohe_feng has joined #openstack-nova10:38
*** josecastroleon has quit IRC10:40
*** thorst_ has joined #openstack-nova10:41
*** smatzek has joined #openstack-nova10:42
*** dikonoor has joined #openstack-nova10:44
*** EinstCrazy has quit IRC10:44
*** roeyc has joined #openstack-nova10:45
*** EinstCrazy has joined #openstack-nova10:45
*** dikonoor has quit IRC10:47
*** thorst_ has quit IRC10:47
*** fragatina has joined #openstack-nova10:47
*** shaohe_feng has quit IRC10:47
*** dikonoor has joined #openstack-nova10:47
*** shaohe_feng has joined #openstack-nova10:48
*** phuongnh has quit IRC10:49
*** EinstCrazy has quit IRC10:50
*** dikonoor has quit IRC10:50
*** shaohe_feng has quit IRC10:57
*** shaohe_feng has joined #openstack-nova10:58
*** roeyc has quit IRC11:01
*** Hosam has joined #openstack-nova11:01
moshelejohnthetubaguy: hi can you review some pci patches https://review.openstack.org/#/c/349060/ and https://review.openstack.org/#/c/347444/11:01
*** dingboopt has joined #openstack-nova11:02
*** dmsimard is now known as dmsimard|afk11:03
*** mlavalle has joined #openstack-nova11:03
*** markvoelker has joined #openstack-nova11:04
*** Hosam_ has quit IRC11:04
*** mingyu has quit IRC11:05
*** mingyu has joined #openstack-nova11:06
*** roeyc has joined #openstack-nova11:06
*** shaohe_feng has quit IRC11:08
*** markvoelker has quit IRC11:08
*** shaohe_feng has joined #openstack-nova11:09
*** mingyu has quit IRC11:10
johnthetubaguymoshele: I am trying to focus on the priority specs and things right now I am afraid, hopefully will get to those soon-ish though11:10
moshelejohnthetubaguy: ok thanks11:12
BlackDexHello there. Is there a way to prevent local storage from being used via both the Horizon Dashboard and the nova-api?11:13
*** baoli has joined #openstack-nova11:14
johnthetubaguyBlackDex: no, the code for that blueprint was not completed this cycle, let me find a link to it for you11:14
johnthetubaguyBlackDex: http://specs.openstack.org/openstack/nova-specs/specs/newton/approved/flavor-root-disk-none.html11:15
johnthetubaguyBlackDex: tracking info is here: https://blueprints.launchpad.net/nova/+spec/flavor-root-disk-none11:15
* johnthetubaguy goes for some lunch11:16
BlackDexthx11:16
BlackDexdarn11:16
BlackDexwe want to force volumes only11:16
*** bvanhav has joined #openstack-nova11:17
*** shaohe_feng has quit IRC11:18
*** sudipto_ has quit IRC11:18
*** sudipto has quit IRC11:18
*** shaohe_feng has joined #openstack-nova11:19
*** baoli has quit IRC11:19
*** fragatin_ has joined #openstack-nova11:19
*** ociuhandu has quit IRC11:20
*** haplo37__ has joined #openstack-nova11:20
*** dikonoor has joined #openstack-nova11:23
*** fragatina has quit IRC11:23
*** Dinesh_Bhor has quit IRC11:23
*** GB21 has quit IRC11:24
BlackDexno other way of forcing volumes only?11:25
*** shaohe_feng has quit IRC11:28
*** josecastroleon has joined #openstack-nova11:28
*** shaohe_feng has joined #openstack-nova11:29
sdaguecdent: I think jaypipes needs gerrit lessons - https://review.openstack.org/#/c/352573/ - patches need both a +2 and +W to be prolog sufficient to move forward11:29
*** GB21 has joined #openstack-nova11:29
sdague:)11:29
cdentsdague: :) those three patches have been rather confused over the past few days11:30
*** haplo37__ has quit IRC11:31
sdaguecdent: I'm a little surprised the update doesn't return a 200 with the updated resource in the body - https://review.openstack.org/#/c/329151/63/nova/api/openstack/placement/handlers/resource_provider.py11:33
cdentsdague there's some discussion about that on the spec, if I recall correctly. The gist being since you've already got the name and it is only the name that is in the representation (besides the links), and if you need the links it is quick GET, why bother with a body?11:34
*** thorst_ has joined #openstack-nova11:35
*** jaosorior has quit IRC11:35
sdagueok, well because I was reading on the Location header the other day, it's not really clear that it's supposed to be used with 204 - https://tools.ietf.org/html/rfc7231#section-7.1.211:35
*** jaosorior has joined #openstack-nova11:36
*** sdake has joined #openstack-nova11:36
*** GB21 has quit IRC11:37
cdent204 is not mentioned, but my read is that those response codes that are mentioned are there to explain special cases, and the meaning of the header remains the same across different response codes: to refer to a11:38
cdent   specific resource in relation to the response11:38
*** shaohe_feng has quit IRC11:38
*** shaohe_feng has joined #openstack-nova11:39
cdentsdague: If we were being hard core hateoas then an empty response is bad, but given how uptight people get about how many bytes get sent over the wire, despite it being tiny...11:41
*** rfolco has joined #openstack-nova11:41
sdaguewell, the empty response sits weird with me11:41
sdaguefor the update cases11:41
*** sfinucan has quit IRC11:41
sdagueI thought the big concern was header bloat11:42
sdaguebecause of quirks of http caching appliances11:42
cdentpeople seem to have lots of concerns11:42
cdent(to clarify: about performance when it doesn't seem warranted)11:44
*** sfinucan has joined #openstack-nova11:44
cdentI've linked to your comments here and my response on the review, I'll give it a few hours to see if there's any other comments11:44
sdaguecdent: ok, sure11:45
cdentit's easy to change if it goes that way11:45
sdagueyeh, lets at least wait for jaypipes to tell me I'm wrong :)11:45
openstackgerritsravan proposed openstack/nova: Add tenant_name when list nova usage  https://review.openstack.org/34995211:46
cdentI suspect jay won't care in this case. And I already think you're wrong, but don't feel strongly enough to quibble ;)11:48
*** shaohe_feng has quit IRC11:49
*** shaohe_feng has joined #openstack-nova11:49
*** Sravan has quit IRC11:50
*** rodrigods has quit IRC11:51
*** rodrigods has joined #openstack-nova11:51
*** Dinesh_Bhor has joined #openstack-nova11:53
*** baoli has joined #openstack-nova11:53
*** baoli_ has joined #openstack-nova11:54
*** tonytan4ever has joined #openstack-nova11:55
Dinesh_Bhorjohnthetubaguy: Thanks, suppose on master I have changed object version from 1.7 to 1.8 and in stable/mitaka the object has version 1.5 then can I backport my patch to stable/mitaka?11:57
*** acabot has quit IRC11:57
*** baoli has quit IRC11:58
*** mlavalle has quit IRC11:58
*** jaypipes has joined #openstack-nova11:59
*** shaohe_feng has quit IRC11:59
*** tonytan4ever has quit IRC11:59
*** shaohe_feng has joined #openstack-nova12:00
*** markvoelker has joined #openstack-nova12:05
*** tiendc has joined #openstack-nova12:06
*** lucasagomes is now known as lucas-hungry12:06
*** tiendc has quit IRC12:06
openstackgerritSergey Nikitin proposed openstack/nova: [WIP] get instance diagnostic  https://review.openstack.org/35554012:06
*** mriedem_away has quit IRC12:07
*** mlavalle has joined #openstack-nova12:07
*** Dinesh_Bhor has quit IRC12:07
johnthetubaguyDinesh_Bhor: I don't think thats possible to backport, can you link to the patch you are trying to backport?12:08
*** ptm_away has joined #openstack-nova12:08
ftersinDoes anyone know who do i ask about numerous floating random faults of XenProject CI?12:08
BobBallftersin: That would be anthonyper12:08
*** PaulMurray has quit IRC12:08
ftersinBobBall: thx12:09
*** haplo37__ has joined #openstack-nova12:09
*** markvoelker has quit IRC12:09
*** shaohe_feng has quit IRC12:09
BobBallftersin: or email https://lists.xenproject.org/cgi-bin/mailman/listinfo/wg-openstack12:09
*** moshele has quit IRC12:09
*** Dinesh_Bhor has joined #openstack-nova12:09
*** acabot has joined #openstack-nova12:10
*** shaohe_feng has joined #openstack-nova12:10
*** Hosam_ has joined #openstack-nova12:10
Dinesh_Bhorjohnthetubaguy: I am working on a fix on master which needs to be backported to stable mitaka and it includes object version bump on master so for confirmation I was asking.12:12
*** haplo37__ has quit IRC12:13
johnthetubaguyDinesh_Bhor: it depends on the exact sort of change you are doing, in general its not possible12:14
*** Hosam has quit IRC12:14
Dinesh_Bhorjohnthetubaguy: ok, thank you for your time and info.12:14
*** su_zhang has joined #openstack-nova12:16
*** ducttape_ has joined #openstack-nova12:18
mdboothjaypipes: Thanks for making it so far through that series. I've got a few things to clear up. On the mock thing, though, mock 2 was released on 6th April this year and adds a bunch of stuff which I assume was previously only in python 3's mock.12:18
mdboothSo, assert_called_once() and assert_not_called() are both things now, but they weren't before.12:18
mdboothEven better, though, assert_foo() will raise an attributeerror even on a MagicMock() unless you tell it not to12:19
*** shaohe_feng has quit IRC12:19
mdboothBy default, it no longer automatically mocks methods which start with assert_12:20
mdboothSo typos in assertions now raise errors, unless you also typo 'assert_'12:20
*** shaohe_feng has joined #openstack-nova12:20
*** prateek_ has quit IRC12:22
*** zenoway has quit IRC12:24
*** Hosam has joined #openstack-nova12:24
*** diga__ has joined #openstack-nova12:24
*** diga_ has quit IRC12:25
*** markvoelker has joined #openstack-nova12:26
*** pradk has joined #openstack-nova12:26
*** Hosam_ has quit IRC12:27
*** burgerk has joined #openstack-nova12:27
*** diga has quit IRC12:27
*** mdrabe has joined #openstack-nova12:29
*** takashin has left #openstack-nova12:30
*** shaohe_feng has quit IRC12:30
*** vladikr has joined #openstack-nova12:30
*** shaohe_feng has joined #openstack-nova12:30
*** gouthamr has joined #openstack-nova12:32
jrollanyone have a good way to play with a metadata service, prior to doing all the network fancy to direct 169.254.169.254 traffic to the control plane? short of trying to spoof src IP? :)12:33
johnthetubaguyjroll: I think mikal just added a functional test for that12:35
pilgrimstackQuestion: is their a reason why starting an instance linked on a net without subnet is not possible but creating a port + attaching this port on the same network is possible?12:35
johnthetubaguyjroll: I think there are headers you can pass instead, basically, to claim the correct IP, with a given shared secret12:35
*** ducttape_ has quit IRC12:35
pilgrimstackERROR: Network 3107fa9f-924e-4f00-8ed8-9976870e1598 requires a subnet in order to boot instances on.12:35
johnthetubaguypilgrimstack: its a bug, they should both fail12:35
jrolljohnthetubaguy: yeah, reading the code now, seeing that... thanks12:35
johnthetubaguypilgrimstack: I am curious why you want to do that, use case wise12:36
pilgrimstackjohnthetubaguy: from my point of view it should both work :)12:36
pilgrimstackgood question12:37
johnthetubaguypilgrimstack: the basic idea was to make nova+neutron easier to use by blocking "bad" things, like not having an IP12:37
pilgrimstackwe have many users who want to manage manually (maybe puppet or something else) their IP address on the second network12:37
pilgrimstackcurrently they can do that but they need to create a subnet with a fake range and DHCP disabled12:38
openstackgerritAndrey Shestakov proposed openstack/nova: Add support of attach_interface for ironic virt driver  https://review.openstack.org/35004712:38
pilgrimstacknova will allocate IPs anyway12:39
