Tuesday, 2016-12-06

*** masayukig has quit IRC00:01
*** masayukig has joined #openstack-nova00:02
*** ljxiash_ has joined #openstack-nova00:04
*** Swami_ has joined #openstack-nova00:07
*** ljxiash has quit IRC00:08
*** dave-mccowan has quit IRC00:08
*** Swami has quit IRC00:11
*** thorst has joined #openstack-nova00:12
*** hongbin has quit IRC00:14
*** thorst has quit IRC00:15
*** thorst has joined #openstack-nova00:16
*** ducttape_ has joined #openstack-nova00:20
*** thorst has quit IRC00:25
*** catintheroof has quit IRC00:25
*** mfisch has quit IRC00:28
*** mfisch has joined #openstack-nova00:29
*** mfisch has quit IRC00:29
*** mfisch has joined #openstack-nova00:29
*** yingjun has joined #openstack-nova00:29
*** penick has quit IRC00:31
openstackgerritTakashi NATSUME proposed openstack/nova-specs: Abort Cold Migration  https://review.openstack.org/33473200:34
*** tbachman has quit IRC00:37
melwittdansmith, mriedem: I had been wondering if it would be worth trying to make the CellDatabases stuff smarter with how it picks a context manager for get_context_manager. maybe nah for now?00:42
mriedemi get the feeling dan wants to move on00:43
melwitt:)00:43
mriedemas in, let's not sweat the small things right now and try to get things working,00:43
mriedemoptimize later00:43
mriedemand i'm totally fine with that00:43
melwittI think it's cool to move on for now. I've been experimenting with some things and realized the main weirdness is that the API cell has a main database too, which is something I've been not thinking about until I tried to make _last_ctxt_manager smarter00:44
*** ljxiash_ has quit IRC00:46
*** b_rake has quit IRC00:46
dansmithmriedem: melwitt I don't want to change anything until that set gets a little smaller00:47
*** b_rake has joined #openstack-nova00:47
dansmithmelwitt: did you see the one change I did make though?00:47
dansmithI was able to fix a ton of stuff by restoring a default database consistently every time00:47
melwittdansmith: yeah, I saw. I don't completely understand it yet but I'm sure it's related to what I ran into in the end, where there's more than just the compute cell main database and the cell0 main database. there's also the API cell main database and I think that wasn't accounted for until your change00:49
mriedemi'd like to get a little smarter too but that's probably not going to happen00:49
dansmithmriedem: I guess we let in another db migration that isn't first in ocata00:49
mriedemdansmith: i'm fixing that right now00:49
mriedemshould be up in minutes00:49
dansmithokay00:49
*** mingyu has joined #openstack-nova00:50
mriedemdid someone remove testscenarios from test-requirements.txt?00:50
mriedemoh need to rebase00:51
*** b_rake has quit IRC00:51
melwittdansmith: and I think with the default set to cell1, that folds the API cell main database and the compute cell database together for the test00:52
dansmithmelwitt: it folds anything that isn't targeting a cell into the default, yeah00:52
*** hfu has joined #openstack-nova00:52
dansmiththat's the point00:52
*** gjayavelu has quit IRC00:52
*** tovin07 has joined #openstack-nova00:54
*** tovin07_ has joined #openstack-nova00:54
*** mingyu has quit IRC00:55
*** diogogmt has quit IRC00:55
mriedemfun fact, if you run nova/tests/unit/db/test_sqlalchemy_migration.py in isolation it fails00:55
*** penick has joined #openstack-nova00:56
melwittdansmith: yeah, I was noting that's not accurate as I think there are some non targeted calls that are supposed to go to the API cell's main database, not the compute cell's main database. example, flavor read from main database during ensure_migrated00:56
melwittnot always accurate, rather00:56
dansmithmelwitt: well, that one needs to go away anyway, and also no longer makes sense if we can have two cells00:57
melwittdansmith: yeah, that is true00:57
*** yingjun has quit IRC00:57
dansmithbut you could configure your api's main connection to be whichever of the cells used to be your main database and it would still be okay00:57
dansmithmelwitt: but if you look at my set now, there are several things yet to fix -- I fixed instance.fault today, which is loaded later and has to be targetd00:57
dansmithso when we get rid of any and all of those legacy things, then we can remove the defaulting thing and maybe even make a non-targeted access fail so we can identify any leaks00:58
*** penick has quit IRC00:59
melwittdansmith: okay, cool. I'm glad to be convinced what's there now is more reasonable than I started worrying about00:59
dansmithwell, there's plenty left to do, but yeah00:59
openstackgerritMatt Riedemann proposed openstack/nova: Require cellsv2 setup before migrating to Ocata  https://review.openstack.org/39222701:00
*** penick has joined #openstack-nova01:01
*** rcernin has quit IRC01:01
melwittI was trying something like associating each service with a host and patching the rpc receive to stash a cell_host in the RequestContext so it could pick the right one associated with the service making the access. similar for init and periodic task. but then I ran into the flavor read thing, so not everything was covered01:01
*** fragatin_ has quit IRC01:02
*** fragatina has joined #openstack-nova01:03
*** tovin07_ has quit IRC01:04
*** rama_y has quit IRC01:04
*** ljxiash has joined #openstack-nova01:08
*** ljxiash has quit IRC01:08
*** thorst has joined #openstack-nova01:09
*** thorst has quit IRC01:09
*** thorst has joined #openstack-nova01:10
*** tbachman has joined #openstack-nova01:10
*** thorst has quit IRC01:14
*** thorst has joined #openstack-nova01:23
*** thorst has quit IRC01:24
*** unicell1 has joined #openstack-nova01:26
openstackgerritAnusha Unnam proposed openstack/nova: Add more field's in InstancePayload  https://review.openstack.org/40712801:26
openstackgerritzhangyanxian proposed openstack/nova: Fix import method to follow community guideline  https://review.openstack.org/40570301:26
*** rfolco has quit IRC01:26
*** unicell has quit IRC01:27
*** yingjun has joined #openstack-nova01:30
*** gjayavelu has joined #openstack-nova01:30
*** Apoorva has quit IRC01:31
*** zhangjl has joined #openstack-nova01:33
*** yuntongjin has joined #openstack-nova01:35
*** unicell has joined #openstack-nova01:36
*** unicell1 has quit IRC01:38
*** gjayavelu has quit IRC01:38
*** kevinz has joined #openstack-nova01:40
*** esberglu has joined #openstack-nova01:47
*** hongbin has joined #openstack-nova01:47
*** sam_nowitzki has joined #openstack-nova01:50
*** esberglu has quit IRC01:51
*** esberglu has joined #openstack-nova01:53
*** Swami_ has quit IRC01:53
*** nic has quit IRC01:54
*** esberglu has quit IRC01:58
*** penick has quit IRC01:59
*** yufei has joined #openstack-nova02:01
*** esberglu has joined #openstack-nova02:04
*** thorst has joined #openstack-nova02:05
*** thorst has quit IRC02:05
*** thorst has joined #openstack-nova02:06
*** esberglu has quit IRC02:08
*** magic has joined #openstack-nova02:12
*** int32bit has joined #openstack-nova02:14
*** thorst has quit IRC02:15
*** hfu has quit IRC02:15
*** xiaoyang has quit IRC02:16
int32bitI run test on my host and the result is ok, but Jenkins seems fail for 'oslo_config.cfg.NoSuchOptError: no such option osapi_max_limit in group [DEFAULT]', the log: http://logs.openstack.org/71/405171/3/check/gate-nova-python27-db-ubuntu-xenial/28e7f6c/console.html02:17
*** mtanino has quit IRC02:19
melwittint32bit: maybe you need to rebase? (to get the same behavior as jenkins)02:19
int32bitOK, I will try it, thanks melwitt.02:21
melwittif a config options move change has landed since you started your branch, you might be out of date on that02:22
mriedemint32bit: yeah you need to rebase. those options moved last week.02:24
*** browne has quit IRC02:28
*** yuntongjin has quit IRC02:29
int32bitMy work is started from latest master branch, so I have little prospect of moving change the config.02:29
*** gcb has joined #openstack-nova02:31
*** tbachman has quit IRC02:31
*** yuntongjin has joined #openstack-nova02:32
diana_clarkeint32bit: I think you'll need to change "max_limit = CONF.osapi_max_limit" to something like "max_limit = CONF.api.max_limit" in your test_compute.py02:34
openstackgerritHuan Xie proposed openstack/nova: WIP: XenAPI Use os-xenapi lib for nova  https://review.openstack.org/40605902:34
diana_clarkeint32bit: And also in api.py02:35
openstackgerritHuan Xie proposed openstack/nova: XenAPI Use os-xenapi lib for nova  https://review.openstack.org/40605902:35
openstackgerritChangBo Guo(gcb) proposed openstack/nova: conf: Remove config option compute_ manager  https://review.openstack.org/40592102:35
*** yuntongjin1 has joined #openstack-nova02:44
*** hfu has joined #openstack-nova02:44
*** yuntongjin has quit IRC02:45
*** takashin has joined #openstack-nova02:45
*** hongbin has quit IRC02:46
*** yuntongjin1 has quit IRC02:48
*** woodster_ has quit IRC02:56
*** yuntongjin has joined #openstack-nova02:56
*** kaisers__ has joined #openstack-nova02:58
*** sudipto has joined #openstack-nova03:00
*** sudipto_ has joined #openstack-nova03:00
*** kaisers_ has quit IRC03:01
*** kristian__ has joined #openstack-nova03:02
*** brault has quit IRC03:04
*** kristian__ has quit IRC03:06
*** thorst has joined #openstack-nova03:12
*** bkopilov has quit IRC03:13
*** yuntongjin has quit IRC03:17
*** thorst has quit IRC03:20
*** stewie925 has quit IRC03:20
*** kevinz has quit IRC03:30
*** kevinz has joined #openstack-nova03:31
*** hfu has quit IRC03:31
*** baoli has joined #openstack-nova03:33
*** Amit_M has joined #openstack-nova03:34
*** hfu has joined #openstack-nova03:36
*** armax has quit IRC03:36
*** hfu has quit IRC03:40
*** hfu has joined #openstack-nova03:40
*** code-R has joined #openstack-nova03:42
*** CristinaPauna has quit IRC03:43
*** hfu has quit IRC03:44
*** hfu has joined #openstack-nova03:45
*** code-R_ has joined #openstack-nova03:45
*** armax has joined #openstack-nova03:45
*** hfu has quit IRC03:47
*** code-R has quit IRC03:48
*** ducnc has joined #openstack-nova03:48
*** efried has quit IRC03:48
*** efried has joined #openstack-nova03:50
*** hfu has joined #openstack-nova03:50
*** links has joined #openstack-nova03:52
*** hfu has quit IRC03:54
*** mriedem has quit IRC03:56
*** code-R_ has quit IRC03:59
*** code-R has joined #openstack-nova03:59
*** zhangjl has quit IRC03:59
*** zhangjl has joined #openstack-nova04:00
*** baoli has quit IRC04:03
*** psachin has joined #openstack-nova04:03
*** esberglu has joined #openstack-nova04:05
*** masber has joined #openstack-nova04:07
*** hfu has joined #openstack-nova04:08
*** esberglu has quit IRC04:10
openstackgerritJay Pipes proposed openstack/nova: placement: correct improper test case inheritance  https://review.openstack.org/40636304:10
openstackgerritJay Pipes proposed openstack/nova: placement: create aggregate map in report client  https://review.openstack.org/40730904:10
*** esberglu has joined #openstack-nova04:11
*** armax has quit IRC04:12
*** armax has joined #openstack-nova04:13
*** armax has quit IRC04:13
*** hfu has quit IRC04:15
*** esberglu has quit IRC04:16
*** thorst has joined #openstack-nova04:18
*** hfu has joined #openstack-nova04:20
*** int32bit has quit IRC04:20
*** GB21 has joined #openstack-nova04:21
*** mdnadeem has joined #openstack-nova04:21
*** thorst has quit IRC04:25
*** sudipto_ has quit IRC04:25
*** sudipto has quit IRC04:25
*** mnaser has quit IRC04:25
*** mordred has quit IRC04:26
*** greghaynes has quit IRC04:26
*** afazekas has quit IRC04:26
*** sreeram_v has joined #openstack-nova04:26
*** hfu has quit IRC04:27
*** GB21 has quit IRC04:33
*** sridharg has joined #openstack-nova04:34
*** cfriesen has quit IRC04:35
*** CristinaPauna has joined #openstack-nova04:35
*** bkopilov has joined #openstack-nova04:41
*** nicolasbock has quit IRC04:48
*** ratailor has joined #openstack-nova04:48
*** GB21 has joined #openstack-nova04:49
*** alex_xu has joined #openstack-nova04:49
*** alex_xu has quit IRC04:49
*** alex_xu has joined #openstack-nova04:51
openstackgerritDiana Clarke proposed openstack/python-novaclient: Revert "Microversion 2.39 - Simple tenant usage pagination"  https://review.openstack.org/40636004:52
*** afazekas has joined #openstack-nova04:58
*** greghaynes has joined #openstack-nova05:03
*** mordred has joined #openstack-nova05:04
openstackgerritDiana Clarke proposed openstack/python-novaclient: Revert "Microversion 2.39 - Simple tenant usage pagination"  https://review.openstack.org/40636005:06
*** zhangguoqing has joined #openstack-nova05:07
*** nandal has joined #openstack-nova05:08
nandalHi, Is there a way to update some info in release notes?05:09
*** mnaser has joined #openstack-nova05:11
*** yufei has quit IRC05:13
*** prateek_ has joined #openstack-nova05:13
*** ayogi has joined #openstack-nova05:14
*** carthaca_1 has joined #openstack-nova05:15
*** mfisch has quit IRC05:16
*** anteaya has quit IRC05:16
*** mordred has quit IRC05:18
*** mnaser has quit IRC05:18
*** sudipto has joined #openstack-nova05:18
*** sudipto_ has joined #openstack-nova05:18
*** carthaca_ has quit IRC05:19
*** mat128 has quit IRC05:19
*** mingyu has joined #openstack-nova05:21
*** cardeois has quit IRC05:22
*** mfisch has joined #openstack-nova05:23
*** mat128 has joined #openstack-nova05:23
*** mfisch is now known as Guest1619705:23
*** code-R has quit IRC05:23
*** coreywright has quit IRC05:24
*** thorst has joined #openstack-nova05:24
*** cardeois has joined #openstack-nova05:24
*** coreywright has joined #openstack-nova05:24
*** mingyu has quit IRC05:26
*** mordred has joined #openstack-nova05:26
*** anteaya has joined #openstack-nova05:29
gcbnadal, show you know how to use reno to generate releasenotes, just want update info ,right ?05:29
*** code-R has joined #openstack-nova05:29
*** thorst has quit IRC05:30
gcbnandal ^05:30
*** shangxdy has joined #openstack-nova05:35
*** hfu has joined #openstack-nova05:37
nandalgcb, thank you for reply05:38