pilgrimstackit's just a declaration in DB, nothing is managed by OpenStack on thoses nics12:39
johnthetubaguypilgrimstack: I think there was a plan for tagging a port with "i don't need an IP", to avoid all those checks12:39
*** zenoway has joined #openstack-nova12:40
johnthetubaguypilgrimstack: ...but I don't remember if that got anywhere12:40
*** shaohe_feng has quit IRC12:40
*** lpetrut1 has joined #openstack-nova12:41
*** shaohe_feng has joined #openstack-nova12:41
*** lpetrut has quit IRC12:41
*** lpetrut1 is now known as lpetrut12:41
johnthetubaguypilgrimstack: sounds like you actually want that on a network, saying "unmanaged IPs" or something like that12:41
pilgrimstackright12:41
pilgrimstackI need a L2 network and nothing more :)12:41
pilgrimstackno IP management12:41
johnthetubaguypilgrimstack: yeah, seems like a think, just not sure we have APIs for that right now12:42
*** sudipto has joined #openstack-nova12:42
*** sudipto_ has joined #openstack-nova12:42
johnthetubaguys/think/thing/12:42
pilgrimstackIn fact the workaround using port creation and attachement is what  want12:42
pilgrimstackbut I would like to do it when I create an instance too12:43
*** rlrossit has joined #openstack-nova12:43
openstackgerritSean Dague proposed openstack/nova: Documentation for the vendordata reboot.  https://review.openstack.org/35379312:43
johnthetubaguypilgrimstack: right, but supporting that breaks the error handing for most users that do want an IP address12:44
pilgrimstackIf I want an IP address, I have to create a subnet and activate DHCP12:45
pilgrimstackQuestion, what is the use case for a subnet management without DHCP management ?12:45
johnthetubaguypilgrimstack: config drive only with cloud-init setup12:46
johnthetubaguypilgrimstack: basically what rackspace public cloud does12:46
*** mlavalle has quit IRC12:46
pilgrimstackok, good point12:46
*** diga has joined #openstack-nova12:46
*** gcb_ has quit IRC12:47
johnthetubaguyin theory, those VIFs should be locked down to only the allowed mac and IP address, if configured correctly12:47
jaypipesmdbooth: assert_called_once() is not a thing.12:47
*** ociuhandu has joined #openstack-nova12:47
mdboothjaypipes: It is on the mock I have locally, from pypi12:48
*** ociuhandu has quit IRC12:48
mdboothrunning 2.712:48
*** devkulkarni has joined #openstack-nova12:49
*** ociuhandu has joined #openstack-nova12:49
mdboothjaypipes: http://paste.openstack.org/show/558374/12:49
mdboothI'm pretty sure it's new in mock 212:49
pilgrimstackSo from OSI point of view, a L2 network connexion should not require a subnet, that's how it implemented on Neutron side, I think Nova should work like this too12:50
mdboothBut whenever it was added, it certainly exists now12:50
*** ociuhandu has quit IRC12:50
*** Hosam_ has joined #openstack-nova12:50
*** bswartz has joined #openstack-nova12:50
*** shaohe_feng has quit IRC12:50
*** ociuhandu has joined #openstack-nova12:50
*** shaohe_feng has joined #openstack-nova12:51
pilgrimstackjohnthetubaguy: does it help if I create a blueprint/backlog-spec ?12:51
*** mitchel_jhonson has quit IRC12:52
johnthetubaguypilgrimstack: yeah. its worth capturing, I would catch armax in the neutron channel, I feel like there is one already about, but I don't remember where12:52
*** kevinz has quit IRC12:52
pilgrimstackko12:53
pilgrimstackok12:53
*** Hosam has quit IRC12:53
*** devkulkarni has quit IRC12:53
*** devkulkarni has joined #openstack-nova12:54
jaypipesmdbooth: it's not a thing in py3.5 then...12:55
jaypipesmdbooth: I see no mention of it in https://docs.python.org/3/library/unittest.mock.html12:55
mdboothjaypipes: Ah...12:55
jaypipesmdbooth: in any case, all the uses of assert_called_once() can easily be replaced with assert_called_once_with(), no?12:56
*** sonofthor has joined #openstack-nova12:56
mdboothjaypipes: Sure, but I didn't have a reason to do that until you mentioned forward compat.12:56
* mdbooth wonders why they would diverge at this stage.12:56
jaypipesmdbooth: good question, wondering the same :/12:56
*** irenab has quit IRC12:57
*** ociuhandu has quit IRC12:57
jaypipesmdbooth: you think you can address those issues in the patches and re-push today? I should be able to re-review this afternoon for you.12:57
mdboothjaypipes: Yeah, I'll update those.12:57
jaypipeskewl.12:57
*** tblakes has joined #openstack-nova12:58
mdboothjaypipes: It looks like unittest.mock() at least has the feature of not adding methods starting with assert_12:58
mdboothiow the typo catcher12:58
bauzasjaypipes: morning :)12:59
bauzasjaypipes: I'm just looking at Gerrit to see what was merged13:00
bauzasjaypipes: AFAICS, the initial patch for the REST placement API was merged, right?13:00
*** shaohe_feng has quit IRC13:00
bauzasso I can try to add the CRUD operations for the dynamic-resource-classes13:01
*** edand has joined #openstack-nova13:01
*** shaohe_feng has joined #openstack-nova13:01
jaypipesbauzas: yup!13:02
*** xyang1 has joined #openstack-nova13:03
*** woodster_ has joined #openstack-nova13:04
edleafebauzas: custom-resource-classes13:07
edleafe(we had a renaming while you were away)13:07
*** andymaier has joined #openstack-nova13:08
bauzaswoohoo13:08
*** dingboopt has quit IRC13:08
*** prateek has joined #openstack-nova13:08
bauzasedleafe: I had this nit before I went away :p13:08
edleafebauzas: yeah, I thought you'd like it13:08
bauzasbut as a French guy, it was this kind of situation when I was afraid to ask for another name :p13:08
*** ociuhandu has joined #openstack-nova13:08
bauzasplus the not dired trait :p13:09
bauzas(trying to make a pun in your non-native language is awfully terrible)13:09
*** houming has quit IRC13:09
edleafebauzas: even being able to attempt it is more than I can do, so kudos!13:09
*** devkulkarni has quit IRC13:10
*** lucas-hungry is now known as lucasagomes13:10
*** edmondsw has joined #openstack-nova13:11
*** shaohe_feng has quit IRC13:11
*** lifeless has quit IRC13:11
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: WIP: live-migration test hook for serial console  https://review.openstack.org/34747113:11
*** shaohe_feng has joined #openstack-nova13:11
*** mriedem has joined #openstack-nova13:12
jrollmikal: do you have any plans to make the metadata API support the openstack-style metadata, rather than just ec2?13:12
*** andymaier has quit IRC13:13
*** tbachman has joined #openstack-nova13:14
*** roeyc has quit IRC13:14
jrolljaypipes: mdbooth: looks like that's coming in 3.6: http://bugs.python.org/issue26323 https://github.com/testing-cabal/mock/commit/7ca5d3afe293d5e4d769c1cb2dfccb0688ba129213:15
mdboothjaypipes jroll: Hmm, in that case is it worth not using the new interfaces? I'm guessing 3.6 will release before Nova is deployable on python 3.13:16
jrollmdbooth: I think it's fine to use it, we use mock from pypi so that we don't have differences in mock between python versions13:17
jrollbut I'm not a nova core :)13:17
mdboothjroll: Well if they diverged we genuinely wouldn't want to use anything not available in both.13:17
mdboothRegardless of source. However, this is an interesting time window.13:18
jrollmdbooth: I don't think we'll ever use the python stdlib version, idk13:18
jrollwe use lots of mock features not in 2.7 :)13:18
mdboothjroll: That's all of them , isn't it?13:19
jrollI think there's some in 2.713:19
jroll2.6 didn't have it at all iirc13:19
jrollbut idk13:19
*** shaohe_feng has quit IRC13:21
*** shaohe_feng has joined #openstack-nova13:22
johnthetubaguymdbooth: the py35 vs py36 call is likely going to be a distro support thing13:23
mdboothjroll is pointing out we don't use that mock anyway13:23
*** jichen has quit IRC13:23
*** roeyc has joined #openstack-nova13:23
* mdbooth is genuinely on the fence on this, btw13:23
johnthetubaguymdbooth: yeah, thats true13:24
jrollright, the pypi version solely exists so we don't have to care about python version13:24
* mdbooth doesn't mind making the requested changes, but sounds like it's worth a discussion at least13:24
johnthetubaguymdbooth: if it works in the gate, and aids readability, I would support it13:24
johnthetubaguyjroll: +113:24
*** su_zhang has quit IRC13:24
mdboothjohnthetubaguy: Specifically we're talking about assert_called_once() and assert_not_called()13:24
openstackgerritChris Dent proposed openstack/nova: Add support for resource_providers urls  https://review.openstack.org/32915113:25
mdboothThere's also the issue that, historically, a typo in a mock assert always passed13:25
jaypipesjohnthetubaguy: I'd prefer to just use the methods that are available in the standard library mock modules and get this patch series through. We could bikeshed on this for a while.13:25
*** su_zhang has joined #openstack-nova13:25
mdboothjaypipes: That, too.13:25
cdentjaypipes, sdague: that change ^^ updates the update method to return a body and a 200 and no location13:25
cdentI continue to think that's not necessary, but it is nice.13:25
johnthetubaguyjaypipes: true13:26
jrolljaypipes: which version of the stdlib? :)13:26
jrolljaypipes: we already use lots of things not available in 2.713:26
jrollor 3.413:26
jrolletc13:26
* mdbooth is just going to make the updates :)13:26
jaypipesjroll: the ones we gate on: py27 and py3513:26
jaypipesmdbooth: ty13:26
openstackgerritChris Dent proposed openstack/nova: placement: add filtering by attrs to resource_providers  https://review.openstack.org/34304613:27
jrolljaypipes: lots to remove if we restrict ourselves to py2713:27
johnthetubaguyso I thought mock was added in 3.x for some reason13:27
cdentjaypipes, sdague: the stack of three starting at https://review.openstack.org/#/c/352573/ ought to be good to go13:27
*** prateek_ has joined #openstack-nova13:28
mriedemspeaking of mock https://review.openstack.org/#/c/355213/13:28
mriedem^ need that in to unblock getting neutronclient 5.1.0 in g-r13:28
*** lifeless has joined #openstack-nova13:29
mriedemjohnthetubaguy: ^ which we need for carl_baldwin's neutronclient ip allocation client error handling thing for server build abort13:29
bauzasmriedem: I can help :)13:29
*** su_zhang has quit IRC13:29
jaypipescdent: cheers13:29
mriedemjesus it's about time13:30
* cdent goes to pub13:30
johnthetubaguymriedem: looking at that patch now13:30
carl_baldwinmriedem: Thanks for helping out with that. I've been traveling for our mid-cycle but I tried to get everything lined up on top of your fix.13:30
jaypipesmriedem: reviewing that now...13:30
*** prateek has quit IRC13:30
johnthetubaguycarl_baldwin: should see you there tomorrow morning around 10am ish!13:31
*** shaohe_feng has quit IRC13:31
mriedemsdague: these are good to go now https://review.openstack.org/#/q/topic:bug/161241013:31
mriedemthe other failures i was hitting are not related to neutron, i fixed them separately13:31
carl_baldwinjohnthetubaguy: Looking forward to it.13:31
*** andymaier has joined #openstack-nova13:32
*** ashyoung has joined #openstack-nova13:32
sdaguemriedem: ok, approved the novaclient one13:32
*** diogogmt has joined #openstack-nova13:33
mriedemassuming the nv job passes right away, i'll push another change later to make it voting13:33
*** parora has joined #openstack-nova13:34
*** shaohe_feng has joined #openstack-nova13:36
bauzasmriedem: so AFAICS, this is just an immediate resolution patch for fixing the problem, but I hope we could see a different way to mock the neutronclient, right?13:36
bauzaslonger-term of course13:36
*** catintheroof has joined #openstack-nova13:36
jaypipesmriedem: done13:37
*** prateek_ has quit IRC13:37
mriedembauzas: yes13:37
mriedemjaypipes: thanks13:37
mriedembauzas: i didn't want to re-write all of the neutron sec group tests on sunday morning13:37
mriedem:)13:37
openstackgerritMatt Riedemann proposed openstack/os-vif: Add a reminder to remove Route.interface field  https://review.openstack.org/35509313:38
bauzasmriedem: sure, I didn't ask for you, just wanted to know if people are thinking about that :)13:38
mriedemyup, i am13:38
mriedemjust don' have the time to rewrite it all before vacation and the freeze13:38
bauzasokay13:39
openstackgerritJay Pipes proposed openstack/nova: [libvirt] Remove live_migration_flag & block_migration_flag  https://review.openstack.org/33486013:40
*** shaohe_feng has quit IRC13:41
*** shaohe_feng has joined #openstack-nova13:42
*** niedbalski has joined #openstack-nova13:42
*** diogogmt has quit IRC13:42
*** ashyoung has quit IRC13:43
*** catintheroof has quit IRC13:43
*** pwnall has joined #openstack-nova13:44
mriedemsdague: random thing to think/fret about, are we going to release the novaclient proxy (image/neutron) changes as a major version bump? i can see arguments for both ways (major or minor update)13:44
mriedemwith 2.36 i think we said that's a major bump13:44
sdaguemriedem: it's not backwards incompatible, right?13:46
mriedemno,13:46
mriedemso even for 2.36 i'm not sure it's a major release,13:46
sdaguefor 2.36 we should signal as major13:46
mriedemi checked the midcycle etherpad and we only said major bump when we drop the network CLIs13:46
mriedemthe image/neutron proxy stuff is only incompatible if we can't find the proper entry in the service catalog that we're looking up...13:47
mriedemwhich we did differently between those two patches13:47
*** TravT has joined #openstack-nova13:47
mriedemthe neutron one falls back to our proxy api if it can't find a 'network' endpoint13:47
sdagueso, honestly, this is case and point to me why semver is a giant waste of time. :)13:47
*** erhudy has joined #openstack-nova13:48
mriedem2.36 should probably be major b/c of the api side changes13:48
sdagueyeh13:48
mriedemwe handle the CLI, but if you're using the API it's going to be bad13:48
mriedemthat's the same for the image proxy patch though13:49
mriedemoh wel,13:49
mriedembut in that patch we don't use 2.36 yet13:49
mriedemok, so the 2.36 change is definitely major bump13:49
erhudycould a +2 look at https://review.openstack.org/#/c/355415/, fixes a regression introduced in https://review.openstack.org/#/c/187857/13:49
mriedemi just didn't want to do like a 6.0 and then a 7.0 a week later13:49
*** ashyoung has joined #openstack-nova13:50
openstackgerritsean mooney proposed openstack/nova: convert libvirt driver to use os-vif for vhost-user with ovs.  https://review.openstack.org/33404813:50
*** ptm_away is now known as PaulMurray13:50
*** shaohe_feng has quit IRC13:52
*** mlavalle has joined #openstack-nova13:52
PaulMurrayLive migration meeting in 8 mins on #openstack-nova-3 PaulMurray paul-carlton2 eliqiao shaohe_feng pkoniszewski jlanoux yuntongjin mdbooth kashyap kmartin claudiub tobasco tdurakov andreas_s abhishekk diana_clarke mrhillsman13:52
*** shaohe_feng has joined #openstack-nova13:52
PaulMurrayoops - #openstack-meeting-3 PaulMurray paul-carlton2 eliqiao shaohe_feng pkoniszewski jlanoux yuntongjin mdbooth kashyap kmartin claudiub tobasco tdurakov andreas_s abhishekk diana_clarke mrhillsman13:53
*** pwnall has quit IRC13:53
*** hshiina has joined #openstack-nova13:53
*** tonytan4ever has joined #openstack-nova13:53
*** permalac has joined #openstack-nova13:53
*** ecelik has quit IRC13:53
sdaguemriedem: yeh13:53
mriedemand i'm wondering if there are things i should be looking at to get in before we land these13:54
*** burgerk has quit IRC13:54
sdaguemriedem: the one other thing that occurred to me is cli flags to override the sc entry13:54
sdaguewhich we used to do on volumes13:54
mriedemyup13:55
sdaguewhich I'm of 2 minds on13:55
*** haplo37__ has joined #openstack-nova13:55
sdague1) we should do those because we used to13:55
*** yamahata has joined #openstack-nova13:55
sdague2) we *should not* and make cloud providers have a non insane sc13:55
mriedem3) don't do it now and see what happens, we could always add in later13:55
sdaguemriedem: yeh, that's always an option as well13:56
mriedemdo 2 now, add 1 later if needed13:56
sdagueright13:56
mriedemthat was my plan anyway13:56
sdaguesee how big an issue it really is for people13:56
mriedemyup13:56
mriedemi think some of the cruft in there from long ago was from rax oddities13:56
*** ashyoung has quit IRC13:56
sdaguesome of it was13:56
sdaguemaybe we could get mordred to manually test on all his clouds13:57
mordreduhoh13:57
* mordred reads13:57
*** burt has joined #openstack-nova13:57
mriedemsdague: i forgot you can have duplicate networks in neutron (by name)13:58
mriedemyour fake was throwing me  off13:58
mordredsdague: is this a "double check that novaclient still works against existing clouds" ?13:58
jrollsdague: still going to be a problem for rackspace, at least internally, external users are probably fine. e.g. we have multiple nova service catalog entries for a single region in preprod13:58
jrollnot saying don't do (2)13:58
jrolljust pointing out the insanity13:58
alaskione point to consider on the cli flags override is that without them it's not possible to use the client to test endpoints before exposing them in the sc13:59
openstackgerritJulian proposed openstack/nova: Add unit tests for nova.virt.firewall.IpTablesFirewallDriver (Part 4)  https://review.openstack.org/35555013:59
jrollalaski: good point14:00
mordredsdague: also, what's the overall thing(s) that should be verified?14:00
*** ashyoung has joined #openstack-nova14:00
*** anush_ has joined #openstack-nova14:00
*** andymaier has quit IRC14:02
*** shaohe_feng has quit IRC14:02
*** jamesdenton has joined #openstack-nova14:03
*** shaohe_feng has joined #openstack-nova14:03
*** jistr is now known as jistr|debug14:03
mriedemmordred: this change and below https://review.openstack.org/#/c/354981/14:04
mriedemmordred: those are using the catalog to lookup the glance v2 image api and the neutron api so we no longer call proxy apis in nova for image/network things14:04
mriedemb/c the nova proxy apis are deprecated in microversion 2.36 and will return a 404 now14:04
*** pbandark has quit IRC14:06
*** efried has joined #openstack-nova14:06
sdaguemordred: the test would be: nova boot --image $nameofimage --nic net-name=$nameofnet --flavor $nameofflavor myvm14:06
*** pwnall has joined #openstack-nova14:06
*** mingyu has joined #openstack-nova14:06
sdaguename of image and name of net are hitting glance / neutron directly after that point14:06
*** Hosam_ has quit IRC14:06
mordredyah - I can do that - it won't be the quickest thing in the world as I don't have any automation set up to run nova cli things14:07
mordredso it'll be extra manual - but I can do it if it's helpful for sure14:07
*** Hosam has joined #openstack-nova14:07
mordredfwiw, I'm fairly certain that it's a safe change for shade/ansible as we stopped using nova proxy apis a while back \o/14:07
sdaguemordred: it would be a good sniff test... though we could also post to the ML and hope cloud providers might care to test14:07
openstackgerritMaciej Szankin proposed openstack/nova: Updated flags for XVP config options  https://review.openstack.org/34969114:09
*** lpetrut1 has joined #openstack-nova14:10
sdaguejroll: what about image / network endpoints?14:10
*** mingyu has quit IRC14:10
sdaguecompute is already covered by other things14:10
*** armax has joined #openstack-nova14:10
*** lpetrut has quit IRC14:11
*** lpetrut1 is now known as lpetrut14:11
*** Hosam has quit IRC14:11
jrollsdague: I think it's similar, e.g. I have NOVA_VOLUME_SERVICE_NAME=cloudBlockStoragePreProd2 laying around in various forms14:11
*** shaohe_feng has quit IRC14:12
*** shaohe_feng has joined #openstack-nova14:13
mriedemjroll: so for the image/network stuff in nova cli today, how would you be testing pre-prod?14:13
mriedemsince the cli is going to the nova proxy APIs14:13
mriedemso you'd have to configure nova.conf to point at your pre-prod glance/neutron APIs right?14:13
*** su_zhang has joined #openstack-nova14:13
*** alextricity25 has joined #openstack-nova14:13
mriedemsince nova doesn't use the catalog14:14
*** liverpooler has quit IRC14:14
mriedemso if we're not losing anything for image/network, i'm fine with not adding that hook right now14:14
*** BjoernT has joined #openstack-nova14:14
*** tri2sing has joined #openstack-nova14:15
mriedemsdague: want me to post something to the ops list?14:16
mriedemas a heads up14:16
sdaguemriedem: yeh, sure14:16
sdaguemriedem: nova does use the catalog14:16
*** tanvirt has joined #openstack-nova14:16
*** jaosorior has quit IRC14:17
*** su_zhang has quit IRC14:18
mriedemsdague: for glance it uses api_servers which is your hard-coded list of glance APIs right?14:19
jrollmriedem: hrm, why's novaclient even accept a volume_service_name?14:20
sdagueoh, sorry, nova server14:20
mriedemjroll: b/c of rax silly pants14:20
sdaguemriedem: no, not exactly14:20
mriedemi'm 3/4 joking14:20
jrollmriedem: if it just goes through the proxy APIs, I mean14:20
mriedemthe volume service name stuff can be removed now though14:20
mriedemjroll: nova cli used to call cinder api directly14:20
mriedemnot via the proxy14:20
sdaguejroll: because when cinder was split off, the fix for the nova cli was to go directly to cinder instead14:21
jrollmriedem: ah, I see14:21
mriedemit was super awesome14:21
mriedemwe removed that awhile back14:21
*** ducttape_ has joined #openstack-nova14:21
mriedemso the volume_service_name stuff should be removed, i think it's deprecated at least14:21
* jroll nods14:21
jrollI have nfi what we do today then, probably downgrade novaclient :)14:21
*** gongysh has joined #openstack-nova14:21
* jroll hasn't worked on downstream much this cycle14:21
mdboothjaypipes: Did you have an opinion on passing an explicit mock.MagicMock() return value in order to make the patch decorator not add an unused method argument?14:21
mriedemjroll: thought rax had it's own cli?14:22
mriedemsupernova or something14:22
*** kevinz has joined #openstack-nova14:22
* mdbooth thinks it's marginally less ugly than unused arguments14:22
jrollmriedem: no, just a metapackage that installs our lame plugins14:22
*** devkulkarni has joined #openstack-nova14:22
jrollmriedem: supernova is a wrapper around nova that sets env variables for you14:22
mriedemlike VOLUME_SERVICE_NAME14:22
mriedem:)14:22
jrollyep!14:22
mriedemi do remember seeing it in there14:22
*** shaohe_feng has quit IRC14:22
mdboothi.e. @mock.patch('foo.bar.baz', mock.MagicMock()) and then it doesn't add an argument to the test14:22
jrollit's actually quite nice, if you use multiple regions14:22
jrollsupernova personal-iad list -> sources env variables under [personal-iad] in ~/.supernova and calls nova list14:23