nandal i was away05:38
nandal¨Now VMwareVCDriver will set disk.EnableUUID=True by default in all guest VM configuration file. To enable udev to generate /dev/disk/by-id¨05:39
nandalThe second sentence of this line is due to a typing mistake05:39
nandalwhich is confusing05:40
*** hfu has quit IRC05:40
nandalIt should be :-05:40
nandal¨Now VMwareVCDriver is set disk.EnableUUID=True by default in all guest VM configuration file to enable udev to generate /dev/disk/by-id¨05:41
gcbIs there a link ?05:42
nandalhttp://docs.openstack.org/releasenotes/nova/newton.html05:42
*** hfu has joined #openstack-nova05:42
*** mnaser has joined #openstack-nova05:42
*** mdnadeem has quit IRC05:44
gcbnandal , just update in https://github.com/openstack/nova/blob/master/releasenotes/notes/vmware_disk_enableuuid_true-99b88e00fc168dd3.yaml .  But this is for newton one, hope the doc can be generated again05:45
*** tlian has quit IRC05:46
*** hfu has quit IRC05:47
*** hfu has joined #openstack-nova05:47
nandalgcb, thank you05:47
gcbnandal, np05:48
openstackgerritChangBo Guo(gcb) proposed openstack/nova: conf: Remove config option compute_ manager  https://review.openstack.org/40592105:49
*** pbandark has joined #openstack-nova05:50
*** penick has joined #openstack-nova05:50
openstackgerritAlex Xu proposed openstack/nova: api-ref: Complete all the verifications of remote consoles  https://review.openstack.org/33805105:50
*** mdnadeem has joined #openstack-nova05:52
*** kiran-r has joined #openstack-nova05:57
*** kristian__ has joined #openstack-nova06:01
*** int32bit has joined #openstack-nova06:01
*** GB21 has quit IRC06:01
*** kiran-r has quit IRC06:05
*** kristian__ has quit IRC06:05
*** hfu has quit IRC06:06
openstackgerritint32bit proposed openstack/nova: Page the db query when filter instances with IP  https://review.openstack.org/40517106:09
*** GB21 has joined #openstack-nova06:14
*** hfu has joined #openstack-nova06:15
*** esberglu has joined #openstack-nova06:18
*** esberglu has quit IRC06:23
*** oanson has joined #openstack-nova06:25
*** hfu has quit IRC06:27
*** sudipto_ has quit IRC06:28
*** sudipto has quit IRC06:28
*** thorst has joined #openstack-nova06:29
*** gmann has quit IRC06:30
*** hfu has joined #openstack-nova06:32
*** rcernin has joined #openstack-nova06:34
*** thorst has quit IRC06:34
*** hfu has quit IRC06:36
*** yuntongjin has joined #openstack-nova06:38
*** code-R has quit IRC06:38
*** kaisers__ has quit IRC06:43
*** hfu has joined #openstack-nova06:46
*** Cristina_ has joined #openstack-nova06:47
*** hfu has quit IRC06:48
openstackgerritHussain Chachuliya proposed openstack/nova: Deny volume swapping with non-bootable volume  https://review.openstack.org/40734606:49
*** kristian__ has joined #openstack-nova06:49
*** CristinaPauna has quit IRC06:50
*** sudipto has joined #openstack-nova06:51
*** sudipto_ has joined #openstack-nova06:52
*** hfu has joined #openstack-nova06:53
*** kristian__ has quit IRC06:54
*** code-R has joined #openstack-nova06:55
*** hfu has quit IRC06:57
*** hfu has joined #openstack-nova06:57
*** code-R has quit IRC07:00
*** markus_z has joined #openstack-nova07:04
*** mjura has joined #openstack-nova07:06
*** rcernin has quit IRC07:12
*** mjura has quit IRC07:13
*** mjura has joined #openstack-nova07:14
*** code-R has joined #openstack-nova07:14
*** kristian__ has joined #openstack-nova07:15
*** nkrinner_afk is now known as nkrinner07:18
*** andreas_s has joined #openstack-nova07:21
*** mat128 has quit IRC07:22
*** voelzmo has joined #openstack-nova07:22
*** voelzmo has quit IRC07:23
*** voelzmo has joined #openstack-nova07:23
*** vladikr has quit IRC07:23
*** mat128 has joined #openstack-nova07:23
*** shz has quit IRC07:24
openstackgerritZhenyu Zheng proposed openstack/nova: Check destination_type when booting with bdm provided  https://review.openstack.org/40237207:24
openstackgerritTakashi NATSUME proposed openstack/nova-specs: Fix formatting errors and warnings  https://review.openstack.org/40736407:25
*** mordred has quit IRC07:25
*** bhagyashris has joined #openstack-nova07:27
*** Guest66666 has quit IRC07:27
*** kaisers_ has joined #openstack-nova07:28
*** cardeois has quit IRC07:29
*** Guest66666 has joined #openstack-nova07:29
*** alexpilo_ has joined #openstack-nova07:29
*** takashin has left #openstack-nova07:29
*** alexpilotti has quit IRC07:30
*** mnaser has quit IRC07:30
*** mrsoul has quit IRC07:32
*** kristia__ has joined #openstack-nova07:32
*** thorst has joined #openstack-nova07:32
*** mordred has joined #openstack-nova07:34
*** kristi___ has joined #openstack-nova07:34
*** rcernin has joined #openstack-nova07:34
*** cardeois has joined #openstack-nova07:34
*** mrsoul has joined #openstack-nova07:35
*** kristian__ has quit IRC07:35
*** Jeffrey4l has quit IRC07:36
*** kaisers_ has quit IRC07:36
*** vladikr has joined #openstack-nova07:37
*** kristia__ has quit IRC07:38
*** kristi___ has quit IRC07:39
*** GB21 has quit IRC07:39
*** thorst has quit IRC07:40
*** pbandark has quit IRC07:41
*** pbandark has joined #openstack-nova07:41
*** pcaruana has joined #openstack-nova07:42
*** pbandark has quit IRC07:42
*** kristian__ has joined #openstack-nova07:42
*** shangxdy has quit IRC07:44
openstackgerritZhenyu Zheng proposed openstack/python-novaclient: Check source_type and destination_type when booting with bdm provided  https://review.openstack.org/40236407:44
*** yamahata has quit IRC07:44
*** pbandark has joined #openstack-nova07:46
*** kristian__ has quit IRC07:47
*** code-R has quit IRC07:47
markus_zpkoniszewski: I'm not sure how to resolve your -1 on https://review.openstack.org/#/c/347471/ Are the backports to newton a pre-condition?07:48
*** hfu has quit IRC07:48
*** hfu has joined #openstack-nova07:49
*** mnaser has joined #openstack-nova07:50
*** tuanluong has joined #openstack-nova07:51
*** moshele has joined #openstack-nova07:51
*** Jeffrey4l has joined #openstack-nova07:52
*** hfu has quit IRC07:53
*** fragatina has quit IRC07:53
*** hfu has joined #openstack-nova07:53
*** fragatina has joined #openstack-nova07:54
openstackgerritint32bit proposed openstack/nova: Page the db query when filter instances with IP  https://review.openstack.org/40517107:55
*** fragatina has quit IRC07:59
*** b_rake has joined #openstack-nova07:59
*** b_rake has quit IRC07:59
*** b_rake has joined #openstack-nova08:00
*** mvk has quit IRC08:01
openstackgerritZhenyu Zheng proposed openstack/nova: Don't delete neutron port when attach failed  https://review.openstack.org/40356808:02
*** GB21 has joined #openstack-nova08:02
*** matrohon has joined #openstack-nova08:02
*** ihrachys has joined #openstack-nova08:02
openstackgerritZhenyu Zheng proposed openstack/nova: Remove mox in tests/unit/objects/test_security_group  https://review.openstack.org/30713508:03
openstackgerritZhenyu Zheng proposed openstack/nova: Remove mox in tests/unit/objects/test_instance_info_cache  https://review.openstack.org/30715508:03
openstackgerritZhenyu Zheng proposed openstack/nova: Remove mox in tests/unit/objects/test_compute_node  https://review.openstack.org/30762508:03
*** FL1SK has quit IRC08:04
*** b_rake has quit IRC08:04
bauzasp!spy08:07
*** shaohe_feng has joined #openstack-nova08:07
*** nandal has quit IRC08:08
openstackgerritHuan Xie proposed openstack/nova: XenAPI Use os-xenapi lib for nova  https://review.openstack.org/40605908:08
*** gszasz has joined #openstack-nova08:11
*** lpetrut has joined #openstack-nova08:13
*** brault has joined #openstack-nova08:13
*** mnaser has quit IRC08:15
*** slaweq_ has quit IRC08:17
*** shaohe_feng has quit IRC08:17
*** shangxdy has joined #openstack-nova08:17
*** shaohe_feng has joined #openstack-nova08:18
*** ralonsoh has joined #openstack-nova08:18
*** code-R has joined #openstack-nova08:18
*** ljxiash has joined #openstack-nova08:20
*** code-R_ has joined #openstack-nova08:21
*** yuntongjin has quit IRC08:21
*** slaweq_ has joined #openstack-nova08:22
*** code-R has quit IRC08:24
openstackgerritwanghao proposed openstack/nova: Set min_disk in the image meta based on the root volume's size  https://review.openstack.org/40739708:24
*** ljxiash has quit IRC08:24
*** ccard_ has quit IRC08:25
*** shaohe_feng has quit IRC08:27
*** tuanluong has quit IRC08:29
*** shaohe_feng has joined #openstack-nova08:30
openstackgerritPavel Kholkin proposed openstack/nova: [proxy-api] microversion 2.39 deprecates image-metadata proxy API  https://review.openstack.org/40546108:32
pkholkinalex_xu: hello and thanks a lot for review! I've added two notes, please take a look https://review.openstack.org/#/c/40546108:32
*** mvk has joined #openstack-nova08:33
*** int32bit has quit IRC08:35
*** mnaser has joined #openstack-nova08:35
alex_xupkholkin: thanks, will check it asap08:35
*** shaohe_feng has quit IRC08:37
*** thorst has joined #openstack-nova08:38
*** shaohe_feng has joined #openstack-nova08:39
openstackgerritChangBo Guo(gcb) proposed openstack/nova: conf: Remove config option compute_ manager  https://review.openstack.org/40592108:42
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: libvirt: create consoles in an understandable/extensible way  https://review.openstack.org/40712008:42
openstackgerritMaho Koshiya proposed openstack/nova: Add metadata functional negative tests  https://review.openstack.org/39174408:43
openstackgerritGábor Antal proposed openstack/python-novaclient: Use more specific asserts in tests  https://review.openstack.org/36566708:43
*** zenoway has joined #openstack-nova08:43
*** thorst has quit IRC08:44
*** mkoderer has joined #openstack-nova08:46
*** diga has joined #openstack-nova08:46
*** zhangguoqing has quit IRC08:47
*** shaohe_feng has quit IRC08:47
*** shaohe_feng has joined #openstack-nova08:50
*** andrey-mp has joined #openstack-nova08:50
openstackgerrittianmaofu proposed openstack/nova: Unity the comparison of hw_qemu_guest_agent  https://review.openstack.org/40740908:51
*** mdnadeem has quit IRC08:52
*** ociuhandu has quit IRC08:53
openstackgerritZhenyu Zheng proposed openstack/nova: Don't apply multi-queue to SRIOV ports  https://review.openstack.org/39754508:54
openstackgerritMoshe Levi proposed openstack/os-vif: [WIP] add OVS acceleration  https://review.openstack.org/39827708:55
*** shaohe_feng has quit IRC08:56
*** tuanluong has joined #openstack-nova08:57
*** shaohe_feng has joined #openstack-nova08:58
*** zzzeek has quit IRC09:00
*** zzzeek has joined #openstack-nova09:00
*** roeyc has joined #openstack-nova09:02
*** psachin has quit IRC09:04
*** mdnadeem has joined #openstack-nova09:04
*** GB21 has quit IRC09:05
*** ociuhandu has joined #openstack-nova09:08
*** larainema has quit IRC09:14
*** larainema_ has joined #openstack-nova09:14
*** gmann has joined #openstack-nova09:15
openstackgerritOleg Bondarev proposed openstack/nova: Live migration: wait for vif-plugged event on pre live migration  https://review.openstack.org/24691009:15
*** psachin has joined #openstack-nova09:17
openstackgerritZhenyu Zheng proposed openstack/nova: Check destination_type when booting with bdm provided  https://review.openstack.org/40237209:18
*** GB21 has joined #openstack-nova09:18
*** ftersin has joined #openstack-nova09:21
*** thorst has joined #openstack-nova09:24
*** acabot_ has joined #openstack-nova09:27
*** rmart04 has joined #openstack-nova09:27
*** kristian__ has joined #openstack-nova09:27
*** kristian__ has joined #openstack-nova09:27
*** sreeram_v has quit IRC09:29
*** sreeram_v has joined #openstack-nova09:30
*** thorst has quit IRC09:31
*** sreeram_v has quit IRC09:34
*** sreeram_v has joined #openstack-nova09:34
*** yassine has joined #openstack-nova09:34
*** paul-carlton1 has joined #openstack-nova09:34
*** yassine is now known as Guest1745009:35
*** daemontool has joined #openstack-nova09:35
*** lucas-afk is now known as lucasagomes09:35
*** paul-carlton2 has joined #openstack-nova09:35
*** paul-carlton2 has quit IRC09:37
*** paul-carlton1 has quit IRC09:37
*** paul-carlton2 has joined #openstack-nova09:38
*** acabot__ has joined #openstack-nova09:40
*** yingjun has quit IRC09:40
*** derekh has joined #openstack-nova09:40
*** kristian__ has quit IRC09:40
gibijohnthetubaguy, bauzas: hi! there is a pretty simple test-only change that has +2 from jaypipes already. Could you check it please? https://review.openstack.org/#/c/407033/09:42
bauzasgibi: oh yeah, saw the change09:45
bauzasgibi: +Wed09:45
gibibauzas: thanks a lot!09:46
*** sudipto_ has quit IRC09:47
*** sudipto has quit IRC09:47
openstackgerritTimofey Durakov proposed openstack/nova: fix for auth during live-migration  https://review.openstack.org/40714709:53
pkoniszewskimarkus_z: can we conditionally start those serial console tests? I mean, can we add a condition so that if this is a grenade job and branch is stable/newton it won't start serial console tests?09:54
*** gmann has quit IRC09:58
*** gmann has joined #openstack-nova09:58
*** huhaoran has joined #openstack-nova10:01
*** acabot_ has quit IRC10:03
*** acabot__ is now known as acabot_10:03
*** voelzmo has quit IRC10:03
*** FL1SK has joined #openstack-nova10:05
*** nkrinner has quit IRC10:05
*** nkrinner has joined #openstack-nova10:06
*** kristian__ has joined #openstack-nova10:06
*** sahid has joined #openstack-nova10:06
*** tovin07 has quit IRC10:09
*** kristian__ has quit IRC10:11
*** voelzmo has joined #openstack-nova10:11
openstackgerritFeodor Tersin proposed openstack/nova: Implement ScaleIO image backend  https://review.openstack.org/40744010:11
*** shangxdy has quit IRC10:13
*** oanson has quit IRC10:15
*** eandersson has joined #openstack-nova10:15
*** nkrinner has quit IRC10:15
*** aweeks has quit IRC10:16
*** shaohe_feng has quit IRC10:17
*** nkrinner has joined #openstack-nova10:17
*** shaohe_feng has joined #openstack-nova10:17
*** esberglu has joined #openstack-nova10:20
*** sam_nowitzki has quit IRC10:23
*** daemontool has quit IRC10:23
*** baoli has joined #openstack-nova10:24
*** JoseMello has joined #openstack-nova10:24
*** esberglu has quit IRC10:24
*** thorst has joined #openstack-nova10:28
*** baoli has quit IRC10:28
*** cdent has joined #openstack-nova10:29
*** kristian__ has joined #openstack-nova10:30
*** voelzmo has quit IRC10:30
openstackgerritChris Dent proposed openstack/nova: Update cors-to-versions-pipeline release note  https://review.openstack.org/40702110:31
*** voelzmo has joined #openstack-nova10:32
*** kristia__ has joined #openstack-nova10:32
*** kristia__ has quit IRC10:33
*** kristia__ has joined #openstack-nova10:34
*** kristian__ has quit IRC10:34
*** kristia__ has quit IRC10:35
*** kristian__ has joined #openstack-nova10:35
*** thorst has quit IRC10:36
*** voelzmo has quit IRC10:36
*** kristia__ has joined #openstack-nova10:40
*** kristian__ has quit IRC10:40
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: libvirt: virtlogd: use virtlogd for char devices  https://review.openstack.org/40745010:41
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: libvirt: virtlogd: use virtlogd for char devices  https://review.openstack.org/40745010:42
*** ducnc has quit IRC10:43
*** sreeram_v has quit IRC10:43
*** sreeram_v has joined #openstack-nova10:44
markus_zpkoniszewski: good question, I'll investigate that and let you know at the l-m meeting today.10:45
*** huhaoran has quit IRC10:46
sfinucan:q10:47
*** huhaoran has joined #openstack-nova10:47
*** nandal has joined #openstack-nova10:49
*** rfolco has joined #openstack-nova10:49
*** kevinz has quit IRC10:50
*** huhaoran has quit IRC10:51
*** huhaoran has joined #openstack-nova10:51
*** tbachman has joined #openstack-nova10:53
*** hfu has quit IRC10:58
*** kristian__ has joined #openstack-nova10:59
*** kristia__ has quit IRC11:02
*** sdague has joined #openstack-nova11:02
*** kristia__ has joined #openstack-nova11:04
*** yingjun has joined #openstack-nova11:05
*** kristian__ has quit IRC11:08
*** zhangjl has quit IRC11:11
*** daemontool has joined #openstack-nova11:11
openstackgerritsahid proposed openstack/nova: objects: add new field cpu_emulator_threads_policy  https://review.openstack.org/38502211:12
openstackgerritsahid proposed openstack/nova: numa: add numa constraints for emulator threads policy  https://review.openstack.org/38502311:12
openstackgerritsahid proposed openstack/nova: libvirt: return a CPU overhead if isolate emulator threads requested  https://review.openstack.org/38536411:12
openstackgerritsahid proposed openstack/nova: objects: add new field cpus_overhead in NUMACell  https://review.openstack.org/40087411:12
openstackgerritsahid proposed openstack/nova: numa: fit instance toplogy with cpu overhead onto cores  https://review.openstack.org/40087511:12
openstackgerritsahid proposed openstack/nova: virt: update overhead to take into account vCPUs  https://review.openstack.org/38502411:12
openstackgerritsahid proposed openstack/nova: libvirt: make emulator threads to run on the reserved pCPU  https://review.openstack.org/40087811:12
openstackgerritsahid proposed openstack/nova: numa: take into account cpus overhead  https://review.openstack.org/40087611:12
openstackgerritsahid proposed openstack/nova: numa: update numa usage incording the cpu overhead  https://review.openstack.org/40087711:12
*** dimtruck is now known as zz_dimtruck11:13
*** owalsh has joined #openstack-nova11:13
*** yingjun has quit IRC11:13
*** hshiina has joined #openstack-nova11:15
*** claudiub|2 has joined #openstack-nova11:22
*** kristia__ has quit IRC11:23
*** kevinz has joined #openstack-nova11:23
*** andrey-mp has left #openstack-nova11:23
*** kristian__ has joined #openstack-nova11:25
*** sreeram_v has quit IRC11:25
*** kristia__ has joined #openstack-nova11:28
*** kristian__ has quit IRC11:28
*** kristia__ has quit IRC11:29
*** yingjun has joined #openstack-nova11:30
openstackgerritPawel Koniszewski proposed openstack/nova: live-mig: Add claims and proper resource tracking  https://review.openstack.org/24448911:31
*** thorst has joined #openstack-nova11:34
*** voelzmo has joined #openstack-nova11:36
*** yingjun has quit IRC11:36
*** alexpilo_ has quit IRC11:36
*** yingjun has joined #openstack-nova11:36
*** alexpilotti has joined #openstack-nova11:37
*** mdnadeem has quit IRC11:37
*** ociuhandu has quit IRC11:37
*** nicolasbock has joined #openstack-nova11:38
*** GB21 has quit IRC11:38
openstackgerritMatthew Booth proposed openstack/nova: libvirt: Add comments in _hard_reboot  https://review.openstack.org/40747011:39
*** ihrachys has quit IRC11:39
mdboothjohnthetubaguy: ^^^ Would take approximately 2 seconds of your time :) It just adds a couple of comments.11:40
*** voelzmo has quit IRC11:40
*** thorst has quit IRC11:41
*** yingjun has quit IRC11:41
*** alexpilotti has quit IRC11:41
*** lpetrut has quit IRC11:44
*** jamielennox is now known as jamielennox|away11:44
*** prateek_ has quit IRC11:45
*** cdelatte has joined #openstack-nova11:46
markus_zsfinucan: congrats on the new role! You've earned it.11:46
sfinucanmarkus_z: Well it's not confirmed yet, but thanks :)11:46
*** alexpilotti has joined #openstack-nova11:47
mdboothsfinucan: woot, congrats!11:47
*** ratailor has quit IRC11:48
*** mdnadeem has joined #openstack-nova11:49
*** tbachman has quit IRC11:49
*** nkrinner has quit IRC11:50
sfinucanmdbooth: Cheers :)11:51
*** alexpilotti has quit IRC11:51
*** GB21 has joined #openstack-nova11:55
*** voelzmo has joined #openstack-nova11:56
*** nkrinner has joined #openstack-nova11:58
*** smatzek has joined #openstack-nova12:00
*** nkrinner has quit IRC12:06
*** nkrinner has joined #openstack-nova12:06
openstackgerritFeodor Tersin proposed openstack/nova: Implement ScaleIO image backend  https://review.openstack.org/40744012:07
*** GB21 has quit IRC12:07
*** tuanluong has quit IRC12:08
ftersinmdbooth: hi. have you seen my comments on https://review.openstack.org/#/c/333244 and the next patch?12:09
*** yingjun has joined #openstack-nova12:10
*** portdirect__ has joined #openstack-nova12:10
*** hfu has joined #openstack-nova12:11
*** lpetrut has joined #openstack-nova12:11
*** GB21 has joined #openstack-nova12:16
*** kristian__ has joined #openstack-nova12:16
*** kristian__ has quit IRC12:17
*** kristian__ has joined #openstack-nova12:17
*** rtheis has joined #openstack-nova12:17
*** esberglu has joined #openstack-nova12:21
*** kristian__ has quit IRC12:22
*** kristian__ has joined #openstack-nova12:23
*** ihrachys has joined #openstack-nova12:24
*** shangxdy has joined #openstack-nova12:25
*** esberglu has quit IRC12:25
*** bvanhav has joined #openstack-nova12:27
*** lpetrut has quit IRC12:27
*** kristian__ has quit IRC12:28
*** thorst has joined #openstack-nova12:30
*** code-R_ has quit IRC12:30
*** lucasagomes is now known as lucas-hungry12:30
*** databus23_ has joined #openstack-nova12:31
*** dgonzalez_ has joined #openstack-nova12:31
*** tommy_ has joined #openstack-nova12:31
*** david_1 has joined #openstack-nova12:31
*** tpatzig_ has joined #openstack-nova12:31
*** code-R has joined #openstack-nova12:31
*** dgonzalez_ has quit IRC12:31
*** sapcc-bot has quit IRC12:31
*** databus23_ has quit IRC12:31
*** tommy_ has quit IRC12:31
*** tpatzig_ has quit IRC12:31
*** carthaca_1 has quit IRC12:31
*** david_1 has quit IRC12:31
*** sreeram_v has joined #openstack-nova12:35
sdaguemarkus_z: once jaypipes gets up, we should get him to look at the alternative virtlogd stack12:37
markus_zsdague: yep, will keep that in mind and ping him later12:37
sdaguebecause it feels a lot easier to understand in the final code (even though that first patch is a bit complex to get there)12:37
sdagueI +2ed the alternative stack12:37
cdentbauzas, jaypipes, edleafe: could one of you sanity check/triage this bug please if you have a chance: https://bugs.launchpad.net/nova/+bug/1647697 ? I discovered it might be the case while editing the grp spec and don't want to lose my motivation on that.12:38
openstackLaunchpad bug 1647697 in OpenStack Compute (nova) "When deleting a resource provider in the placement api the rp's associated aggregates are not cleaned up" [Undecided,New]12:38
*** hferenc has joined #openstack-nova12:38
openstackgerritsahid proposed openstack/nova: libvirt: return a CPU overhead if isolate emulator threads requested  https://review.openstack.org/38536412:38
openstackgerritsahid proposed openstack/nova: libvirt: make emulator threads to run on the reserved pCPU  https://review.openstack.org/40087812:38
markus_zsdague: Thanks a lot. FWIW, if it weren't for the deadline next week I would have done it differently.12:39
sdaguemarkus_z: yep, no worries12:39
*** bkopilov has quit IRC12:39
*** sreeram_v has quit IRC12:40
*** ociuhandu has joined #openstack-nova12:41
openstackgerritzhaolihui proposed openstack/nova: Fix keypair quota error  https://review.openstack.org/40352912:44
*** larainema_ has quit IRC12:46
*** larainema_ has joined #openstack-nova12:46
*** Kevin_Zheng has quit IRC12:46
*** Kevin_Zheng has joined #openstack-nova12:47
*** cdelatte has quit IRC12:51
*** GB21 has quit IRC12:54
*** RuiChen has quit IRC12:56
*** belmoreira has joined #openstack-nova12:56
*** RuiChen has joined #openstack-nova12:57
bauzascdent: looks correct12:57
bauzasI mean, looks reasonable12:57
cdentthanks bauzas12:57
*** cdelatte has joined #openstack-nova12:58
openstackgerritjichenjc proposed openstack/nova: Remove AdminRequired usage in flavor  https://review.openstack.org/40658012:58
*** ayogi has quit IRC12:59
*** nkrinner has quit IRC13:01
*** nkrinner has joined #openstack-nova13:02
openstackgerritOleg Bondarev proposed openstack/nova: Live migration: wait for vif-plugged event on pre live migration  https://review.openstack.org/24691013:02
*** kristian__ has joined #openstack-nova13:04
openstackgerritjichenjc proposed openstack/nova: Add more log when delete orphan node  https://review.openstack.org/40127413:06
*** tbachman has joined #openstack-nova13:06
openstackgerritjichenjc proposed openstack/nova: Refactored query user_id code for quota-sets  https://review.openstack.org/28016513:06
jaypipessdague: I'm up! :)13:07
*** hfu has quit IRC13:07
jaypipessdague: link for me?13:07
openstackgerritHussain Chachuliya proposed openstack/nova: Deny volume swapping with non-bootable volume  https://review.openstack.org/40734613:08
*** kristian__ has quit IRC13:08
jaypipescdent: yup, that's def a bug13:08
* cdent nods at jaypipes 13:08
cdentthanks for confirming13:08
*** Amit_M has quit IRC13:09
sdaguejaypipes: https://review.openstack.org/#/c/407450/13:11
*** jheroux has joined #openstack-nova13:11
*** tbachman has quit IRC13:11
sdagueI asked markus_z to restructure the virtlogd patch so it's easier to understand in final form (though the 2 patches to get there are a little more confusing because the base code is so odd)13:12
sdagueyou have a +2 on the other patch, so wanted to see if you are up for this one instead13:12
*** gouthamr has joined #openstack-nova13:12
*** tbachman has joined #openstack-nova13:12
*** edmondsw_ has joined #openstack-nova13:13
*** fragatina has joined #openstack-nova13:15
*** fragatin_ has joined #openstack-nova13:15
wznoinsksfinucan, congrats13:16
*** edmondsw_ has quit IRC13:17
openstackgerritjichenjc proposed openstack/nova: Enhancement comments on CountableResource  https://review.openstack.org/40205113:18
*** david_1 has joined #openstack-nova13:18
*** sapcc-bot has joined #openstack-nova13:18
*** tpatzig_ has joined #openstack-nova13:18
*** dgonzalez_ has joined #openstack-nova13:18
*** databus23_ has joined #openstack-nova13:18
*** carthaca_ has joined #openstack-nova13:18
*** fragatina has quit IRC13:19
*** dave-mccowan has joined #openstack-nova13:19
*** lamt has joined #openstack-nova13:19
mdboothftersin: I haven't read them yet because I'm still working on something else, but I will definitely address them before re-pushing.13:20
*** tpatzig_ has quit IRC13:20
*** databus23_ has quit IRC13:20
*** dgonzalez_ has quit IRC13:20
*** carthaca_ has quit IRC13:20
*** david_1 has quit IRC13:20
*** shangxdy has quit IRC13:21
mdboothftersin: Ah, yes. I have read those, actually. I thought I'd responded, but apparently it got lost. I'm going to have to look at that closely.13:22
*** pradk has joined #openstack-nova13:22
*** hfu has joined #openstack-nova13:22
mdboothI have a sneaking suspicion that fallback is there because we start using it when we ditch the legacy cache() implementation.13:23
mdboothBut I'll look in detail and update it if necessary before I re-push.13:24
*** huhaoran has quit IRC13:25
*** yingjun has quit IRC13:25
* mdbooth was trying to avoid pulling in the changes to remove create_images_and_backing in this series, but I realised I couldn't.13:25
mdboothCould I have a really quick couple of +2 on this patch which adds 4 lines of comments to the libvirt driver: https://review.openstack.org/#/c/407470/13:26
*** hamzy has quit IRC13:26
*** mdnadeem has quit IRC13:27
*** hfu has quit IRC13:29
*** hfu has joined #openstack-nova13:29
*** kristian__ has joined #openstack-nova13:32
*** nkrinner has quit IRC13:33
jaypipessdague, markus_z: done.13:33
*** code-R_ has joined #openstack-nova13:33
tdurakovmdbooth: hi, got question on that one13:34
markus_zjaypipes: Thanks a ton!13:34
tdurakovmdbooth: what is the reason we need auth_token?13:34
mdboothtdurakov: It's in the comment!13:34
*** kristian__ has quit IRC13:34
mdboothAh, you mean what's it subsequently used for?13:35
*** kristian__ has joined #openstack-nova13:35
tdurakovyes13:35
mdboothFetching from glance13:35
tdurakovmdbooth: could you please point me to?13:35
mdboothI could draw you a map13:35
tdurakovlol13:35
mdboothHowever, it would take me a really long time :)13:35
markus_zjaypipes: If you could check https://review.openstack.org/#/c/407120/ too, then bp/libvirt-virtlogd would be done. :)13:35