*** shaohe_feng has joined #openstack-nova14:23
*** su_zhang has joined #openstack-nova14:26
*** kevinz has quit IRC14:27
*** kevinz has joined #openstack-nova14:27
bauzasmdbooth: AFAIK, use the 'new' attribute14:30
*** ekuris has quit IRC14:30
* mdbooth looks that up14:30
*** zz_dimtruck is now known as dimtruck14:31
*** kfarr has joined #openstack-nova14:32
*** shaohe_feng has quit IRC14:33
mdboothbauzas: Ah, I have actually been using that all along, but without using new=14:33
mriedemyeah new is nice14:33
*** edand has quit IRC14:33
*** shaohe_feng has joined #openstack-nova14:34
mdboothmriedem: Is it possible to use without specifying what it should return?14:34
jaypipesmdbooth: yes, my opinion is don't do that :)14:34
jaypipesmdbooth: be consistent with the code around it.14:34
bauzasmdbooth: http://www.voidspace.org.uk/python/mock/patch.html14:34
mdboothbauzas: Yeah, read that. It's not exactly clear.14:34
bauzasmdbooth: try refraining yourself from using it if that's not super needed14:34
*** mdrabe has quit IRC14:34
mdboothjaypipes: If there was any way to avoid a ton of unused arguments, though, that would be easier to read14:35
*** mdrabe has joined #openstack-nova14:35
bauzasmdbooth: in general, we prefer setting the explicit return value in the method itsefl14:35
*** jed56 has quit IRC14:35
mdboothjaypipes: I know we have a lot of tests like that, but they're pretty ugly14:35
mriedemi might have been thinking of create=14:35
openstackgerritChristopher Martin proposed openstack/nova: Fix for Nova API Dev Guide  https://review.openstack.org/35597814:35
mdboothMeh14:36
mdboothLike I said, I'm not a huge fan of either solution14:36
bauzasmdbooth: do you have a patch in hand so I could bite it ?14:36
bauzasthere is no golden rule14:36
*** burgerk has joined #openstack-nova14:37
cdentmlavalle: you had any additional thoughts/issues about the placement api lately? (my todo list has just reminded me to check in with you)14:37
*** diogogmt has joined #openstack-nova14:37
bauzasbut in case you want to see a difference between a good UT module and a wrong one, look at between test_tracker and test_resource_tracker and you'll understand14:37
bauzasmdbooth:14:37
bauzasthe test_compute module is just insane too14:38
bauzasbut the one who would fix that could be running for president if so14:38
mlavallecdnet: not lately. Right now I am adding to Neutron the interaction with Nova using the placement api. I expect to start testing it later this week. So expect me to come back with comments next week14:38
cfriesenmriedem: dansmith: bauzas:  I think I've found an issue that could bite Liberty/Mitaka upgrades.  Fix proposed at https://review.openstack.org/35598114:38
bauzasI mean, giving peace to the world should be easier than fixing test_compute14:38
cdentmlavalle: cool, thanks14:39
*** links has quit IRC14:39
cfriesenthe trigger for the above issue is that in class PciDevice nova-conductor will write to the "parent_addr" field which is new in Mitaka, then in the downversioned object that field doesn't exist anymore but it's still listed as a changed field.14:39
mlavallecdent: quck question. funny that you pinged exactly right now, when I am coding the creation of the clients. Are the nova client and the placement client always going to be separte? r are they going to fuse in one?14:39
openstackgerritMatt Riedemann proposed openstack/nova: nfs fix for xenial images  https://review.openstack.org/32946614:40
bauzascfriesen: k, looking but I don't have a magic wand for o.vo repo :)14:40
edleafemlavalle: they will be separate14:40
cfriesenbauzas: yeah...I was just wondering if what PciDevice is doing is legit14:40
mlavalleedleafe: ok, thanks14:41
cdentmlavalle: timing is weird. My expectation is that the placementclient library will always be separate and its cli will be generated by osc. This is in line with the expectation that eventually the placement service will be it's own separate thing.14:41
bauzascfriesen: mmm, that's weird14:41
*** marst has quit IRC14:41
bauzascfriesen: because a backport method should make sure to not update the object field when the version is old14:41
bauzascfriesen: so that the field should not be part of obj_what_changed14:42
mlavallecdent: ok, thanks. btw, I still owe you the cleaning of the execptions in the placement client. I'll take care of it over this week or next14:42
cdentmlavalle: that's great, but there's no huge rush: from the nova side there's no immediate dependencies brewing on that code14:42
*** shaohe_feng has quit IRC14:43
cfriesenbauzas: PciDevice._from_db_object() writes to pci_device.parent_addr unconditionally14:43
mlavallecdent: I know. I just want you to know it is in my to do list ;-)14:43
cdent:)14:43
*** shaohe_feng has joined #openstack-nova14:44
*** rama_y has joined #openstack-nova14:44
bauzascfriesen: but obj_make_compatible does the necessary https://github.com/openstack/nova/blob/master/nova/objects/pci_device.py#L13314:45
bauzascfriesen: so that the DB object is modified *before* calling out _from_db_object()14:45
dansmithcfriesen: commented on the review and the bug14:46
mlavallecdent: I am on your side of the pond, attending the Neutron mid-cycle in Cork Ireland14:46
*** brian_price has quit IRC14:46
cdentmlavalle: Awesome, I've never been to Cork but I hear it is lovely14:46
mlavallecdent: it is. Specially if you consider that in San Antonio, TX, where I am from, we get 40C+ degrees this time of the year :-)14:48
cfriesendansmith: thanks.  I'll check out the latest version of Liberty.  What I was seeing is that the older version of the object preserves the nonexistant field in self._changed_fields, and the next time it calls self.save() it causes problems.14:49
dansmithcfriesen: then we should fix *that*, IMHO14:50
*** jistr|debug is now known as jistr14:50
dansmithcfriesen: https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L43414:51
dansmithcfriesen: that looks like it does the right thing to me14:51
dansmithcfriesen: doesn't take changes from the primitive unless they match a field we have14:51
*** ashyoung has quit IRC14:52
cfriesenhttps://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L223dansmith:14:52
cfriesendansmith: https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L22314:52
cfriesensorry, paste error14:52
*** zenoway has quit IRC14:53
*** shaohe_feng has quit IRC14:53
dansmithcfriesen: ah, remotable return right, well, all we have to do is filter that too, IMHO14:53
*** zenoway has joined #openstack-nova14:53
cfriesendansmith: then when we call self.save() from the lower versioned object, we hit https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L651 and choke on the getattr()14:54
*** shaohe_feng has joined #openstack-nova14:54
dansmithcfriesen: we can make that get_changes() be more defensive too for good measure14:55
bauzasdansmith: question, why do we need to do https://github.com/openstack/oslo.versionedobjects/blob/master/oslo_versionedobjects/base.py#L223 ?14:55
*** marst has joined #openstack-nova14:55
bauzasdansmith: if we recreate the object on the receiving side, then it will automatically set the fields, which will update the values, nope ?14:56
*** tongli has joined #openstack-nova14:56
*** kenperkins has joined #openstack-nova14:56
dansmithbauzas: that code mirrors the changed fields list from the calling side14:56
dansmithbauzas: so that if some remotable call touches a field, we see it as dirty on the calling side,14:57
dansmithand if it cleans one, we see it as clean14:57
bauzasoh ok14:57
*** permalac has quit IRC14:58
*** rrecio has joined #openstack-nova14:58
cfriesendansmith: just to clarify...you don't think we need the changes in obj_to_primitive(), but should maybe add some robustness in obj_get_changes()?14:58
*** lpetrut1 has joined #openstack-nova14:58
*** zenoway has quit IRC14:58
dansmithcfriesen: no, I see why now, I think both hunks are fine, but I think you should add the robustness to get_changes while you're in there14:59
*** rrecio_ has joined #openstack-nova14:59
dansmithcfriesen: but if being on the latest liberty fixes this for you for nova, then I think we should just apply this to master and not try some complicated backport14:59
cfriesendansmith: right, makes sense.15:00
*** lpetrut has quit IRC15:00
*** lpetrut1 is now known as lpetrut15:00
*** raj_singh has left #openstack-nova15:00
dansmithcfriesen: also would appreciate a trace in the bug.. I see that most of the detail is there, but my find-the-trace-and-spiral-out-from-there bug technique failed me because you used human words15:00
*** raj_singh has joined #openstack-nova15:00
*** PaulMurray has left #openstack-nova15:00
mriedemtdurakov: this doesn't look to bad for the live migration job after skipping the volume-backed tests http://tinyurl.com/zlbkq4715:01
mriedemtdurakov: that's also check only so it's subject to more spikes15:01
Dinesh_BhorHi all, Can someone please take a look at a small cleanup patch: https://review.openstack.org/#/c/355745/ ?15:02
*** rrecio has quit IRC15:03
mriedemtracking the gate-tempest-dsvm-full-xenial job pretty closely actually http://goo.gl/NLRvl715:03
*** su_zhang has quit IRC15:03
*** shaohe_feng has quit IRC15:03
mriedemdansmith: sdague: johnthetubaguy: ^ we might want to consider making the live migration job voting now that the volume-backed tests are being skipped - those were causing the majority of the random fails on xenial15:04
*** su_zhang has joined #openstack-nova15:04
mriedemthe multinode job also runs live migration but would remain non-voting15:04
dansmithshould we maybe wait until n3 just to avoid a "thought it was ready but it's not" sort of situation?15:05
dansmithgenerally happy to have anything that is solid be voting, just worried about the timing15:05
mriedemthat could also mean letting a regression in before n-315:05
mriedemwe do control the tests for this job in tree15:05
openstackgerritDanil Akhmetov proposed openstack/nova: Handle user-defined ephemeral disk size if provided  https://review.openstack.org/35252215:05
*** su_zhang_ has joined #openstack-nova15:05
*** su_zhang has quit IRC15:06
*** shaohe_feng has joined #openstack-nova15:06
dansmithyeah, true15:06
*** zenoway has joined #openstack-nova15:07
johnthetubaguymriedem: I mean we are using the nodes anyways, so I am OK trying with it voting15:08
mriedemjohnthetubaguy: yeah, and it takes 1/3 of the time of the gate-tempest-dsvm-neutron-multinode-full job15:09
mriedem~30m vs ~90m15:09
*** dikonoor has quit IRC15:11
johnthetubaguymriedem: ah, that would be a win, I didn't know that was a thing15:12
*** devkulkarni has quit IRC15:13
*** devkulkarni has joined #openstack-nova15:13
*** shaohe_feng has quit IRC15:14
*** cdelatte has joined #openstack-nova15:15
*** shaohe_feng has joined #openstack-nova15:15
*** zenoway has quit IRC15:15
*** josecastroleon has quit IRC15:15
*** zenoway has joined #openstack-nova15:15
*** mfisch has joined #openstack-nova15:15
*** mfisch has quit IRC15:15
*** mfisch has joined #openstack-nova15:15
cfriesendansmith: I'm not seeing 'parent_addr' in stable/liberty.....what am I missing?15:15
pkholkinsahid: hello, are you here?15:15
pkholkinI am interested in implementing your spec https://review.openstack.org/#/c/8539915:16