mdboothtdurakov: Sec...13:36
*** nkrinner has joined #openstack-nova13:36
*** code-R has quit IRC13:36
tdurakovmdbooth: it sounds similar to  https://review.openstack.org/#/c/407147/ so, I wonder could we fix nova glance communication instead13:37
*** acabot__ has joined #openstack-nova13:37
mdboothtdurakov: Have a look in _create_images_and_backing13:37
*** hfu has quit IRC13:38
mdboothIf it has a local root disk, we fetch its backing file from glance13:38
mdbooth_try_fetch_image_cache13:38
mdboothThat needs auth_token13:38
mdboothBut as this code isn't strictly necessary at all in _hard_reboot, it doesn't really matter if we don't do it13:39
mdboothHence the if is ok13:39
mdboothBut.... my patch just explains why auth_token might not be set, and therefore why that if statement is there13:39
* mdbooth had to dig that out of git, and having done that I thought it would be useful to put the information in a comment instead.13:40
*** nkrinner has quit IRC13:43
*** catintheroof has joined #openstack-nova13:43
tdurakovpkoniszewski: hi, are you around?13:43
pkoniszewskitdurakov: yes13:43
*** swamireddy has quit IRC13:43
tdurakovpkoniszewski: don't we just wait for backports?13:43
tdurakovfor serial console hook13:44
pkoniszewskithat's an option13:44
*** tlian has joined #openstack-nova13:44
markus_zpkoniszewski: tdurakov: I would show my proposals on that during the l-m meeting. jroll pointed me to a promising direction: http://eavesdrop.openstack.org/irclogs/%23openstack-qa/%23openstack-qa.2016-12-06.log.html#t2016-12-06T12:44:1613:46
*** huhaoran has joined #openstack-nova13:47
openstackgerritFeodor Tersin proposed openstack/nova: Implement ScaleIO image backend  https://review.openstack.org/40744013:48
openstackgerritjichenjc proposed openstack/nova: Catch ProjectQuotaNotFound when do update_limit  https://review.openstack.org/28815313:48
tdurakovpkoniszewski: for your proposal, we could check BRANCH_OVERRIDE  DEVSTACK_GATE_GRENADE variables13:48
openstackgerritFeodor Tersin proposed openstack/nova: Implement ScaleIO image backend  https://review.openstack.org/40744013:49
tdurakovmarkus_z: acked, let's discuss it on a meeting13:50
*** lucas-hungry is now known as lucasagomes13:50
*** palexster has quit IRC13:51
*** bvanhav_ has joined #openstack-nova13:51
tdurakovmarkus_z: https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/devstack-gate.yaml#L145313:51
tdurakovit won't work13:51
tdurakovwe explicilty set it to false for live-migration grenade job13:52
*** fragatin_ has quit IRC13:52
*** bvanhav has quit IRC13:52
markus_ztdurakov: no worries, I have a list of proposal which don't have a dependency to the smoke tests.13:52
*** fragatina has joined #openstack-nova13:52
tdurakovmarkus_z: ok)13:52
*** hfu has joined #openstack-nova13:53
*** nkrinner has joined #openstack-nova13:55
*** baoli has joined #openstack-nova13:56
openstackgerritHussain Chachuliya proposed openstack/nova: Fix 500 error while passing 4-byte unicode data  https://review.openstack.org/40751413:57
*** yingjun has joined #openstack-nova13:57
*** hfu has quit IRC13:58
*** baoli has quit IRC13:58
*** baoli has joined #openstack-nova13:58
*** swamireddy has joined #openstack-nova13:58
*** esberglu has joined #openstack-nova13:58
*** baoli has quit IRC13:59
tdurakovPaulMurray paul-carlton2 eliqiao shaohe_feng pkoniszewski yuntongjin mdbooth kashyap kmartin claudiub tobasco andreas_s abhishekk mrhillsman raj_singh wznoinsk siva_krishnan live migration meeting in 2 minutes on #openstack-meeting-313:59
*** baoli has joined #openstack-nova13:59
*** palexster has joined #openstack-nova13:59
openstackgerritPawel Koniszewski proposed openstack/nova: Move migration_downtime_steps to libvirt/migration  https://review.openstack.org/40751714:00
openstackgerritPawel Koniszewski proposed openstack/nova: Change exponential function to linear  https://review.openstack.org/40751814:00
*** prateek has joined #openstack-nova14:00
*** lamt has quit IRC14:01
*** lamt has joined #openstack-nova14:02
*** Guest16197 is now known as mfisch14:03
*** mfisch is now known as Guest7303714:04
*** daemontool has quit IRC14:04
*** mdrabe has joined #openstack-nova14:07
*** hfu has joined #openstack-nova14:08
*** oanson has joined #openstack-nova14:08
*** prateek has quit IRC14:09
*** ihrachys has quit IRC14:10
*** READ10 has joined #openstack-nova14:13
*** ihrachys has joined #openstack-nova14:13
*** ducttape_ has quit IRC14:15
*** Guest73037 is now known as mfisch14:15
*** prateek has joined #openstack-nova14:15
*** mfisch is now known as Guest1328114:16
*** Guest13281 is now known as mfisch14:16
*** mriedem has joined #openstack-nova14:16
*** mfisch is now known as Guest4066414:17
*** Guest40664 is now known as mfisch14:17
*** mfisch has quit IRC14:17
*** mfisch has joined #openstack-nova14:17
*** swamireddy has quit IRC14:17
diana_clarkemriedem, andreykurilin: It took me a while, but I'm pretty certain the python-novaclient gate failures are cells v2 related. If I revert the devstack change "Always setup cells v2", the gate is green again.14:19
andreykurilinhm14:19
mriedemdiana_clarke: yeah i suspected that, i think the marker stuff might be messed up14:19
mriedemi posted some patches to debug it, still need to get into those14:19
mriedemhttps://review.openstack.org/#/c/407204/14:20
diana_clarkemriedem: yeah, I was following along looking at similar stuff at the same time14:20
mriedemoops actually https://review.openstack.org/#/c/407205/14:20
jaypipesmarkus_z: reviewed https://review.openstack.org/#/c/407120. +0... I have a question in there.14:21
openstackgerritTimofey Durakov proposed openstack/nova: dnm to identify grenade job  https://review.openstack.org/40753214:21
markus_zjaypipes: ok, will answer in a minute14:22
*** tbachman has quit IRC14:22
*** tbachman_ has joined #openstack-nova14:22
*** jamesdenton has joined #openstack-nova14:24
openstackgerritjichenjc proposed openstack/nova: Add description for 2.9 microversion  https://review.openstack.org/37148114:25
markus_zjaypipes: crap, that was a mistake, I'll push a new ps. thanks for spotting this!14:25
mriedemjaypipes: while you're feeling saucy, can you also get this? https://review.openstack.org/#/c/402097/ it's the bottom change on the vhostuser series, which you were +W on the top14:27
*** hfu has quit IRC14:28
*** hfu has joined #openstack-nova14:29
*** tblakes has joined #openstack-nova14:30
*** hshiina has quit IRC14:30
mriedemdiana_clarke: hmm, so here is the marker not found with extra debug http://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/70528ca/console.html#_2016-12-06_00_42_18_93836814:31
cdentsfinucan: woot14:31
mriedemERROR (BadRequest): marker [dc89a764-5d12-4d17-8379-ccd110c81335] not found; current instance UUID: ['f4cc5ce3-cfcb-40bd-981c-6b1b4901c697', '6b2aae34-0308-4b92-af93-dd26d3569760', 'dc89a764-5d12-4d17-8379-ccd110c81335'] (HTTP 400) (Request-ID: req-2a81fd48-517b-4577-885a-0ad5a6dc8dec)14:31
mriedemthe marker uuid is in the list of available instances14:31
diana_clarkemriedem: yup14:31
bauzascdent: quick question, can I GET /RPs?uuid=<foo>&uuid=<bar> ?14:31
bauzascdent: per the code, it seems no14:31
cdentbauzas: looking, one sec14:31
diana_clarkemriedem: strange, huh. perhaps a cells canary that something isn't quite right...14:32
*** swamireddy has joined #openstack-nova14:32
jaypipesmriedem: done.14:32
mriedemdiana_clarke: at first i suspected something was up with a race to query the instances table before the instance existed, but the build request would have existed, but this test is waiting for all 3 instances to be ACTIVE before it does the query14:32
mriedemjaypipes: thanks14:32
mriedemso if the instances are ACTIVE then they are in the DB14:33
jaypipesmarkus_z: wait... so I spotted something sdague missed?! today is a good day.14:33
markus_zjaypipes: I'm shocked too ;P14:33
cdentbauzas: unclear, looking at master, it depends on the behavior of sd.query.filter(foo == bar) when bar is a list14:33
*** smatzek has quit IRC14:33
jaypipesmarkus_z: :)14:33
bauzascdent: well, I'm looking at the API14:34
sfinucancdent: I feel like I should have prepared a speech or something14:34
bauzascdent: given we're looping over a list of allowed filter keys14:34
jaypipessfinucan: lol14:34
jaypipessfinucan: just need to get your casual friday nick in order.14:34
sfinucanjaypipes: heh. Working on it14:35
cdentbauzas: it also depends on whether req.GET['uuid'] returns a scalar or a list14:35
cdentbauzas: looks like it needs a test14:35
mriedemdiana_clarke: the other wrinkle is that all 3 instances have the same name, and we're filtering on name14:35
*** hfu has quit IRC14:35
bauzascdent: ah good point, we don't check the scalar14:36
bauzascdent: I was supposing a scalar for that key, not a list14:36
mriedemso filtering on name and passing a marker14:36
cdentbauzas: the uuidutils.is_uuid_like call may cause it to bail out14:36
bauzascdent: like uuid=foo,bar14:36
bauzasyeah :(14:36
cdentso seems a few different bugs there?14:36
cdentor at least unexpecte behaviors14:37
bauzasnot really, just that I'd love to see a way to ask for more than one uuid14:37
bauzasthat's not a bug honestly14:37
*** tbachman_ has quit IRC14:37
bauzasor the attribute could be named uuid*s*14:37
cdentwell I agree that it should be possible to do uuid=something&uuid=somethingelse14:37
cdentthat would be the normal way to do a list of the same query14:37
bauzasthat's what I'd prefer than just a uuid=<list>14:38
bauzasexactly14:38
*** hfu has joined #openstack-nova14:38
cdentit should be doable, and I can't think of a reason why not to do it14:38
bauzasfor me, uuid is expecting a scalar, not a list but if you want more than one, then just add a new qs14:38
bauzasokay14:38
tdurakovpkoniszewski: for grenade job we tests only newton and master compatibility, right?14:39
cdentbauzas: in related news, jaypipes remembered in the last 24 hours that we need a similar filter on the same url for aggregates14:39
pkoniszewskitdurakov: it also tests mitaka<->newton14:40
cdentone that states: resource providers that are in these aggregates (is that <- the proper translation jaypipes ? )14:40
jaypipesbauzas: right, otherwise we don't have a way of finding providers of shared resources.14:40
mriedem2016-12-06 00:42:17.880 27703 DEBUG nova.compute.api [req-2a81fd48-517b-4577-885a-0ad5a6dc8dec admin admin] Searching by: {'deleted': False, 'project_id': u'2ba71d0343c74b2aa8b70fcead4ad407', u'name': u'365f4533-46fd-4879-b75b-0d34c6d66585'} get_all /opt/stack/new/nova/nova/compute/api.py:231514:40
*** hfu has quit IRC14:40
*** links has quit IRC14:40
pkoniszewskitdurakov: i mean, it is supposed to test this, but it fails due to wrong version of a script14:40
bauzasjaypipes: cdent: I'm about to propose my REST API patch very soon, I'm just blocked by the above14:40
bauzasI hope my proposal will be loved by all of you, guys14:41
bauzaskind of a trade-off :)14:41
tdurakovpkoniszewski: so, right now, I'd prefer to skip serial console for all grenade, not just for newton -> master14:41
jaypipesbauzas: uuid=<uuid1>&uuid=<uuid2> is perfect. go with that.14:41
diana_clarkemriedem: It's failing before the sorting and filtering though. It's unable to do a simple get by id here, I think: https://github.com/openstack/nova/blob/f61db221f31d9ba86f61c13a7d1c5a951654fdc0/nova/db/sqlalchemy/api.py#L224014:41
*** ducttape_ has joined #openstack-nova14:41
bauzasjaypipes: okay, I'll put in the same change for reviews, but I could split that14:41
tdurakovpkoniszewski: smth like http://xsnippet.org/362141/ I've prepared patch, but want to be sure that it will work,  so wating for https://review.openstack.org/#/c/407532/14:42
mdboothPeeps, I use this kind of pattern in a bunch of places, and mostly it works really well:14:42
mdbooth        for name in ('kernel', 'ramdisk', 'disk', 'disk.eph0', 'disk.swap'):14:42
mdbooth            policyfixture.disks[name].create.assert_called_once_with()14:42
pkoniszewskitdurakov: but will this change affect mitaka<->newton tests? I mean, we would need to backport hooks, wouldn't we?14:42
mdboothProblem is, when it fails you've got to annotate it to work out what it failed on14:42
mdboothDoes anybody have a better pattern?14:43
jaypipesbauzas: whatever you want, lemme know a link when you're ready for review.14:43
*** tbachman has joined #openstack-nova14:43
*** kaisers has quit IRC14:43
bauzasjaypipes: hopefully in the next hour14:43
dansmithmdbooth: assert_has_calls14:43
dansmithvladikr: any progress on that failure yesterday?14:43
mdboothdansmith: It's multiple different objects14:43
dansmithmdbooth: oh I see yeah14:44
tdurakovpkoniszewski: will check that14:44
mdboothdansmith: Unless assert_has_calls lets you add additional output?14:44
dansmithmdbooth: no, I just didn't read the example very clearly14:44
tdurakovas for now stable branches tests only limited scenarios for live-migration job, simple block migration only14:44
lbeliveau_tdurakov: pkoniszewski: I'm getting serial console issues with devstack, is this related to the issue you are mentionning above http://paste.openstack.org/show/591544/14:45
mriedemmdbooth: you can name a mock14:45
mriedemwould that help with the failure output?14:45
mdboothmriedem: Yeah, if the name ended up in the output14:45
mriedemmock.Mock(..., name='disks.kernel')14:45
pkoniszewskilbeliveau_: no, it's unrelated to config options14:45
* mdbooth looks that up14:45
vladikrdansmith, not much, just got deeper into the mess.. on some tests , at some point the instance disappears right after one of the instance.save()14:46
mdboothmriedem: ooh. Trying it now.14:46
pkoniszewskilbeliveau_: the issue we talk about is about acquiring/releasing serial console ports14:46
mriedem"name: If the mock has a name then it will be used in the repr of the mock. This can be useful for debugging. The name is propagated to child mocks."14:46