*** devkulkarni has quit IRC15:16
*** devkulkarni has joined #openstack-nova15:16
dansmithcfriesen: I'll have to go do some digging15:17
*** brian_price has joined #openstack-nova15:17
*** roeyc has quit IRC15:17
*** belmoreira has quit IRC15:17
cfriesendansmith: trace added to bug report15:17
sahidpkholkin: hello15:18
sahidyes I think that have never been implemented15:18
dansmithcfriesen: okay I'm thinking of k->l and a different field: https://bugs.launchpad.net/nova/+bug/147407415:19
openstackLaunchpad bug 1474074 in OpenStack Compute (nova) kilo "PciDeviceList is not versioned properly in liberty and kilo" [High,Fix released] - Assigned to Dan Smith (danms)15:19
pkholkinsahid: why not)?15:20
pkholkinI think this is a good spec15:20
*** mtanino has joined #openstack-nova15:20
pkholkinand I can take it for Ocata15:20
sahidpkholkin: no real reason, please do15:21
pkholkinsahid: so, you don't mind?15:21
pkholkinare not optimistic with it?15:22
sahidnot sure to understand but for sure you can work on it, it a valid usecase15:23
sahidI expected to find time one day but...15:23
pkholkinunderstand15:23
*** derekh has quit IRC15:24
*** oanson has quit IRC15:24
*** shaohe_feng has quit IRC15:24
pkholkinso I want to start working on it15:24
pkholkinthanks15:24
Dinesh_Bhorjohnthetubaguy: Hi, Could you please take a look at it : https://review.openstack.org/#/c/355745/ ?15:25
*** shaohe_feng has joined #openstack-nova15:25
dansmithcfriesen: you're going to add the robustness to the get_changes() call as well right?15:25
*** ravikama_ has joined #openstack-nova15:25
*** tongli has quit IRC15:27
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Improve mocking of imagebackend disks  https://review.openstack.org/33324215:27
*** tongli has joined #openstack-nova15:27
openstackgerritJens Rosenboom proposed openstack/nova: Fix ephemeral disk creation  https://review.openstack.org/35541515:29
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Rewrite test_rescue and test_rescue_config_drive  https://review.openstack.org/33596615:29
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Rewrite _test_finish_migration  https://review.openstack.org/33715915:31
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Test disk creation in test_hard_reboot  https://review.openstack.org/33899315:32
*** matrohon has quit IRC15:32
mriedemmelwitt: if you get some time, would be good to get this in https://review.openstack.org/#/c/354349/ - the change after it is ready to land15:34
cfriesendansmith: yeah, that's the plan.  I'm heading off on vacation tomorrow so hopefully can get this done.   After tomorrow anyone can feel free to run with it if they want.15:34
*** shaohe_feng has quit IRC15:34
dansmithcfriesen: okay, I can do it myself if you would prefer15:34
openstackgerritBenjamin VanHavermaet proposed openstack/nova: Integrate OSprofiler and Nova  https://review.openstack.org/25470315:35
*** paul-carlton2 has quit IRC15:35
cfriesendansmith: if you've got time that'd be great15:35
*** su_zhang_ has quit IRC15:35
*** shaohe_feng has joined #openstack-nova15:35
dansmithcfriesen: I don't, but it's my code that's broken, so.. :)15:35
mriedemmelwitt: a lot of that change is just making the fakes / stubs in the tests not suck by passing invalid image ids15:35
cfriesendansmith: am I right that this could bite Liberty/Mitaka upgrades?15:36
dansmithcfriesen: I thought you are specifically seeing that? this has been this way for a while and it has never come up, which is part of my skepticism15:37
*** diga has quit IRC15:37
cfriesendansmith: I'm actually on Kilo (yeah, I know).  But the relevent code is the same for Kilo/Liberty, though it's moved to o.vo of course.15:37
*** su_zhang has joined #openstack-nova15:38
dansmithcfriesen: well, still, I'm not sure why you're hitting it but nobody else has reported it15:38
cfriesendansmith: you need to be using pci devices15:38
dansmithmaybe because of something stupid in PCI specifically, I dunno15:38
dansmithyeah15:38
siva_krishnanHi alaski! Currently I am working on https://bugs.launchpad.net/nova/+bug/1599201.  I found the root cause to be this https://github.com/openstack/nova/blob/master/nova/compute/api.py#L1240.  I just found your TODO on top of it, Can you please let me know, if you had a  chance to look into it ?15:40
openstackLaunchpad bug 1599201 in OpenStack Compute (nova) "Device tagging does not work with block_device_mapping with source=blank,destination=local" [Medium,Confirmed] - Assigned to Sivasathurappan Radhakrishnan (siva-radhakrishnan)15:40
*** amotoki has joined #openstack-nova15:40
*** andreykurilin has quit IRC15:41
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Rewrite _test_finish_migration  https://review.openstack.org/33715915:41
*** andreykurilin has joined #openstack-nova15:42
*** ekuris has joined #openstack-nova15:42
*** amotoki has quit IRC15:43
*** amotoki has joined #openstack-nova15:43
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Test disk creation in test_hard_reboot  https://review.openstack.org/33899315:43
openstackgerritVu Tran proposed openstack/nova: Improve error log when snapshot fails  https://review.openstack.org/35602015:44
*** shaohe_feng has quit IRC15:44
pumaranikarHi johnthetubaguy , please have a look at expected_attr list fix when you get time https://review.openstack.org/#/c/342299/15:45
*** shaohe_feng has joined #openstack-nova15:45
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Cleanup test_create_configdrive  https://review.openstack.org/33911415:45
*** rcernin has quit IRC15:46
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Cleanup spawn tests  https://review.openstack.org/34417515:47
*** haplo37__ has quit IRC15:47
dansmithcfriesen: seems you broke the fixture test too somehow.. did all tests run for you locally?15:48
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Rename Backend snapshot and image  https://review.openstack.org/33327215:48
cfriesenno, code was just a proof of concept.  was just looking at the added robustness stuff.15:49
cfriesenit seems to work in practice though15:49
johnthetubaguypumaranikar: sorry, not sure I will be able to get to that till after the freeze, given all the other things to look at15:50
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Never copy a swap disk during cold migration  https://review.openstack.org/33111515:50
dansmithcfriesen: actually my change makes one of yours not necessary15:51
pumaranikarjohnthetubaguy, ok15:52
cfriesendansmith: the remotable() one?15:52
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Don't re-resize disks in finish_migration()  https://review.openstack.org/33111815:52
dansmithcfriesen: no the other15:52
dansmithcfriesen: see what I just pushed15:52
dansmithcfriesen: I guess I need to write tests for your other change too15:52
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Don't check config disk during cold migration  https://review.openstack.org/33715815:53
*** kevinz has quit IRC15:53
*** hieulq_ has joined #openstack-nova15:55
*** shaohe_feng has quit IRC15:55
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Combine injection info in InstanceConfig  https://review.openstack.org/33779015:55
cfriesendansmith: not sure about your obj_what_changed() modification.  I think that's running on the new side, so self.fields will include the new field15:55
*** shaohe_feng has joined #openstack-nova15:56
dansmithcfriesen: you mean for the middle hunk I deleted? yeah, you're right15:56
*** jojden has quit IRC15:57
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Add create_from_image and create_from_func to Backend  https://review.openstack.org/33324415:57
cfriesendansmith: I was thinking maybe something like this if we wanted more robustness: http://paste.openstack.org/show/558414/15:57
dansmithcfriesen: no, that's confusing the set-ness on the current thing from the definition15:58
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Pass is_block_dev explicitly from create_from_func  https://review.openstack.org/33326315:58
*** rawlin has joined #openstack-nova15:59
cfriesendansmith: the issue I saw is that on the new side we're calling obj_get_changes() on the new side with self._changed_fields including a field that isn't specified in self.  So getattr(self,key) chokes.15:59
dansmithcfriesen: right16:00
dansmithcfriesen: obj_attr_is_set is not the proper thing though.. checking to see if it's in fields is the right thing16:00
cfriesendansmith: but I don't see an easy way of the new side saying "is this field valid in the old object"16:00
cfriesendansmith: but it's in fields on the new side, where the code is running16:00
dansmithcfriesen: you can't.. you can only see that after the primitive conversion16:00
*** Sukhdev has joined #openstack-nova16:01
dansmithcfriesen: the robustness for obj_get_changes() is to protect the old side from a potentially bad primitive sent to it from a newer one16:01
*** markus_z has quit IRC16:01
*** anush_ has quit IRC16:02
cfriesendansmith: ah, okay.  so then we should never send invalid self._changed_fields to the newer side in the first place16:02
dansmithcfriesen: from the newer side you mean16:02
*** bwallis has quit IRC16:03
dansmithcfriesen: actually, I think I meant to remove the other one of your changes -- the remotable one.. because if we're defensive in obj_get_changes() then the remotable fix didn't really need to clean things up,16:03
cfriesendansmith: well, the problem I saw was the newer side sending invalid self._changed_fields to the older side, which sent it back to the newer side, which choked in obj_get_changes() calling getattr() on a field that wasn't set.16:04
dansmithand if the to-primitive change is there, then we'll stop returning bad things16:04
dansmithcfriesen: only because we persisted the change list, but filtered out the actual changed field16:04
*** karthiks is now known as karthiks_afk16:05
cfriesendansmith: right.  and object_action() in ConductorManager sets  updates['obj_what_changed'] on the new side.16:05
*** Sukhdev has quit IRC16:05
*** shaohe_feng has quit IRC16:05
dansmith...so is there still something to resolve?16:05
*** shaohe_feng has joined #openstack-nova16:06
dansmithheh, your change breaks a test because the test is returning invalid fields .. nice :)16:06
cfriesendansmith: I think object_action() calls obj_what_changed on the newer side, so it's still going to include fields that only exist on the new side.16:07
cfriesendansmith: so either we need to fix that, or else we need the remotable() change on the old side.16:07
*** amotoki has quit IRC16:07
dansmithcfriesen: yes, but then obj_to_primitive should filter them from the primitive after the backport16:07
mdboothjaypipes dansmith: Everything updated for your further perusal.16:08
*** sridharg has quit IRC16:09
*** sdake has quit IRC16:09
*** xyang1 has quit IRC16:09
*** rl_person has quit IRC16:09
cfriesendansmith: where would obj_to_primitive filter them from the set of changed fields?16:09
dansmithcfriesen: L612 of your patch?16:09
openstackgerritChris Dent proposed openstack/nova: In placement API send microversion header when error  https://review.openstack.org/35603616:10
cfriesendansmith: oh, I thought you meant without the patch. :)16:10
dansmithcfriesen: no16:10
cdentsdague, jaypipes, alex_xu: patch for placement api (microversion handling improvement) outside of the rest of the stack: https://review.openstack.org/#/c/356036/16:11
cfriesendansmith: okay, so if we take my v1 patch and remove the remotable() changes, then yes I think that would do it.16:11
dansmithcfriesen: plus the defensiveness16:11
dansmiththat's what I'm doing now16:12
*** sahid has quit IRC16:13
dansmithcfriesen: pushed16:13
dansmithcfriesen: note that the stable json fixture test was broken in the same way, caught by making this change16:14
*** paul-carlton2 has joined #openstack-nova16:14
*** nic has joined #openstack-nova16:14
*** hshiina has quit IRC16:15
*** ftersin has left #openstack-nova16:15
*** shaohe_feng has quit IRC16:15
*** shaohe_feng has joined #openstack-nova16:16
cfriesendansmith: does that cover the ConductorManager.object_action() case?16:17
dansmithcfriesen: say more things16:18
*** xyang1 has joined #openstack-nova16:18
cfriesendansmith: in object_action we set  updates['obj_what_changed'] = objinst.obj_what_changed()16:18
*** chandankumar has quit IRC16:19
dansmithcfriesen: which will be new because it has to be16:19
cfriesendansmith: which is used by remotable() on the other (lower-versioned) side to set self._changed_fields16:20
dansmithcfriesen: right, so we'll set the internal list to something bad, but nothing will ever see the bad thing because we filter it in get_changes,16:20
dansmithcfriesen: and for your case when we serialize for save, our outgoing list will be filtered thusly, which means the second trip into conductor will not see the changed-but-not-set field16:21
cfriesendansmith: but remotable() doesn't call obj_what_changed(), rather it looks at updates.get('obj_what_changed')16:21
dansmithcfriesen: read what I said again16:22
dansmithcfriesen: remotable will set the internal list to the "new" list from conductor,16:22
dansmithbut nothing will ever see it because we filter in obj_what_changed()16:22
cdentthanks sdague16:23
*** pilgrimstack has quit IRC16:23
*** Sukhdev has joined #openstack-nova16:23
*** yamahata has quit IRC16:23
cfriesendansmith: and then when we call self.save(), won't we pass the internal (wrong) self._changed_fields back to nova-conductor?16:23
dansmithcfriesen: no because when we make the save call, we call obj_to_primitive(), which filters (after your change) the changes field16:24
dansmithactually, my defensiveness change will have filtered it before we even hit your new code16:25
*** shaohe_feng has quit IRC16:25
dansmithcfriesen: L605 here: https://review.openstack.org/#/c/355981/3/oslo_versionedobjects/base.py16:26
dansmithcfriesen: won't return the bogus field(s) anymore, which is where we're serializing the object to make the save() call16:26
*** shaohe_feng has joined #openstack-nova16:26
cfriesendansmith: okay, I'm sold.  We'll still need a backport to stable/liberty, right?16:27
dansmithcfriesen: note that I'm not saying the remotable change is wrong, I'm saying it should not be necessary, and that's critical code that's semi-hard to test, so less change is better, IMHO16:27
cfriesendansmith: fair enough16:28
dansmithcfriesen: "we" are not backporting non-security things to liberty anymore16:28
dansmithcfriesen: and this is in a library, which also has all kinds of other implications that I would have to ask someone about16:28
dansmithcfriesen: for the specific pci thing, we should be able to craft a smaller fix that just trims the changes list in the backport method, which would be purely nova code as a fix16:29
*** bwallis has joined #openstack-nova16:29
cfriesendansmith: right, that'd be more limited16:30
cfriesendansmith: thanks for digging into this16:30
dansmithparent_addr was added in 2015-Nov.. is that liberty?16:30
*** gongysh has quit IRC16:31
*** baoli_ has quit IRC16:31
melwittmriedem: looking at the novaclient patch now16:32
*** unicell has quit IRC16:34
cfriesendansmith: as far as I can tell parent_addr was added in Mitaka16:34
dansmithoh you said liberty->mitaka, but you're on kilo right?16:34
cfriesendansmith: yes, jumping to Mitaka16:34
*** zenoway has quit IRC16:34
cfriesendansmith: I think we'll hit the bug doing Liberty/Mitaka upgrades16:35
dansmithcfriesen: meaning skipping liberty?16:35
*** zenoway has joined #openstack-nova16:35
cfriesendansmith: yes.  But I think the same bug would be hit upgrading from Liberty16:35
*** shaohe_feng has quit IRC16:36
dansmithmaybe16:36
*** shaohe_feng has joined #openstack-nova16:36
dansmithhaven't had it reported from anyone, some of which I know have gone through the upgrade with pci devices16:37
dansmithbut I can't point to why not16:37
mdboothFun. Just noticed that ephemeral disks are cached according to instance.os_type, not guest_format16:37
mdboothSo if you specify a non-default guest_format for an ephemeral disk on a host which has previously generated an ephemeral disk with a different format16:38
mdboothYou'll get the previous format instead of the one you asked for.16:39
cfriesenmdbooth: nice16:39
mdboothIt gets better, if you migrate it to another host and you're using qcow216:40
*** zenoway has quit IRC16:40
mdboothThe backing file created on the destination will be of the default format, not the one you asked for16:40
mdboothSo, instant data corruption :/16:40
cfriesenmdbooth: so the takeaway is don't use ephemeral disks till O? :)16:40
mdboothMy guess is that it's been this way since the feature was implemented16:41
mdboothSo at this point it probably means that nobody actually ever does this.16:41
*** danpb has quit IRC16:41
mdboothIt's a lot of hassle for a feature nobody seems to be using :/16:42
mriedemlet's remove ephemeral bdms and bdm v116:42
mriedemand cloudpipe16:42
*** thumpba has joined #openstack-nova16:43
mdboothmriedem: Actually, this bug is exclusive to bdm v216:44
mdboothI think...16:44
mriedemoh16:44
mriedemwell,16:44
mriedemlet's just remove bdms16:44
*** shaohe_feng has quit IRC16:46
*** shaohe_feng has joined #openstack-nova16:47
*** su_zhang has quit IRC16:48
*** su_zhang has joined #openstack-nova16:50
dansmithmriedem: do we still have something broken that is resetting the gate?16:52
dansmithmriedem: I'm just seeing a few patches of mine constantly getting reset as they try to make it through16:53
dansmithI thought with the devstack revert we should be good16:53
jaypipesmdbooth: cool, thanks, after meetings will re-review.16:53
melwittjaypipes, alaski, dansmith: fyi, I have proposed a patch for the local disk scheduling/reporting for volume-backed instances that was discussed at the nova meeting last week, at https://review.openstack.org/35509116:56
dansmithmelwitt: yeah, I don't like it16:56
dansmithmelwitt: was trying to decide how to reply on it16:56
*** shaohe_feng has quit IRC16:56
melwittokay16:56
mriedemdansmith: yes16:56
*** gjayavelu has joined #openstack-nova16:56
dansmithbasically, as I've said, I just think it's a bunch of debt and mess to fix a thing that has been around for years on the eve of fixing it for reals16:56
dansmithmelwitt: I'm sure I'm in the minority on that though16:57
mriedemdansmith: https://launchpad.net/bugs/161374916:57
openstackLaunchpad bug 1613749 in OpenStack-Gate "Timeouts when requesting a glance image created with a remote image from git.o.o" [Undecided,New]16:57
*** shaohe_feng has joined #openstack-nova16:57
dansmithmriedem: okay16:57
mriedemdansmith: possibly fixed with https://review.openstack.org/#/c/356010/16:57
dansmithmriedem: oh that, heh :)16:57
melwittdansmith: I get that. for whatever reason at present it seems like several people are hitting the bug and a backportable fix is worthwhile. I tried to make it something that would be easy to delete when we get resource-providers17:00
*** su_zhang has quit IRC17:00
*** kfarr has quit IRC17:00
dansmithmelwitt: yeah and I get that17:01
*** su_zhang has joined #openstack-nova17:01
*** lucasagomes is now known as lucas-dinner17:02
oomichimriedem: I am not sure the meaning of -1 on https://review.openstack.org/#/c/35196017:05
*** cdent has quit IRC17:05
oomichimriedem: are you digging it more deeply?17:05
dansmithmriedem: ah, that fix got promoted I guess?17:05
*** yamahata has joined #openstack-nova17:05
mriedemdansmith: yeah17:06
*** shaohe_feng has quit IRC17:06
*** devkulkarni has quit IRC17:07
alextricity25cfriesen: Hey. Are you around?17:07
johnsomHi Nova folks.  Is there a python client binding for "Show Port Interface Details"?17:07
mriedemoomichi: so, the method your changing is called from 2 different APIs,17:07
mriedemoomichi: one for deleting a floating IP and one for just removing a floating IP from an instance,17:07
mriedemin the latter case, disassociate, if we don't find the port/instance, we don't disassociate the floating ip17:07
*** shaohe_feng has joined #openstack-nova17:07
mriedemif the former case, delete, we pass None to the network API17:07
mriedemoomichi: which last night i thought that instance wasn't used, but it's used by the refresh_cache decorator in the network API17:08
mriedemhowever, the delete floating IP flow doesn't use the refresh_cache decorator for some reason17:08
johnsomI see interface_list, interface_attach/detach but I'm not seeing an interface_show or similar17:08
mriedemso passing instance=None probably doesn't hurt17:08
mriedemjohnsom: no, that's a proxy, use neutron or osc17:09
*** oanson has joined #openstack-nova17:09
*** irenab has joined #openstack-nova17:09
mriedemjohnsom: those proxies are all returning 404s with the 2.36 microversion17:09
mriedemso move off of them17:09
johnsomHmmm, ok, that is good information17:09
*** david-lyle_ has joined #openstack-nova17:09
mriedemoomichi: so i *think* your change won't break anything in the delete floating IP flow, for neutron anyway, but we probably have a latent bug here with not refreshing a server's network info cache when we delete a floating ip17:10
mriedemi haven't been able to test that though17:10
*** tonytan4ever has quit IRC17:11
*** diga__ has quit IRC17:11
*** rawlin has left #openstack-nova17:11
openstackgerritStephen Finucane proposed openstack/nova: objects: Move 'arch' to 'fields.Architecture'  https://review.openstack.org/35606717:12
*** efoley has quit IRC17:12
oomichimriedem: yeah, that is as you said on the review. and that is separated bug17:12
mriedemdisassociate_and_release_floating_ip is the network API method i'm worried about17:12
mriedemfor nova-net that refreshes the cache, for neutron it doesn't17:12
*** david-lyle has quit IRC17:13
mriedemwould be a good tempest test actually17:13
*** david-lyle_ is now known as david-lyle17:13
mriedembut would be racey since nova has to refresh the cache17:13
*** unicell has joined #openstack-nova17:14
mriedemactually no that shouldn't race17:14
mriedemthe REST API is blocking on the network API call17:14
*** shaohe_feng has quit IRC17:17
*** shaohe_feng has joined #openstack-nova17:17
oomichimriedem: in the dissociation case, if the instance is None as the patch, the API returns an error as the expectation17:19
oomichimriedem: on https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/floating_ips.py#L31217:20
*** smatzek has quit IRC17:20
oomichimriedem: that seems valid behavior, right?17:20