dansmithvladikr: hmm14:46
*** burt has joined #openstack-nova14:46
vladikrdansmith, on other tests it saves an empty pci device object, while I see that the constructed devices are correct14:46
*** lpetrut has joined #openstack-nova14:47
*** kaisers has joined #openstack-nova14:47
tdurakovmriedem: hi, what is the resolution on this one https://review.openstack.org/#/c/407147/ ?14:48
mriedemtdurakov: i don't have one yet14:48
mriedemi'm busy with something else atm14:48
tdurakovmriedem: acked14:48
vladikrdansmith, also, some tests that failed in the ci are passing locally, but others are failing :/14:48
edleafebauzas: cdent: If you ran GET /RPs?uuid=<foo>&uuid=<bar> it ANDs the filters together. It should be GET /RPs?uuid=in:foo,bar14:49
dansmithvladikr: really.. just in that one test module or elsewhere?14:49
dansmithvladikr: because I've since pushed up another revision where those are the only failures14:49
vladikrdansmith, I've tested this module only14:49
cdentedleafe: excellent point, good catch14:49
bauzasedleafe: from an API perspective, duplicating the key is fine by me14:49
cdentbauzas: edleafe's point is that api-wg guidelines have it that duplicated keys are ands14:49
tdurakovpkoniszewski: are there any other issues with grenade job?14:50
cdentso to avoid that you need to use a different form14:50
* bauzas facepalms14:50
edleafebauzas: but two filters are ANDed together, so that will always return nothing14:50
pkoniszewskitdurakov: haven't spotted any other issue14:50
mdboothmriedem: Unfortunate that doesn't make it into the output of an assertion failure :(14:50
tdurakovpkoniszewski: https://review.openstack.org/#/c/407147/14:50
mdboothMeh, nm14:50
bauzasso, say I'm asking for fruit=apple&fruit=banana, it will AND ?14:50
bauzasoh man14:50
* mdbooth makes coffee14:51
mriedemmdbooth: so it worked?14:51
mriedemdiana_clarke: ok i'm going to just push up another nova change with more debug logging14:51
bauzasedleafe: is that a RFC or HTTP argument ?14:51
jaypipesbauzas: edleafe is absolutely correct. that is what the API WG placed in the guidelines. sorry, man.14:51
diana_clarkemriedem: I think it's cells & _build_instance_get query related... but I'm not yet sure how14:51
pkoniszewskitdurakov: it looks like a rabbit issue http://logs.openstack.org/47/407147/2/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/a6a9a8b/logs/subnode-2/old/screen-n-cpu.txt.gz?level=TRACE14:51
cdentbauzas: it's in here: http://specs.openstack.org/openstack/api-wg/guidelines/pagination_filter_sort.html14:51
bauzasI always expected as a customer that I'm providing two attributes with the same name, it will OR them14:51
edleafebauzas: cdent: https://github.com/openstack/api-wg/blob/master/guidelines/pagination_filter_sort.rst14:52
cdentedleafe: buy me a coke14:52
edleafe"Multiple filters result in an implicit AND, so in our example /app/items?foo=buzz&baz=quux would provide no results."14:52
* bauzas is desesperated by that pile of news14:52
tdurakovpkoniszewski: the reason I'm asking: https://review.openstack.org/#/c/400878/ https://review.openstack.org/#/c/407120/14:52
cdentbauzas: it may be worthwhile to note that the precedent for this is coming out of nova, that guideline is one of the things that was copied in from the existing apis14:53
edleafebauzas: is there any reason that you can't use in:?14:53
*** kashyap has quit IRC14:53
bauzasedleafe: that looks pretty special, but okay14:53
*** pbandark has quit IRC14:53
bauzasI can do that to respect the WG :-)14:53
jaypipesedleafe: I can understand where bauzas is coming from actually. there's really no point in the ability to say "uuid = A AND uuid = B". it's an unresolvable condition.14:54
pkoniszewskitdurakov: that's weird14:54
tdurakovjaypipes: hi, if you have some time, could you please take a look https://review.openstack.org/#/c/407147/14:54
pkoniszewskitdurakov: http://logs.openstack.org/78/400878/7/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/272a2c8/logs/subnode-2/old/screen-n-cpu.txt.gz?level=TRACE14:54
edleafebauzas: it's not respecting the WG as much as much as respecting OpenStack14:54
*** lpetrut has quit IRC14:54
bauzasI'm always having hard time to respect myself even14:55
tdurakovpkoniszewski: yeah, looks like constant issue(14:55
pkoniszewskitdurakov: neutron agent is also failing on rabbit http://logs.openstack.org/78/400878/7/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/272a2c8/logs/subnode-2/old/screen-q-agt.txt.gz?level=TRACE14:55
tdurakovneed to figure out14:55
bauzasbut I'm missing snow, that's probably why14:55
edleafejaypipes: well, yeah, that would make no sense, so use the in: operator for that case14:55
cdentbauza, edleafe, yeah, in fact if we were writing the filtering, sorting and pagination guidelines from scratch, rather than precedent they would likely be _much_ different14:55
edleafejaypipes: What makes no sense at all is having some APIs where filters are ANDed, and some where filters are ORed.14:55
pkoniszewskitdurakov: and always subnode-2 is affected14:55
bauzascdent: edleafe: that's where I think we name that 'guidelines' and not rules :)14:56
jaypipesedleafe: there's never a case where an AND'd condition on a single attribute would ever make sense. But, hey, the main point is for consistency here.14:56
bauzasanyway, let's put in:14:56
dansmithvladikr: could it be related to the issue I fixed where the compute service was being restarted?14:56
dansmithvladikr: like maybe that was covering something up in the test that was broken?14:57
edleafebauzas: sure. The idea was that no one should have to change existing code, but all new development should follow the guidelines14:57
bauzashonestly, I don't want to discuss yet again 4 weeks for that point :)14:57
jaypipesagreed.14:57
bauzasso, given the API WG is saying "in:", let's write that "in:"14:57
*** tbachman has quit IRC14:57
tdurakovpkoniszewski: https://review.openstack.org/#/c/407120/ - there is a similar job that passes gate-grenade-dsvm-neutron-multinode-ubuntu-xenial14:58
jaypipesbauzas: ++14:58
bauzasanyway, edleafe, thanks for pointing that out now :)14:58
vladikrdansmith, it definitely feels like a failure that is being masked..14:58
edleafebauzas: :)14:59
tdurakovpkoniszewski: and both of them suffer from messaging timeout btw14:59
*** sudipto has joined #openstack-nova14:59
*** sudipto_ has joined #openstack-nova14:59
vladikrdansmith, maybe something related to saving pci devices.. as these are the only tests that are failing ..14:59
*** psachin has quit IRC15:00
pkoniszewskitdurakov: yeah, http://logs.openstack.org/20/407120/2/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/7b7054a/logs/subnode-2/old/screen-q-agt.txt.gz?level=TRACE15:00
pkoniszewskitdurakov: so not only lm grenade job is affected15:00
*** cdelatte has quit IRC15:00
jaypipestdurakov: done15:00
dansmithvladikr: so, if I apply the start-compute-only-once change to master, I get a failure15:00
*** siva_krish has joined #openstack-nova15:00
*** anush has joined #openstack-nova15:00
dansmithvladikr: which I'm pretty sure isn't expected, right?15:00
dansmithvladikr: because it's supposed to be testing with one compute service to make sure that the second instance doesn't get VFs from a used PF right?15:01
*** danpawlik has quit IRC15:01
openstackgerritMatt Riedemann proposed openstack/nova: DNM: debug novaclient limit/marker failure  https://review.openstack.org/40720415:02
*** mlavalle has joined #openstack-nova15:02
*** smatzek has joined #openstack-nova15:03
*** cdelatte has joined #openstack-nova15:03
vladikrdansmith, yea, there shouldn't be two instances running. are you referring to "if not self.compute_started:" that you've added in the test?15:04
*** hfu has joined #openstack-nova15:04
*** hamzy has joined #openstack-nova15:04
dansmithvladikr: yep15:04
dansmithvladikr: without that the test calls start_service(compute) more than once15:04
dansmithwhich it should not do15:04
dansmithregardless of the base test changes15:04
vladikrdansmith, yes15:05
mriedemdiana_clarke: what's confusing is the same request is hitting ComputeAPI.get_all() twice it looks like15:06
mriedemhttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/70528ca/logs/screen-n-api.txt.gz#_2016-12-06_00_42_17_88015:06
mriedemhttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/70528ca/logs/screen-n-api.txt.gz#_2016-12-06_00_42_17_95515:06
* diana_clarke looks15:06
mriedemsame request id15:06
mriedemsame 'searching by' log line15:06
mriedembut that should only be once per request15:06
*** penick has quit IRC15:07
*** tbachman has joined #openstack-nova15:07
mriedemi guess the 2nd query could be because of my debug call w/o the marker15:08
mriedembut the order seems messed up15:08
mriedemi.e. i have a 2nd get_all query here https://review.openstack.org/#/c/407204/3/nova/api/openstack/compute/servers.py15:08
mriedemactually nvm i'm just dumb15:08
mriedemthat's the right order15:08
mriedemmy new patch set should make this more clear which query has the marker15:08
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: libvirt: virtlogd: use virtlogd for char devices  https://review.openstack.org/40745015:09
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: libvirt: create consoles in an understandable/extensible way  https://review.openstack.org/40712015:09
markus_zsdague: jaypipes: ^ If you could have a look again please?15:10
lbeliveau_guys (and gals), I'm running into this issue with latest devstack (nova-serialproxy), I'll investigate it, but was wondering if anybody was working on it or seen it http://paste.openstack.org/show/591544/15:11
*** larainema_ is now known as larainema15:11
*** lbeliveau_ is now known as lbeliveau15:11
sdaguemarkus_z: lgtm. And good catches by jay on those conditionals15:12
*** mjura_ has joined #openstack-nova15:13
*** nandal has quit IRC15:13
*** cfriesen has joined #openstack-nova15:13
*** prateek has quit IRC15:14
vladikrdansmith, one of the test is failing because it's trying to save an empty pci device, however, the allocated pci object (on a second line) is complete https://paste.fedoraproject.org/500472/14810371/15:14
*** mjura has quit IRC15:15
dansmithvladikr: which test is that?15:15
tdurakovjaypipes 2c about mox, there are 2 classes: TestNeutronv2 and TestNeutronv2WithMock, so fix is merged, I'd be happy to rewrite these to mock as well15:15
*** yingjun has quit IRC15:15
vladikrdansmith, nova.tests.functional.libvirt.test_pci_sriov_servers.SRIOVServersTest.test_create_server_with_pci_dev_and_numa15:15
*** yingjun has joined #openstack-nova15:16
dansmithvladikr: that one passes for me15:16
vladikrdansmith, weird .. it failed in the CI15:17
vladikr2016-12-05 23:56:01.553491 | {7} nova.tests.functional.libvirt.test_pci_sriov_servers.SRIOVServersTest.test_create_server_with_pci_dev_and_numa [4.205866s] ... FAILED15:17
dansmithvladikr: strange15:18
dansmithvladikr: I wonder if it might be better to fix the test before even trying to make it work on top of this base test change/15:18
mriedemtdurakov: unwinding the mox in TestNeutronv2 is probably not the best use of your time or reviewer time atm15:18
*** tbachman has quit IRC15:18
mriedemas it's a terrible monster15:18
jaypipestdurakov: nah, it's cool :)15:18
*** tbachman has joined #openstack-nova15:19
tdurakovmriedem: right, it's a monster15:19
*** erlon-airlong has quit IRC15:20
openstackgerritChris Dent proposed openstack/nova-specs: Update the generic resource pools to reflect reality  https://review.openstack.org/40756215:20
dansmithvladikr: it didn't fail with that constraint issue though right?15:20
*** yingjun has quit IRC15:20
*** erlon-airlong has joined #openstack-nova15:20
*** eharney has joined #openstack-nova15:20
jaypipesmarkus_z: done.15:21
*** mjura_ has quit IRC15:21
vladikrdansmith, no, and the test was working fine before the cells change ..15:21
*** jwcroppe has joined #openstack-nova15:21
dansmithvladikr: sure, but the test is doing bad stuff already (starting two computes with the same hostname) and fails if you don't do that even on master15:22
dansmithvladikr: which makes me suspect its correctness :)15:22
markus_zjaypipes: awesome, thanks again!15:23
openstackgerritOleg Bondarev proposed openstack/nova: Live migration: wait for vif-plugged event on pre live migration  https://review.openstack.org/24691015:23
vladikrdansmith, well, I can push a quick fix for the two computes issue, but I doubt it was related to this issue.. the test it self doesn't do anything different from the other functional tests15:24
mriedemjaypipes: forgot to +W https://review.openstack.org/#/c/407120/15:24
vladikrdansmith, but sure, I'll submit a patch is a sec15:24
dansmithvladikr: you should push that fix, yeah15:24
dansmithvladikr: other functional tests don't start two computes with the same hostname :)15:25
jaypipesmriedem: sdorry, done.15:26
*** diga has quit IRC15:27
*** rama_y has joined #openstack-nova15:27
openstackgerritJulian proposed openstack/nova: Transform instance.snapshot notifications  https://review.openstack.org/33012815:29
*** mjura_ has joined #openstack-nova15:33
*** kfarr has joined #openstack-nova15:35
*** anush has quit IRC15:37
*** anush has joined #openstack-nova15:38
tdurakovmarkus_z: ping15:39
markus_ztdurakov: pong15:39
*** hongbin has joined #openstack-nova15:39
tdurakovmarkus_z: I couldn't see any output for grenade job at all http://logs.openstack.org/32/407532/1/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/33c4319/console.html#_2016-12-06_14_57_06_60234615:39
tdurakovam I missing smth?15:39
markus_ztdurakov: hm, I only looked at the top: http://logs.openstack.org/32/407532/1/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/33c4319/console.html#_2016-12-06_14_22_13_27093415:41
*** tblakes has quit IRC15:41
tdurakovmarkus_z: to compare http://logs.openstack.org/32/407532/1/check/gate-tempest-dsvm-multinode-live-migration-ubuntu-xenial/9ab7d67/console.html#_2016-12-06_14_46_39_77584215:41