*** smatzek has joined #openstack-nova17:20
alaskisiva_krishnan: I have not quite determined the reason behind that TODO I added17:21
mriedemoomichi: i didn't even see the else case here https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/floating_ips.py#L32417:21
*** lpetrut has quit IRC17:21
openstackgerritAlex Szarka proposed openstack/nova: [WIP] Reduce code complexity to 25  https://review.openstack.org/35117617:22
mriedemoomichi: i'm +2 on the change, left a big comment17:22
mriedemoomichi: as for the DELETE call to os-floating-ips, that is a 404 with 2.36 now17:22
mriedemoomichi: so if tempest is using that, it should probably move off17:22
alaskisiva_krishnan: there was a discussion with mriedem and mdbooth and it seems that ephemeral disks can be passed in as block device mappings but then should not be put into the database. But it's not clear if that's what we want or what would be the right thing to change. there just hasn't been much thought put into it17:22
mriedemit should call nova to disassociate the floating IP and then call neutron to delete the floating IP17:22
mriedemoomichi: is there somewhere in tempest i can hack to check the server.addresses before and after a floating IP is deleted?17:23
mriedemto verify the refresh cache bug?17:24
* oomichi checking the code..17:24
mriedemi guess whatever calls self.floating_ips_client.delete_floatingip17:25
mriedemoomichi: at the end of test_minimum_basic_scenario i could maybe delete the floating IP17:26
mriedembefore teardown runs17:26
*** tonytan4ever has joined #openstack-nova17:26
openstackgerritStephen Finucane proposed openstack/nova: objects: Move 'arch' to 'fields.Architecture'  https://review.openstack.org/35606717:27
*** shaohe_feng has quit IRC17:27
*** shaohe_feng has joined #openstack-nova17:28
*** hieulq_ has quit IRC17:29
*** gyee has joined #openstack-nova17:30
oomichimriedem: yeah, the scenario seems basic one, and the test seems good place. The other existing tests are too complex for this case17:30
*** cleong has joined #openstack-nova17:30
*** chandankumar has joined #openstack-nova17:32
mriedemplus that test will hit both nova-net and neutron17:33
*** baoli has joined #openstack-nova17:33
*** hieulq_ has joined #openstack-nova17:36
*** cdent has joined #openstack-nova17:36
*** shaohe_feng has quit IRC17:37
*** sudipto_ has quit IRC17:37
*** sudipto has quit IRC17:37
*** shaohe_feng has joined #openstack-nova17:38
*** paul-carlton2 has quit IRC17:38
*** shaohe_feng has quit IRC17:47
openstackgerritVladik Romanovsky proposed openstack/nova: tests: verify pci passthrough with numa  https://review.openstack.org/35607717:48
*** shaohe_feng has joined #openstack-nova17:48
oomichisdague: can you take a look at https://review.openstack.org/#/c/351960/ ? that will fix some gate jobs(gate-openstack-ansible-os, etc)17:49
*** devkulkarni has joined #openstack-nova17:49
*** hieulq_ has quit IRC17:50
sdaguedansmith / alaski what's this all about - http://logs.openstack.org/60/351960/9/check/gate-tempest-dsvm-cells/33b510f/logs/devstacklog.txt.gz#_2016-08-16_00_55_26_481 ?17:51
*** oanson has quit IRC17:52
dansmithsdague: we need to skip that for cellsv117:52
dansmithsdague: because there are no hosts at the top level17:52
sdaguedansmith: is there a patch?17:52
dansmithsdague: we reverted the devstack patch immediately to stop it17:52
dansmithbecause ironic breaks in the same way17:52
dansmithand I have a todo to finish my replacement17:52
sdaguedansmith: but it hasn't landed yet?17:52
dansmithsdague: https://review.openstack.org/#/c/355599/17:53
sdagueoh, those test results are yesterday at 11:30pm17:53
sdagueok17:53
dansmithI shall go finish that thing now17:53
*** Sukhdev has quit IRC17:53
sdagueok17:53
alaskidansmith: let me know if you need reviews17:54
*** sdake has joined #openstack-nova17:55
dansmithalaski: yeah I just haven't been pushing up anything non-critical because of the backlog17:55
*** baoli has quit IRC17:56
dansmithalaski: so for the cells thing I guess we just don't do it at all if cellsv1 is in play/17:57
*** mlavalle has quit IRC17:57
*** baoli has joined #openstack-nova17:57
*** parora has quit IRC17:57
*** shaohe_feng has quit IRC17:58
alaskidansmith: yeah. I think we may want something in the nova-manage cmd that will bail and explain that the simple command won't work with v117:58
alaskirather than in devstack17:58
dansmithoh17:58
dansmithwell, that's more work17:58
*** shaohe_feng has joined #openstack-nova17:58
dansmithalaski: so we detect that in nova-manage how?17:58
*** tonytan4ever has quit IRC17:59
openstackgerritVladik Romanovsky proposed openstack/nova: adding debug info for pinning calculation  https://review.openstack.org/35515017:59
dansmithalaski: although we still need to ignore the failure status in devstack right? surely you don't want to exit with success for the v1 case right?17:59
mriedemcouldn't nova-manage check the config?18:00
mriedemto see if you're running cells?18:00
alaskiif we have a parsed config in nova-manage then CONF.cells.enable will work18:00
dansmithyeah, but.. I guess18:00
mriedemps aux | grep cells18:00
mriedem:)18:00
*** dmsimard|afk is now known as dmsimard18:01
dansmithjust sucks to have to always have a full config for some of this, especially for things like cellsv1 which requires different configs in different places18:01
mriedemhow about the db?18:01
cdentjroll: is there no ironic v2 thing this week, I forget18:01
*** rcernin has joined #openstack-nova18:01
jrollcdent: just started it, but idk if we have anything to talk about18:01
mriedemthere are tables in the db if you're using cells v1 or not right18:01
mriedem?18:01
cdentjinx18:02
alaskidansmith: that's a fair point on the return code. I have no problem with a devstack skip, was just thinking of being more informative to users if they mistakenly use the simple cmd18:02
dansmithalaski: yeah I think we need both18:02
alaskimriedem: there would be entries in the cells table18:02
alaskibut it's really to config that matters. I think you can configure it without any db entries18:03
mriedemwell then config it up mothergrabber18:03
dansmithI shall18:03
dansmithwe'll be able to merge things in just another few hours18:04
* dansmith headdesks18:04
alaskidansmith: lend infra some of your personal computing resources :)18:04
dansmithheh18:05
*** code-R has joined #openstack-nova18:06
*** ociuhandu has quit IRC18:07
*** shaohe_feng has quit IRC18:08
*** shaohe_feng has joined #openstack-nova18:09
*** Hosam has joined #openstack-nova18:11
openstackgerritSean Dague proposed openstack/python-novaclient: WIP: add support for user specified service types  https://review.openstack.org/35608518:14
*** code-R has quit IRC18:14
sdaguemriedem: that's a first pass on service type flags on the cli18:14
sdaguebecause, I have a feeling we are going to need it18:15
*** cdent has quit IRC18:16
*** lpetrut has joined #openstack-nova18:16
*** shaohe_feng has quit IRC18:18
*** shaohe_feng has joined #openstack-nova18:19
*** jaosorior has joined #openstack-nova18:21
*** chandankumar has quit IRC18:24
*** fragatin_ has quit IRC18:24
*** fragatina has joined #openstack-nova18:24
*** jaosorior has quit IRC18:25
*** tri2sing has quit IRC18:27
*** shaohe_feng has quit IRC18:28
*** baoli has quit IRC18:29
*** baoli has joined #openstack-nova18:29
*** sameo has joined #openstack-nova18:29
*** shaohe_feng has joined #openstack-nova18:29
* dansmith makes a big flushing sound18:30
*** Jeffrey4l has quit IRC18:31
*** ociuhandu has joined #openstack-nova18:32
*** fragatina has quit IRC18:35
*** mvk has quit IRC18:38
*** shaohe_feng has quit IRC18:39
*** shaohe_feng has joined #openstack-nova18:39
*** rlrossit has quit IRC18:40
*** amotoki has joined #openstack-nova18:45
*** jdurgin1 has quit IRC18:45
*** prateek has joined #openstack-nova18:46
*** amotoki has quit IRC18:48
*** su_zhang has quit IRC18:49
*** shaohe_feng has quit IRC18:49
*** shaohe_feng has joined #openstack-nova18:49
*** cdent has joined #openstack-nova18:53
openstackgerritDan Smith proposed openstack/nova: Refuse to run simple_cell_setup on CellsV1  https://review.openstack.org/35609718:53
*** zenoway has joined #openstack-nova18:53
*** prateek_ has joined #openstack-nova18:54
*** prateek has quit IRC18:57
*** tonytan4ever has joined #openstack-nova18:58
*** devkulkarni1 has joined #openstack-nova18:59
siva_krishnanalaski: thanks for the info. Bug mentions that even if the block device mapping doesn't go through, volume/device tagging should have entry in the table.  I am not quite sure if there is way to do it, i thought both we interdependent on each other. Any suggestion would be welcome :)18:59
*** shaohe_feng has quit IRC18:59
*** Sukhdev has joined #openstack-nova18:59
mriedemsiva_krishnan: is that the bug that artom reported?19:00
*** shaohe_feng has joined #openstack-nova19:00
mriedemi think it was basically a known issue/gap when doing the 2.32 microversion that we weren't tagging ephemeral bdms19:00
siva_krishnanmriedem: yes19:00
mriedemok19:00
*** edand has joined #openstack-nova19:00
mriedemi hope that wasn't marked as low-hanging-fruit19:00
*** prateek has joined #openstack-nova19:01
*** devkulkarni has quit IRC19:01
*** baoli has quit IRC19:01
alaskiit's not19:01
*** prateek_ has quit IRC19:02
*** jichen has joined #openstack-nova19:02
alaskisiva_krishnan: they are interdependent. Being able to tag those will mean adding a bdm entry for them, and it's unclear what the side effects of doing that will be19:03
siva_krishnanmriedem since you confirmed it, i thought i would have a look at it.  I was able to create a entry in block_device_mapping table if have volume size mentioned in the  API request which shouldn't  have an entry for it in the bdm table19:04
*** fifieldt has quit IRC19:06
siva_krishnanmriedem: alaski Sorry my last statement was unclear, I was about to say if volume size is mentioned in the request, I was able to create an bdm entry for epehemeral devices as well19:07
*** kfarr has joined #openstack-nova19:07
mriedemoh right19:08
mriedemyeah forgot about volume size19:08
mriedemit doesn't create the bdm entry in the db if the volume size is 0, but it doesn't fail either19:09
*** shaohe_feng has quit IRC19:09
*** shaohe_feng has joined #openstack-nova19:10
*** fragatina has joined #openstack-nova19:13
*** fragatina has quit IRC19:13
siva_krishnanmriedem Is this the way how it is supposed to work ?19:13
*** fragatina has joined #openstack-nova19:13
*** sdake_ has joined #openstack-nova19:14
*** sdake has quit IRC19:14
siva_krishnanIf volume size  is not mentioned, hypervisor driver calculates the size and would given a name for it which is why we dont tag it . Am I right ? mriedem19:15
*** sdake_ has quit IRC19:15
*** sdake has joined #openstack-nova19:16
*** prateek_ has joined #openstack-nova19:16
mriedemsiva_krishnan: don't know19:16
mriedemwhat i do know19:16
mriedemis this is awesome http://www.sugarbushsquirrel.com/19:16
*** fifieldt has joined #openstack-nova19:19
cdentOMG19:19
*** prateek has quit IRC19:19
mriedemlegitimate success tag19:19
*** shaohe_feng has quit IRC19:20
*** shaohe_feng has joined #openstack-nova19:20
* siva_krishnan likes the tagline "Grab ya nuts and follow me..."19:22
*** xyang1 has quit IRC19:29
*** shaohe_feng has quit IRC19:30
*** code-R has joined #openstack-nova19:30
*** shaohe_feng has joined #openstack-nova19:31
*** code-R_ has joined #openstack-nova19:32
*** parora has joined #openstack-nova19:34
*** code-R has quit IRC19:35
*** tongli has quit IRC19:35
*** prateek_ has quit IRC19:37
*** lpetrut has quit IRC19:39
*** shaohe_feng has quit IRC19:40
*** shaohe_feng has joined #openstack-nova19:41
*** bvanhav has quit IRC19:47
*** mnestratov has quit IRC19:50
*** shaohe_feng has quit IRC19:50
*** shaohe_feng has joined #openstack-nova19:51
*** zenoway has quit IRC19:54
*** zenoway has joined #openstack-nova19:55
*** su_zhang has joined #openstack-nova19:55
*** Apoorva has joined #openstack-nova19:55
*** annegentle has joined #openstack-nova19:59
*** zenoway has quit IRC19:59
*** parora has quit IRC20:00
*** shaohe_feng has quit IRC20:01
*** shaohe_feng has joined #openstack-nova20:02
*** edand has quit IRC20:09
*** matrohon has joined #openstack-nova20:09
*** shaohe_feng has quit IRC20:11
*** shaohe_feng has joined #openstack-nova20:12
*** TravT has quit IRC20:19
*** cdelatte has quit IRC20:19
*** shaohe_feng has quit IRC20:21
*** shaohe_feng has joined #openstack-nova20:22
*** TravT has joined #openstack-nova20:22
*** baoli has joined #openstack-nova20:24
mriedemsdague: https://review.openstack.org/#/q/status:merged+project:openstack/python-novaclient+branch:master+topic:bp/deprecate-api-proxies are merged, so i'll rebase dan's change on top of that20:25
mriedemif there is anything we want to release before we land dan's change for 2.36 we should probably do that20:25
mriedemVek: andreykurilin: ^20:25
andreykurilin__I'm ok about it20:26
*** shaohe_feng has quit IRC20:31
*** tblakes_ has joined #openstack-nova20:32
andreykurilin__mriedem: what about keystone related changes? Sorry for asking about them, but I really want to abandon them since nobody is interested in them...20:32
*** shaohe_feng has joined #openstack-nova20:32
mriedemandreykurilin__: so those are just for keystone v3 or to just be using ksa sessions all the time?20:33
mriedemi'm honestly not well versed in keystone things20:33
*** gouthamr has quit IRC20:33
andreykurilin__mriedem: usage of keystone session instead of own implementation solves several issues. One of them good keystone v3 support20:34
*** hemanthm is now known as hemanthm|afk20:34
andreykurilin__Keystone folks were ok with these changes20:34
mriedemandreykurilin__: how was the keystone v3 only job passing on novaclient before that change?20:34
*** tblakes has quit IRC20:35
*** tblakes_ is now known as tblakes20:35
andreykurilin__mriedem: in cli level we already use keystone session20:35
andreykurilin__And our own httpclient works for base keystone v3 case20:35
*** xyang1 has joined #openstack-nova20:36
andreykurilin__But in case of specific cases, it fails20:37
andreykurilin__And nobody wants to maintain HTTPClient20:37
andreykurilin__I can raise my patches at keystone meeting20:37
andreykurilin__mriedem: btw, in our functional jobs I  made change that creates keystone session and initialized novaclient with it20:39
andreykurilin__See base class20:39
mriedemandreykurilin__: ok in https://review.openstack.org/#/c/304035/ you need to fix up the release note20:41
mriedemit's not formatting properly20:41
*** jamesden_ has joined #openstack-nova20:41
andreykurilin__https://github.com/openstack/python-novaclient/blob/master/novaclient/tests/functional/base.py#L20120:42
*** shaohe_feng has quit IRC20:42
mriedemoh i didn't realize there were 3 changes stacked up here20:42
*** shaohe_feng has joined #openstack-nova20:43
andreykurilin__Ok, I'll update release notes when I get home20:43
andreykurilin__Now I'm in karaoke:) lol20:43
*** jamesdenton has quit IRC20:44
*** armax has quit IRC20:45
*** cleong has quit IRC20:49
mriedemuh oh20:49
*** shaohe_feng has quit IRC20:52
*** shaohe_feng has joined #openstack-nova20:53
*** xavpaice has left #openstack-nova20:53
*** Apoorva_ has joined #openstack-nova20:54
openstackgerritSarah Ulmer proposed openstack/nova: Remove deprecated security_group_api option  https://review.openstack.org/35612820:54
openstackgerritJens Rosenboom proposed openstack/nova: Fix ephemeral disk creation  https://review.openstack.org/35541520:54
*** matrohon has quit IRC20:55
*** Apoorva has quit IRC20:57
*** smatzek has quit IRC20:58
*** tonytan4ever has quit IRC20:59
mriedemsdague: blam -2 https://review.openstack.org/#/c/310722/21:02
mriedemwell, -2 via abandon21:02
*** erhudy has quit IRC21:02
*** thorst_ has quit IRC21:02
*** shaohe_feng has quit IRC21:02
*** shaohe_feng has joined #openstack-nova21:03
*** sdague has quit IRC21:04
*** tblakes_ has joined #openstack-nova21:07
*** tblakes has quit IRC21:11
*** tblakes_ is now known as tblakes21:11
*** Apoorva_ has quit IRC21:11
*** Apoorva has joined #openstack-nova21:12
*** shaohe_feng has quit IRC21:12
*** shaohe_feng has joined #openstack-nova21:13
*** matrohon has joined #openstack-nova21:18
*** lpetrut has joined #openstack-nova21:19
*** gyee has quit IRC21:20
*** cdent has quit IRC21:21
*** mvk has joined #openstack-nova21:23
*** shaohe_feng has quit IRC21:23
openstackgerritAndrew Laski proposed openstack/nova: Add BuildRequestList object  https://review.openstack.org/35613721:23
openstackgerritAndrew Laski proposed openstack/nova: WIP Pull from cell0 and build_requests for instance list  https://review.openstack.org/35613821:23
*** gyee has joined #openstack-nova21:23
*** tanvirt has quit IRC21:24
mriedemVek: https://review.openstack.org/#/c/295681/ looks like a legit fix that we should keep, maybe you want to take that one over21:25
*** shaohe_feng has joined #openstack-nova21:26
*** code-R_ has quit IRC21:26
*** rcernin has quit IRC21:27
mriedemhmm it's marked opinion/wishlist in the bug report though for some reason21:28
mikalMorning21:28
*** matrohon has quit IRC21:32
*** annegentle has quit IRC21:32
*** baoli has quit IRC21:33
*** shaohe_feng has quit IRC21:33
melwittmriedem: looks like that was me. I think I was thinking whether it's positional or not is an opinion21:34
*** shaohe_feng has joined #openstack-nova21:34
mriedemmelwitt: well flavor is required21:34
mriedemright?21:34
mriedemoh, well,21:34
mriedemBFV...21:34
mriedembut you still provide a flavor when BFV don't you?21:35
melwittI'm not sure we can change it to a positional parameter because of backward compat but we can make the optional param "required" I think21:35
melwittyeah, you do21:35
melwittI'm not sure that would help in the help text though. marking it as required21:35
melwitt(I'm not sure if required optional params get called out nicely in the help)21:36
mriedemwe can make backward incompatible changes in a major version bump, if we needed to21:36
*** jichen has quit IRC21:36
melwittokay. I think at the time we weren't doing major bumps or something, I can't remember. feel free to change it away from wishlist/opinion if it's something you'd like to do in a major bump21:37
*** thorst_ has joined #openstack-nova21:39
* melwitt runs to an appointment21:41
mriedemi don't really care i guess, i just want to clean up the old reviews for novaclient21:41
*** shaohe_feng has quit IRC21:43
*** thorst_ has quit IRC21:43
*** shaohe_feng has joined #openstack-nova21:44
*** matrohon has joined #openstack-nova21:46
*** jamesden_ has quit IRC21:47
*** kevinz has joined #openstack-nova21:50
*** annegentle has joined #openstack-nova21:51
*** shaohe_feng has quit IRC21:53
*** thorst_ has joined #openstack-nova21:54
*** shaohe_feng has joined #openstack-nova21:54
*** kevinz has quit IRC21:55
*** annegentle has quit IRC21:58
*** edmondsw has quit IRC21:58
*** shaohe_feng has quit IRC22:04
*** mriedem has quit IRC22:04
*** shaohe_feng has joined #openstack-nova22:04
*** devkulkarni1 has quit IRC22:08
*** burt has quit IRC22:08
*** kfarr has quit IRC22:12
*** chrichip has quit IRC22:13
*** shaohe_feng has quit IRC22:14
*** marst has quit IRC22:14
openstackgerritMatt Riedemann proposed openstack/nova-specs: Amend spec for 'get me a network'  https://review.openstack.org/35614722:14
*** shaohe_feng has joined #openstack-nova22:15
*** mdrabe has quit IRC22:17
*** tblakes has quit IRC22:24
*** shaohe_feng has quit IRC22:24
*** shaohe_feng has joined #openstack-nova22:25
*** marst has joined #openstack-nova22:27
*** sdake has quit IRC22:27
*** sdake has joined #openstack-nova22:28
*** dimtruck is now known as zz_dimtruck22:29
*** zz_dimtruck is now known as dimtruck22:29
*** yamahata has quit IRC22:31
*** lpetrut has quit IRC22:31
*** shaohe_feng has quit IRC22:34
*** annegentle has joined #openstack-nova22:34
*** shaohe_feng has joined #openstack-nova22:35
*** dimtruck is now known as zz_dimtruck22:39
*** gouthamr has joined #openstack-nova22:39
*** gouthamr_ has joined #openstack-nova22:40
*** matrohon has quit IRC22:40
*** pwnall has quit IRC22:42
*** gouthamr has quit IRC22:44
*** gouthamr_ is now known as gouthamr22:44
*** shaohe_feng has quit IRC22:45
*** shaohe_feng has joined #openstack-nova22:45
*** thumpba has quit IRC22:46
*** thumpba has joined #openstack-nova22:49
*** thumpba has quit IRC22:49
*** thorst_ is now known as thorst22:49
*** yamahata has joined #openstack-nova22:51
*** burgerk has quit IRC22:51
*** shaohe_feng has quit IRC22:55
*** shaohe_feng has joined #openstack-nova22:56
*** chrichip has joined #openstack-nova22:58
*** thumpba has joined #openstack-nova22:59
*** thumpba has quit IRC22:59
*** thumpba has joined #openstack-nova22:59
*** shaohe_feng has quit IRC23:05
*** shaohe_feng has joined #openstack-nova23:06
*** chrichip has quit IRC23:07
*** bwallis has quit IRC23:08
*** xyang1 has quit IRC23:14
*** shaohe_feng has quit IRC23:15
*** shaohe_feng has joined #openstack-nova23:16
*** thorst has quit IRC23:17
*** thorst has joined #openstack-nova23:17
*** ducttape_ has quit IRC23:18
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Change api-site to v2.1 format  https://review.openstack.org/35616323:20
*** shaohe_feng has quit IRC23:26
*** thorst has quit IRC23:26
*** shaohe_feng has joined #openstack-nova23:26
*** takashin has joined #openstack-nova23:27
*** shaohe_feng has quit IRC23:36
*** shaohe_feng has joined #openstack-nova23:37
*** thumpba_ has joined #openstack-nova23:37
*** thumpba has quit IRC23:39
*** jklare has quit IRC23:39
*** zz_dimtruck is now known as dimtruck23:40
*** bwallis has joined #openstack-nova23:42
*** jamesdenton has joined #openstack-nova23:44
*** shaohe_feng has quit IRC23:46
*** BjoernT has quit IRC23:47
*** shaohe_feng has joined #openstack-nova23:47
*** bswartz has quit IRC23:49
*** annegentle has quit IRC23:50
*** ihrachys has joined #openstack-nova23:50
*** Hosam has quit IRC23:51
*** shaohe_feng has quit IRC23:56
*** shaohe_feng has joined #openstack-nova23:57
*** ducttape_ has joined #openstack-nova23:57
oomichialex_xu: hi, https://bugs.launchpad.net/nova/+bug/1606740 exists in long-term, I guess. How did you know that?23:58
openstackLaunchpad bug 1606740 in OpenStack Compute (nova) "The schema of quota's update isn't include networks quota" [High,In progress] - Assigned to Alex Xu (xuhj)23:58
*** jklare has joined #openstack-nova23:58

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