markus_ztdurakov: Is it possible that empty lines get captured and not printed?15:42
tdurakovfor l-m job it's outputs smth15:42
markus_zah, yeah, forget what I said, there is the empty line too15:42
*** kiran-r has joined #openstack-nova15:43
*** rama_y has quit IRC15:43
* markus_z tries to figure that out15:44
*** mkoderer has quit IRC15:46
*** mkoderer has joined #openstack-nova15:47
markus_ztdurakov: mind if I add a patchset to that?15:48
tdurakovsure15:49
openstackgerritVladik Romanovsky proposed openstack/nova: tests: avoid starting compute service twice in sriov functional test  https://review.openstack.org/40757215:50
dansmithvladikr: are those two changes related?15:52
*** fragatina has quit IRC15:53
dansmithvladikr: also, that patch for me: {0} nova.tests.functional.libvirt.test_pci_sriov_servers.SRIOVServersTest.test_create_server_with_PF_no_VF [5.241744s] ... FAILED15:53
vladikrdansmith, no, just wanted to get this test to be as similar as the other numa test. the other thing is very trivial15:53
vladikrdansmith, I see.. this one fails for me, as in the CI..15:54
*** mdnadeem has joined #openstack-nova15:54
dansmithvladikr: that's what I was trying to say before.. just making that change causes a fail on master15:54
vladikrdansmith, does anything fails for you in that SRIOVServersTest test?15:54
dansmithvladikr: on the first run that's the only fail I saw, let me try in parallel15:55
dansmithvladikr: ah, they pass in parallel, so something else is going on :)15:55
vladikryea..15:56
*** armax has joined #openstack-nova15:56
sfinucanvladikr: Missing context, but assuming you're working on functional tests, have you seen this? https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:sriov_numa_func_test15:56
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: dnm to identify grenade job  https://review.openstack.org/40753215:56
*** mtanino has joined #openstack-nova15:58
vladikrsfinucan, sure, I've commented on https://review.openstack.org/#/c/399653 :)15:58
sfinucanvladikr: Cool - just checking :)15:59
vladikrsfinucan, I'll review it again today15:59
vladikr:)15:59
sfinucan...so as to avoid overlap etc.15:59
*** rama_y has joined #openstack-nova16:01
dansmithsfinucan: you're distracting vladikr from caring about *my* problem :)16:01
ildikovHi Nova cores! If any of you would have some time and check my attempt to remove 'check_attach' from the code base I very much appreciated it: https://review.openstack.org/#/c/335358/ :)16:02
sfinucandansmith: Barely in the door and already messing with other cores. Excellent :)16:02
vladikrdansmith, I think all of the tests are working on master. It only fails in https://review.openstack.org/#/c/40638016:03
dansmithvladikr: but not with your compute-once fix16:03
dansmithvladikr: but even still, the base test change really shouldn't be affecting those tests ...16:04
*** rama_y has quit IRC16:04
*** ayoung-afk is now known as ayoung16:05
*** HenryG has left #openstack-nova16:06
*** tblakes has joined #openstack-nova16:06
vladikrdansmith, locally all tests are passing with the compute-once fix  - will wait for the CI, and will look at the instance.save() with the /406380 change16:06
dansmithvladikr: have you tried running with --concurrency=1?16:07
dansmithvladikr: if it fails with some test arrangements and not others, that means there are interactions and it's a timebomb in CI that will eventually go off16:08
markus_zFor awareness: https://bugs.launchpad.net/neutron/+bug/164743116:08
openstackLaunchpad bug 1647431 in neutron "grenade job times out on Xenial" [Critical,Confirmed]16:08
vladikrdansmith, I see.. I think I did earlier.. let me try again now16:08
markus_zSeems to hit "gate-grenade-dsvm-neutron-multinode-ubuntu-xenial"16:08
dansmithvladikr: fails 100% of the time for me with --concurrency=116:08
*** sridharg has quit IRC16:08
dansmithvladikr: and the worrisome part is that it fails because it succeeds in starting the instance that should be failing16:09
dansmithvladikr: it also fails if you run that test by itself16:09
vladikrdansmith, got it.. this is the test_create_server_with_VF_no_PF test?16:10
dansmithvladikr: yeah16:10
vladikrok16:10
mriedemmarkus_z: hmm, not terrible deterministic http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22timeout%2030%20sh%20-c%20'while%20openstack%20server%20show%20cinder_server1%5C%22%20AND%20tags%3A%5C%22grenade.sh.txt%5C%22&from=7d16:11
mriedem*terribly16:11
*** ihrachys has quit IRC16:11
markus_zmriedem: yeah, the bug reporter noticed that too16:12
openstackgerritEd Leafe proposed openstack/nova: Corrects the type of a base64 encoded string  https://review.openstack.org/38568616:12
macszls16:12
*** kevinz has quit IRC16:12
macszoops, sorry :D16:12
macszlucky it was just ``ls`` :D16:13
mriedemlooks like it's mostly only on multinode jobs16:13
*** JoseMello has quit IRC16:13
markus_zmriedem: looks like this started on Nov 27th16:14
*** catinthe_ has joined #openstack-nova16:14
*** yamahata has joined #openstack-nova16:14
*** owalsh has quit IRC16:14
markus_zmaybe grenade jobs only?16:14
markus_zwhops, nevermind, you had that in the ELK query16:15
mriedemmarkus_z: hmm, this looks pretty gd awful http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-c-vol.txt.gz?level=TRACE16:15
*** bkopilov has joined #openstack-nova16:16
markus_zEverything red and orange, it's not *that* bad ... :/16:16
*** daemontool has joined #openstack-nova16:16
*** catintheroof has quit IRC16:16
*** code-R_ has quit IRC16:17
mriedemthat started around 12/216:17
*** rcernin has quit IRC16:17
*** catinthe_ has quit IRC16:18
*** annegentle has joined #openstack-nova16:18
*** nkrinner is now known as nkrinner_afk16:18
*** pcaruana has quit IRC16:19
*** rmart04 has quit IRC16:19
*** gszasz has quit IRC16:19
* mdbooth would like to re-pimp this change, which simply adds 4 lines of comments to the libvirt driver: https://review.openstack.org/#/c/407470/16:20
*** acabot_ has quit IRC16:20
*** acabot__ has quit IRC16:20
*** moshele has quit IRC16:22
*** oanson has quit IRC16:22
*** cdent has quit IRC16:22
*** zul has quit IRC16:22
*** sapcc-bot1 has joined #openstack-nova16:25
*** zul has joined #openstack-nova16:25
*** sapcc-bot has quit IRC16:25
*** mvk has quit IRC16:26
*** cdent has joined #openstack-nova16:26
vladikrdansmith, that's crazy.. it all passes, checked on two different envs. https://paste.fedoraproject.org/500517/10413071/16:26
*** bvanhav_ has quit IRC16:27
dansmithvladikr: wrong test16:28
dansmithPF_no_VF16:28
dansmithsorry I guess I confirmed the opposite above, but I pasted the full name of the test at some point16:28
vladikrdansmith, I've tested all of them.. all seems to pass, will double check in a sec16:29
vladikrdansmith, https://paste.fedoraproject.org/500521/41859148/16:31
markus_zmriedem: Version 3.1 is already 2 months old: https://github.com/openstack/cinder/commit/154404f5ddfe73f843e75196ed8e950d9edf9ba316:32
*** baoli has quit IRC16:32
markus_zI'm wondering if it is something outside of cinder16:32
mriedemmarkus_z: https://bugs.launchpad.net/cinder/+bug/164778916:33
openstackLaunchpad bug 1647789 in Cinder "RPCVersionCapError: Requested message version, 3.1 is incompatible. It needs to be equal in major version and less than or equal in minor version as the specified version cap 3.0. in c-vol in grenade jobs" [Undecided,New]16:33
mriedemred herring16:33
markus_zah crap, I looked at the wrong place16:35
openstackgerritChristopher Martin proposed openstack/nova: Bugfix: Changes to Instance Soft Delete  https://review.openstack.org/38675616:38
*** unicell has quit IRC16:39
dansmithvladikr: try this: .tox/functional/bin/python -m subunit.run /home/dan/nova/nova/tests/functional/libvirt/test_pci_sriov_servers.py | /home/dan/nova/.tox/py27/bin/subunit-trace16:39
vladikrit finally failed :)16:39
dansmithvladikr: with your home dir16:39
dansmithah, okay16:39
dansmithvladikr: it's a similar failure to what I see with the test base, but on more tests, which is why I think it might be related16:40
*** kiran-r has quit IRC16:40
vladikrdansmith, got it, ok, looking into it16:40
dansmithvladikr: thanks :)16:40
mriedemmarkus_z: it's weird, the server delete should be starting here http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/grenade.sh.txt.gz#_2016-12-05_09_07_08_77816:41
mriedemthe server uuid is 7e8aad9d-50e1-4b6c-8418-7395a035256f16:41
mriedemthe server delete should be around 09:07:08.77816:41
mriedembut i don't see any DELETE request for that server around that time in the api logs http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-n-api.txt.gz16:42
mriedemthere is a POST from neutron for an external network changed event because the floating IP was removed from the server16:42
mriedemah here we go http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-n-api.txt.gz#_2016-12-05_09_07_10_73416:44
*** bvanhav_ has joined #openstack-nova16:44
*** zenoway has quit IRC16:45
*** kaisers_ has joined #openstack-nova16:45
openstackgerritLudovic Beliveau proposed openstack/nova: Fix nova-serialproxy when registering cli options  https://review.openstack.org/40760216:45
openstackgerritmelanie witt proposed openstack/nova: Pass context as kwarg instead of positional arg to get_engine  https://review.openstack.org/40760316:45
*** zenoway has joined #openstack-nova16:46
lbeliveausfinucan: ^ can you please have a look since you are well versed with config options :)16:46
*** belmoreira has quit IRC16:47
mriedemno failures in the n-cpu logs16:47
*** stvnoyes has joined #openstack-nova16:48
*** ihrachys has joined #openstack-nova16:48
*** tbachman has quit IRC16:48
mriedemmarkus_z: last thing n-cpu does is post to cinder http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-n-cpu.txt.gz#_2016-12-05_09_07_13_26316:49
mriedemfor os-terminate_connection16:49
tdurakovmarkus_z, mriedem what's the problem you are working on?16:50
*** tjones has joined #openstack-nova16:50
*** zenoway has quit IRC16:50
mriedemtdurakov: https://bugs.launchpad.net/neutron/+bug/164743116:50
openstackLaunchpad bug 1647431 in neutron "multinode neutron grenade job times out on Xenial" [Critical,Confirmed]16:50
*** kaisers_ has quit IRC16:50
tdurakovmriedem: acked, will take a look too16:51
markus_zmriedem: I don't get why the cinder-scheduler automatically chooses version cap 3.016:51
*** rama_y has joined #openstack-nova16:53
mriedemmarkus_z: dulek is looking into that too16:53
markus_zOK, cool, then I call it a day16:53
markus_zIt's getting late here16:53
mriedemo716:53
*** sahid has quit IRC16:54
*** markus_z has quit IRC16:54
mriedemweird, so the volume terminate_connection rpc call to c-vol happens from c-api here http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-c-api.txt.gz#_2016-12-05_09_07_13_39116:54
mriedemthat request never shows up in c-vol16:55
mriedemreq-42b76e49-9bec-4ac5-b457-938bc82b364916:55
mriedemthere is a crazy amount of logging in c-vol16:57
*** mjura_ has quit IRC17:00
*** STEWIE925 has joined #openstack-nova17:01
bauzasedleafe: I'm getting mad at writing the regex that would match in:foo,bar :p17:01
mriedemtdurakov: at this point the instance is stopped http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-n-cpu.txt.gz#_2016-12-05_09_07_25_97817:01
sdaguebauzas: I expect you can't do that with just a regex17:02
*** roeyc has quit IRC17:02
bauzassdague: supposingly, I wrote re.findall('^in:(?:([^,]+),?)+, mystr)17:02
bauzasbut anyway, I'll just split the string without using re...17:03
bauzasthat's just sad17:03
mriedemtdurakov: this is really odd, we seem to just stop in _shutdown_instance in the compute manager17:03
mriedemwe post the os-terminate_connection to cinder17:03
sdagueyeh, I would just do that. Regex complexity for that just gets weird17:03
mriedembut we never make the os-detach call to cinder17:03
*** efoley has joined #openstack-nova17:03
sdaguewhereas if you do a couple of splits first, you can then easily validate the components17:03
dulekmriedem: Be advised that you have two c-vols - on primary and subnode.17:03
*** catintheroof has joined #openstack-nova17:03
*** catintheroof has quit IRC17:03
*** catintheroof has joined #openstack-nova17:04
mriedemah17:04
edleafebauzas: I'm knee-deep in some other stuff, but if you send me what you have I can take a look at it17:04
bauzasedleafe: sdague: nah, no worries, I'll just write a .split() stuff :(17:04
tdurakovmriedem: have you tried to find a root cause patch?17:05
mriedemdulek: the volume, 6645dbcf-3a6e-4444-8afc-16df7fd70267, is on the 'new' c-vol though, it doesn't show up in the logs for the subnode17:05
dulekmriedem: Hm…17:05
*** eharney has quit IRC17:05
dulekmriedem: So is this even related to the error I'm fixing?17:06
mriedemtdurakov: no because i don't know what's causing the server to not get deleted17:06
mriedemdulek: no17:06
mriedemtdurakov: we seem to just deadlock or something17:06
*** voelzmo has quit IRC17:07
tdurakovright, otherwise it will be 100% reproducible17:07
sfinucanlbeliveau: Sure. Checking17:07
dulekmriedem: I've observed something like that when I've messed up with oslo.messaging topics in my patch.17:07
mriedemwe're this far in n-cpu to call os-terminate_connection in cinder http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-n-cpu.txt.gz#_2016-12-05_09_07_13_26317:08
dulekmriedem: Actually exactly like that, it was terminate_connection.17:08
mriedemi can see c-api call c-vol17:08
mriedemright here http://logs.openstack.org/40/402140/7/check/gate-grenade-dsvm-neutron-multinode-ubuntu-xenial/ad0cf41/logs/new/screen-c-api.txt.gz#_2016-12-05_09_07_13_39117:08
mriedemthen we hang17:08
mriedemand die17:08
dulekI'll look into it once I'll push patch for bug 1647789.17:09
openstackbug 1647789 in Cinder "RPCVersionCapError: Requested message version, 3.1 is incompatible. It needs to be equal in major version and less than or equal in minor version as the specified version cap 3.0. in c-vol in grenade jobs" [High,In progress] https://launchpad.net/bugs/1647789 - Assigned to Michal Dulko (michal-dulko-f)17:09
*** andreas_s has quit IRC17:09
*** gjayavelu has joined #openstack-nova17:10
*** Swami has joined #openstack-nova17:10
*** dinobot_ has joined #openstack-nova17:10
lbeliveausfinucan: btw congrats on your "promotion" to core, well deserved :)17:11
*** rama_y has quit IRC17:11
openstackgerritEd Leafe proposed openstack/nova: Do not post allocations that are zero  https://review.openstack.org/40718017:11
mriedemso the c-api rpc call to c-vol for terminate_connection uses msg id dd4e70fdda9a4234a4377adbb24f983117:11
mriedemso i guess we can check the rabbit logs for that?17:11
dulekmriedem: Not much of them in what's published.17:12
*** gjayavelu has quit IRC17:12
*** gjayavelu has joined #openstack-nova17:12
*** anush has quit IRC17:13
*** matrohon has quit IRC17:14
mriedemdulek: yeah there is nothing in the rabbitmq logs17:14
dulekIt looks like there's a lot of guru meditation reports in new primary c-vol logs?17:15
*** anush has joined #openstack-nova17:15
*** prateek has joined #openstack-nova17:15
dulekRPCVersionCapError is probably uncatched and kills the threads.17:16
mriedemdulek: https://github.com/openstack-dev/devstack/commit/1f739c867a7fe6d54ca061c6352eed12873db5cc17:16
dulekBut we still shouldn't lose messages, right?17:16
mriedemidk17:16
dulekmriedem: Oh, okay, I see now.17:17
dulekmriedem: c-vol has this notify_service_capabilities in initialization. So it fails to spawn *any* thread that could be able to process *any* message.17:18
*** tbachman has joined #openstack-nova17:18
dulekmriedem: So this terminate_connection is first message that gets to new c-vol after upgrade. It fails, so everything fails.17:18
dulekmriedem: This would mean my commit (https://review.openstack.org/407612) fixes that.17:19
*** eharney has joined #openstack-nova17:20
*** irenab_ has joined #openstack-nova17:21
*** irenab has quit IRC17:23
*** irenab_ is now known as irenab17:23
mdboothhttps://review.openstack.org/#/c/407470/ <--- Any chance we can knock that on the head? It's trivial and it will save me some time.17:24
*** annegentle has quit IRC17:25
sdaguemdbooth: +2 from me17:25
mdboothsdague: Thanks17:25
* mdbooth is just trying to minimise rebases :)17:25
sdagueyep17:26
*** nic has joined #openstack-nova17:26
*** rcernin has joined #openstack-nova17:26
*** mdrabe has quit IRC17:27
*** bvanhav__ has joined #openstack-nova17:28
*** bvanhav_ has quit IRC17:29
*** eharney has quit IRC17:30
*** mlavalle has quit IRC17:31
*** diogogmt has joined #openstack-nova17:32
*** mdrabe has joined #openstack-nova17:32
*** sudipto_ has quit IRC17:32
*** sudipto has quit IRC17:32
*** lucasagomes is now known as lucas-afk17:33
*** code-R has joined #openstack-nova17:33
*** mdnadeem has quit IRC17:34
mriedemnow back to the other super fun gate failure for novaclient...17:35
*** aweeks has joined #openstack-nova17:35
*** baoli has joined #openstack-nova17:36
*** code-R_ has joined #openstack-nova17:36
*** code-R has quit IRC17:38
diana_clarkemriedem: yeah... I'm still stumped. Aside from knowing the cells v2 devstack change triggered it.17:39
sdaguemriedem: what failure are we looking at?17:40
mriedemsdague: https://review.openstack.org/#/c/407205/17:40
mriedemsdague: http://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/02d19a4/console.html#_2016-12-06_15_24_58_72002617:41
mriedemgetting a marker not found from a test in the client17:41
mriedemcreates 3 servers with the same name, waits for them to be active, then does a nova list --limit=1 --name <name>17:41
*** tblakes has quit IRC17:41
mriedemthe --limit=1 makes novaclient loop and pass in a marker17:41
diana_clarkesdague: And reverting this makes it green again: https://review.openstack.org/#/c/407319/17:41
mriedemdiana_clarke: i'm not even sure why the limit=-1 logic makes it make a 2nd call17:42
mriedembecause i thought limit=-1 meant, only make another call if there are more results to get17:42
mriedem?17:42
diana_clarkemriedem: something like that, I think. It's in while True loop, IIRC17:42
mriedemyeah https://github.com/openstack/python-novaclient/blob/master/novaclient/v2/servers.py#L84117:43
sdaguewhat is limit -1 supposed to mean semantically?17:44
mriedemsdague: see ^17:44
mriedemmeans page until there are no more results17:45
mriedemin the client17:45
*** eharney has joined #openstack-nova17:45
mriedemwe have 3 servers, so in the first query i expect it returns those 3, then since limit=-1 it takes the last one as the marker, which is 212edd36-6dc0-4c6f-a4fb-69e08e1c233517:45
mriedemthen makes a 2nd query with marker=212edd36-6dc0-4c6f-a4fb-69e08e1c233517:46
mriedemand that fails with a marker not found, which is confusing me17:46
mriedemhere is where we call with the marker set http://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/02d19a4/logs/screen-n-api.txt.gz#_2016-12-06_15_24_57_83517:47
*** yamahata has quit IRC17:47
*** fragatina has joined #openstack-nova17:48
*** rmart04 has joined #openstack-nova17:48
mriedemi dumped the sql query object in there17:48
mriedemand i don't actually see the marker in the query17:48
mriedemright here https://review.openstack.org/#/c/407204/3/nova/db/sqlalchemy/api.py@189117:49
mriedemis this http://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/02d19a4/logs/screen-n-api.txt.gz#_2016-12-06_15_24_57_85117:49
*** sreeram_v has joined #openstack-nova17:50
*** tblakes has joined #openstack-nova17:50
mriedemWHERE instances.uuid = :uuid_117:51
mriedemwtf is that17:51
mriedem?17:51
diana_clarkemriedem: the marker clause is there, just indented left: "WHERE instances.uuid = :uuid_1" I think anyway17:51
sdagueyeh, so this is an off by one error right?17:51
*** Zer0Byte__ has joined #openstack-nova17:51
mriedemthat should be WHERE instances.uuid = 212edd36-6dc0-4c6f-a4fb-69e08e1c233517:51
sdaguehttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/02d19a4/logs/screen-n-api.txt.gz#_2016-12-06_15_24_57_996 looks like this is a 400 on the API side17:51
*** bkopilov has quit IRC17:51
*** abalutoiu_ has joined #openstack-nova17:51
*** abalutoiu_ has quit IRC17:51
sdaguethe client is just translating it as such17:52
sdaguethe client is right, the server is wrong17:52
mriedemright but clearly we have an instance with that uuid17:52
mriedemERROR (BadRequest): marker [212edd36-6dc0-4c6f-a4fb-69e08e1c2335] not found; current instance UUIDs: ['66e7d166-2a6d-4810-9a70-5743cb2159e5', 'ef018cbe-a878-4ebc-b519-7fed72fd014e', '212edd36-6dc0-4c6f-a4fb-69e08e1c2335'] (HTTP 400) (Request-ID: req-042909e7-45e1-4485-b72d-c62391eeff29)17:52
mriedemb/c it's in that list17:52
mriedemi make the same query after the marker not found to list the same instances but w/o the marker17:53
sdaguemriedem: there is a name param on the filter as well17:53
sdague"GET /v2.1/servers/detail?marker=212edd36-6dc0-4c6f-a4fb-69e08e1c2335&name=95979b2e-599b-4d05-aa66-14d718c23747 HTTP/1.1"17:53
mriedemyeah, we don't get to the name filter i don't think in the db api17:53
mriedemb/c we're first looking up the instance by marker17:54
*** rama_y has joined #openstack-nova17:54
*** sreeram_v has quit IRC17:54
mriedemright here https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L223517:54
mriedemthat's before we apply the display_name filter in sqlalchemyutils.paginate_query17:55
dansmithvladikr: I might have figured out the other problem with those tests that show up after this fixture changes17:55
*** browne has joined #openstack-nova17:55
vladikrdansmith, great, what is it?17:55
dansmithvladikr: some interleaving with the mucking of global db state it's doing17:56
dansmithvladikr: so with that in place, the only fail I see is that one PF_no_VF failure thing17:56
*** baoli has quit IRC17:56
vladikrdansmith, ok, yea, I'm still debugging it - hope to finish soon17:57
dansmithvladikr: coolt hanks17:57
*** baoli has joined #openstack-nova17:57
mriedemsdague: diana_clarke: https://github.com/openstack/nova/commit/9ffa622b6ff062379e489809c0ed5a1b61881916 - i wonder if after we turned cells v2 on, if the query is going to the API DB or something?17:57
*** weshay_ has quit IRC17:58
*** bkopilov has joined #openstack-nova17:58
diana_clarkemriedem: But those tables don't exist in the API database, correct? I would expect an exception in that case. I was thinking that perhaps one of the joins are failing on get because perhaps in the cells v2 case something isn't created that's joined on.17:58
*** rmart04 has quit IRC17:59
diana_clarkeBut honestly, I don't know... that was my only remaining hunch.17:59
*** weshay has joined #openstack-nova17:59
mriedemmelwitt: do we lose anything needed for picking the proper cell when we elevate a context?18:00
*** mvk has joined #openstack-nova18:00
*** rmart04 has joined #openstack-nova18:00
melwittmriedem: let me check. it will be a matter of whether the db_connection attr is preserved during the clone18:00
dansmithmriedem: it does a deepcopy18:00
dansmithso I think it should be fine18:00
*** Guest17450 has quit IRC18:01
*** tbachman has quit IRC18:01
*** alexpilotti has joined #openstack-nova18:02
cfriesenjaypipes: are we expecting the resource provider work to solve https://bugs.launchpad.net/nova/+bug/1469179 in Ocata?  Or would that not land till Pike?18:02
openstackLaunchpad bug 1469179 in OpenStack Compute (nova) "instance.root_gb should be 0 for volume-backed instances" [Medium,In progress] - Assigned to melanie witt (melwitt)18:02
mriedemi guess i'm still confused about "WHERE instances.uuid = :uuid_1"18:02
mriedemand how to actually get the full sql we'd see when it's executed18:02
diana_clarkemriedem: Put another way, perhaps the instance create path differs subtly in the cells v2 case, making the get not work later.18:02
*** jwcroppe has quit IRC18:03
sdaguemriedem: why do you think it's not https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L2242 when it blows up?18:03
*** tbachman has joined #openstack-nova18:03
mriedemsdague: we don't get that far18:03
*** syjulian_ has joined #openstack-nova18:03
mriedemsdague: we get here https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L224018:03
mriedemand raise that up to the rest api18:03
melwittthat means the uuid is missing right i.e. nothing was passed? that marker was None?18:03
diana_clarkesdague: It raises a MarkerNotFound exception a few lines before that18:03
mriedemmelwitt: that wouldn't happen https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L223518:04
mriedemmelwitt: and: 2016-12-06 15:24:57.835 27484 DEBUG nova.compute.api [req-042909e7-45e1-4485-b72d-c62391eeff29 admin admin] Searching with limit '1000' and marker '212edd36-6dc0-4c6f-a4fb-69e08e1c2335' get_all /opt/stack/new/nova/nova/compute/api.py:231618:04
dansmithvladikr: although a bunch now fail in new and different ways with the top patch :(18:04
*** siva_krish has quit IRC18:04
melwittokay18:04
openstackgerritChris Dent proposed openstack/nova: Support filtering resource providers by aggregates  https://review.openstack.org/40762918:04
dansmithvladikr: I dunno why these are so different from everything else18:04
openstackgerritSylvain Bauza proposed openstack/nova: Extend get_all_by_filters to support resource criteria  https://review.openstack.org/38624218:05
openstackgerritSylvain Bauza proposed openstack/nova: WIP: Expose a REST API for a specific list of RPs  https://review.openstack.org/39256918:05
openstackgerritSylvain Bauza proposed openstack/nova: Support multiple uuids and names for RP list  https://review.openstack.org/40763018:05
sdaguediana_clarke: our MarkerNotFound error messages are different though, right? I guess I was trying to figure out where it was translating18:05
*** derekh has quit IRC18:05
vladikrdansmith, are these only the sriov tests?18:05
mriedemsdague: in the rest api, we throw away the MarkerNotFound message18:05
dansmithvladikr: yeah18:05
mriedemsdague: https://review.openstack.org/#/c/407204/3/nova/api/openstack/compute/servers.py@35818:05
mriedemmaybe i should run with CONF.database.connection_trace = True18:05
bauzasjaypipes: dansmith: cdent: edleafe: my pleasure to tell to you that I wrote a new REST API for the RP list https://review.openstack.org/39256918:06
sdaguemriedem: ah, I see now18:06
*** siva_krish has joined #openstack-nova18:06
*** ralonsoh has quit IRC18:06
bauzasjaypipes: dansmith: cdent: edleafe: I'll be writing the gabbi tests tonighty18:06
*** code-R_ has quit IRC18:06
*** alexpilotti has quit IRC18:06
*** huhaoran has quit IRC18:07
*** code-R has joined #openstack-nova18:07
openstackgerritSylvain Bauza proposed openstack/nova: WIP: Expose a REST API for a specific list of RPs  https://review.openstack.org/39256918:09
diana_clarkemriedem: To answer your earlier question, this is 100% normal: " :uuid_1". That's what a compiled query looks like before the bind parameters are applied.18:09
mriedemdiana_clarke: is there a way to see what it looks like after the parameters are applied?18:10
*** eandersson has quit IRC18:10
*** kristian__ has quit IRC18:10
mriedembtw18:11
mriedemoslo.db 4.15.0 on 12/1 https://review.openstack.org/#/c/405491/18:11
diana_clarkemriedem: you can increase the connection_debug logging in the nova conf (which is what I do when I play with devstack)18:11
mriedemmaybe there is something coincidental in there with the cells v2 stuff18:11
sdagueor turn on slow query on the mysql side to 0, and have it dump all the queries18:11
sdagueit might be easier to read than the sqlalchemy versions of it18:12
mriedemdiana_clarke: heh18:12
mriedem# Verbosity of SQL debugging information: 0=None, 100=Everything. (integer # value) # Minimum value: 0 # Maximum value: 100 # Deprecated group/name - [DEFAULT]/sql_connection_debug #connection_debug = 018:12
diana_clarkemriedem: I think you can also .compile().params18:12
mriedemmaybe 75 has what i want, maybe 25 does, who knows?!18:13
diana_clarkeon the query, but it's been a long time since I last tried. edleafe was asking the same thing a few days ago18:13
* mriedem summons zzzeek18:13
diana_clarkemriedem: greater than 50 is what you want, I think18:13
zzzeekhey, got about 10 minutes here18:13
mriedemzzzeek: https://review.openstack.org/#/c/407204/3/nova/db/sqlalchemy/api.py18:13
mriedemi want to get the actual compiled params from that query before executing it18:14
mriedemand dump the straight up sql that's going to be executed18:14
mriedemis there an easy way?18:14
sdagueso, a little bit of a distraction, but we're doing an aweful lot of work in getting a full instance (with join tables) when we litterally only need an object with .project_id and .uuid18:14
diana_clarkemriedem: http://git.openstack.org/cgit/openstack/rack/tree/rack/openstack/common/db/sqlalchemy/session.py#n60018:14
zzzeekmriedem: it's not "easy" but it is:  compiled = query.statement.compile();  compiled.params18:14
mriedemsdague: yes i noticed that too18:14
mriedemdiana_clarke: ah nice, we should maybe update the help text for that config option18:15
sdagueit might be easier to see what is going on if we did that optimization there first18:15
mriedemzzzeek: and compiled.params is the sql string?18:15
zzzeekmriedem: that's a dictionary of the parameters themselves18:15
*** anush has quit IRC18:15
zzzeekstr(compiled) is the SQL string18:15
mriedemok that's what i want, thanks18:15
zzzeekmriedem: if you want it to be the MySQL compiled, statemnet.compile(dialect=my_engine.dialect)18:16
diana_clarkemriedem: this example might help visualize it: http://docs.sqlalchemy.org/en/latest/core/tutorial.html#insert-expressions18:17
openstackgerritDan Smith proposed openstack/nova: Make servers api view load instance fault from proper cell  https://review.openstack.org/40723818:17
openstackgerritDan Smith proposed openstack/nova: Add schedule_and_build_instances conductor method  https://review.openstack.org/36755718:17
openstackgerritDan Smith proposed openstack/nova: WIP Multicell support for instance listing  https://review.openstack.org/39677518:17
openstackgerritDan Smith proposed openstack/nova: Make RPCFixture support multiple connections  https://review.openstack.org/39641718:17
openstackgerritDan Smith proposed openstack/nova: Setup CellsV2 environment in base test  https://review.openstack.org/40638018:17
openstackgerritDan Smith proposed openstack/nova: Add SecurityGroup.identifier to prefer uuid over name  https://review.openstack.org/40626018:17
openstackgerritDan Smith proposed openstack/nova: Add SingleCellSimple fixture  https://review.openstack.org/40568418:17
openstackgerritDan Smith proposed openstack/nova: WIP Move instance creation to conductor  https://review.openstack.org/31937918:17
openstackgerritDan Smith proposed openstack/nova: Fix up non-cells-aware context managers in test_db_api  https://review.openstack.org/40637918:17
openstackgerritDan Smith proposed openstack/nova: Add a CellDatabases test fixture  https://review.openstack.org/39971018:17
dansmithmelwitt: so I updated the fixture patch.. I think we *have* to lock before we muck with global db state ^18:18
dansmithmelwitt: basically, what was happening in one case was that a test created an instance, which was in the middle of going through the build process in cell1,18:18
openstackgerritSarafraj Singh proposed openstack/nova: conf: remove deprecated barbican options  https://review.openstack.org/39777418:19
dansmithmelwitt: and that test would poll the api, which would switch between cell1 and cell0 to compile the full list, and occasionally overlap with the build process, causing it to fail because suddenly the compute manager methods were hitting cell018:19
melwittdansmith: ugh, okay18:19
*** tbachman has quit IRC18:19
dansmithmelwitt: the fix isn't as ugly as I thought it was going to be, luckily18:19
dansmithand it makes sense18:20
openstackgerritMatt Riedemann proposed openstack/nova: DNM: debug novaclient limit/marker failure  https://review.openstack.org/40720418:20
melwittdansmith: yeah. I'm not surprised there were more issues with _last_ctxt_mgr. I just didn't know what it was going to be18:21
melwittthanks for fixing that :)18:22
dansmithyup18:22
dansmithnarrowing the problems I think18:22
dansmithI think there might be one more with cell0 stuff, on top of what vladikr is debugging right now, but hoping the light I see is the end of the tunnel18:22
dansmithand not the one you see right before death18:22
melwittlol18:23
melwittthat's a good one18:23
*** efoley has quit IRC18:23
* dansmith notes that his potential demise is funny to melwitt 18:23
openstackgerritSarafraj Singh proposed openstack/nova: conf: remove deprecated barbican options  https://review.openstack.org/39777418:25
diana_clarkemriedem: Either way, I don't think seeing the query will help. I suspect it's passing the correct marker, and that a join is failing.18:25
vladikrdansmith, I think I think I know what the problem is with the test.. the parent addresses of the vf are not generated in fakelibvirt - it doesn't matter for the rest of the tests, but this one is completely wrong - fixing it now18:25
melwittheh18:25
dansmithvladikr: sweet18:25
edleafebauzas: sweet18:25
*** rama_y has quit IRC18:26
*** rfolco has quit IRC18:26
*** yamahata has joined #openstack-nova18:26
*** krtaylor has quit IRC18:28
openstackgerritmelanie witt proposed openstack/nova: Handle ImageNotFound exception during instance backup  https://review.openstack.org/38851918:29
*** alexpilotti has joined #openstack-nova18:31
mriedemsdague: going back to your question, it looks like paginate_query in oslo.db only uses the marker for pulling sort keys, which by default are https://review.openstack.org/#/c/407204/3/nova/db/sqlalchemy/api.py18:31
mriedemoops18:31
*** kaisers_ has joined #openstack-nova18:31
mriedemdefault_keys=['created_at', 'id']18:31
*** tbachman has joined #openstack-nova18:31
*** jwcroppe has joined #openstack-nova18:32
mriedemand as we shouldn't sort on joined tables, yeah we probably don't need those18:32
mriedemeven though today you can sort on joined tables, but we're fixing that18:32
*** kaisers_ has quit IRC18:33
*** alexpilotti has quit IRC18:33
*** kaisers_ has joined #openstack-nova18:33
*** alexpilotti has joined #openstack-nova18:33
*** Apoorva has joined #openstack-nova18:33
edleafebauzas: oh, wait - you're doing the claim logic? I thought that wasn't on the table for Ocata18:34
*** penick has joined #openstack-nova18:35
*** dharinic is now known as dharinic|lunch18:35
*** penick_ has joined #openstack-nova18:36
melwittdiana_clarke, mriedem: did anyone try not elevating the context and seeing if it succeeds otherwise? I noticed elevated only deep copies self.roles and shallow copies everything else, but I didn't think that should matter18:36
mriedemhaven't tried it18:36
*** mkoderer has quit IRC18:36
mriedemi also want to try not joining on all of these other tables18:36
melwittokay18:37
*** rama_y has joined #openstack-nova18:37
*** b_rake has joined #openstack-nova18:37
*** b_rake has quit IRC18:38
mriedemmelwitt: that would still copy db_connection right?18:38
*** tbachman has quit IRC18:38
melwittmriedem: shallow copy, yeah18:38
*** b_rake has joined #openstack-nova18:39
*** penick has quit IRC18:39
*** penick_ is now known as penick18:39
tdurakovsdague: hi, what will you recommend me on this one: https://review.openstack.org/#/c/389767/ if it's should be tempest plugin, should I exclude all live-migration related tests from tempest as well?18:39
*** tbachman has joined #openstack-nova18:42
*** b_rake has quit IRC18:43
*** rfolco has joined #openstack-nova18:44
*** tbachman has quit IRC18:44
*** lpetrut has joined #openstack-nova18:46
*** anush has joined #openstack-nova18:47
*** penick has quit IRC18:48
jrolljaypipes: jfyi, started on the CI stuff for the resource class things. gonna be a couple days (given project-config is frozen) but I think I've got the right direction here: https://review.openstack.org/#/q/topic:bug/164781318:49
*** lpetrut has quit IRC18:50
jaypipesjroll: rock on brother.18:53
jaypipesbauzas: reviewing now.18:54
*** krtaylor has joined #openstack-nova18:54
jaypipescfriesen: dansmith already solved that for the resource providers inventory/allocation records: https://github.com/openstack/nova/blob/master/nova/scheduler/client/report.py#L103-L10718:55
*** weshay has quit IRC18:55
*** rmart04 has quit IRC18:58
*** weshay has joined #openstack-nova18:58
*** tbachman has joined #openstack-nova19:02
stvnoyeshi, i'm having trouble using a remote debugger (pdev/eclipse) with nova-compute in devstack. I started nova-compute with  --remote_debug-host 192.168.9.94 --remote_debug-port 5678 , and I hit the breakpoint at the log msg warning about eventlet. But when I resume execution,  nova-compute hangs after printing "Connecting to libvirt: qemu:///system" in the log. Is this familiar to anyone?19:02
mriedembauzas: jaypipes: could use your input on the last comment from me here https://review.openstack.org/#/c/406627/19:04
mriedemabout the proper order and way handle deleting a compute service record and the compute node19:04
mriedem*way of handling19:04
*** huhaoran has joined #openstack-nova19:07
*** jamielennox|away is now known as jamielennox19:07
*** nic has quit IRC19:08
*** nic has joined #openstack-nova19:09
mriedemjaypipes: also, can't we get rid of create/update_inventory methods in the ComptueNode object now?19:09
mriedemas that's done in the RT19:10
mriedemi thought that stuff was all reverted out in newton19:10
jaypipesmriedem: yes we absolutely can now.19:11
mriedemaye aye19:11
mriedemsdague: you could drop the glance v1 code in nova now too if you wanted :)19:11
mriedemi don't think anyone has claimed that yet19:11
jaypipesmriedem: b/c Mitaka was the last code that called that, IIRC. dansmith pls confirm?19:12
*** huhaoran has quit IRC19:12
dansmithjaypipes: mriedem they're not remotable so they have nothing to do with when they were last called19:12
openstackgerritChris Dent proposed openstack/nova: Fix a typo in a comment in microversion history  https://review.openstack.org/40766419:12
*** Apoorva has quit IRC19:12
dansmithalthough they are still using remotable methods on the RP objects19:13
cdentsdague: ^ that's in response to something you noticed19:13
dansmithwhich has impacts for cdent's removal of those19:13
dansmithbut I assume we're just breaking all the rules for that anyway19:13
*** Apoorva has joined #openstack-nova19:13
cdent@I might as well begin to put some action in my life@19:13
*** catinthe_ has joined #openstack-nova19:14
mriedemdansmith: idk anything about that, i just don't see anything calling create/update_inventory on the ComputeNode object anywhere19:14
mriedemso assume it's dead code19:14
*** bapalm_ has quit IRC19:14
dansmithmriedem: I assume it is as well, if you have confirmed, then they can come out yeah19:14
mriedemok19:14
*** catintheroof has quit IRC19:16
*** liverpooler has joined #openstack-nova19:16
*** bapalm has joined #openstack-nova19:18
*** code-R has quit IRC19:20
jaypipesmriedem: done.19:20
sdaguecdent: thanks19:21
mriedemjaypipes: reading, thanks19:21
sdaguecdent: +A19:21
jaypipesbauzas: you may want to weigh in on https://review.openstack.org/#/c/406627 as well since it was you and I who removed the linkage between compute_nodes and services.19:21
sdaguemriedem: sure, I could probably cycle around on that19:21
mriedemjaypipes: i see some RT/computenode/servicegroup love of yours in here :)19:21
jaypipesmriedem: :)19:21
jaypipesmriedem: bad memories.19:21
*** fragatina has quit IRC19:21
mriedemit's over now jay19:22
mriedemwell, not really19:22
jaypipesmriedem: :)19:22
mriedembut something something PTSD comforting19:22
tdurakovmelwitt: are you around?19:22
melwitttdurakov: yes19:22
tdurakovmelwitt: quick question on https://review.openstack.org/#/c/407603/19:23
*** Sukhdev has joined #openstack-nova19:23
melwittokay19:24
tdurakovmelwitt: how will this affect execution flow, if main_context_manager will be used instead of db_connection, that is stored in context?19:24
* tdurakov trying to figure out whether this one should be backported or not19:25
*** dharinic|lunch is now known as dharinic19:27
mriedemsmcginnis: was the bug smash last week advertised all that well? i wasn't even aware of it happening.19:27
mriedemsmcginnis: or did you just direct them away from all of the nova bugs and make them focus on cinder bugs19:28
melwitttdurakov: I'm not sure I follow. the way it is currently, the context passed in is never being used because it's assigned to use_slave19:28
*** annegentle has joined #openstack-nova19:31
*** Sukhdev has quit IRC19:31
*** claudiub has joined #openstack-nova19:32
*** Sukhdev has joined #openstack-nova19:32
*** McNinja has quit IRC19:33
tdurakovmelwitt: yes, but I'm still not sure, do we have bugs because of that or not19:33
*** morgabra has quit IRC19:34
*** morgabra has joined #openstack-nova19:34
*** morgabra has joined #openstack-nova19:34
*** claudiub|2 has quit IRC19:35
*** slunkad has quit IRC19:35
*** andreaf has quit IRC19:35
*** toabctl has quit IRC19:35
melwitttdurakov: oh, I see what you mean. I think not yet because we don't have multiple cells v2 cells yet19:35
*** McNinja has joined #openstack-nova19:36
*** toabctl has joined #openstack-nova19:36
*** slunkad has joined #openstack-nova19:36
*** andreaf has joined #openstack-nova19:37
tdurakovmelwitt: acked, thanks19:38
*** Jeffrey4l has quit IRC19:42
*** Jeffrey4l has joined #openstack-nova19:42
*** kristian__ has joined #openstack-nova19:43
*** dinobot_ has quit IRC19:44
*** siva_krish has quit IRC19:48
*** ijw has quit IRC19:49
*** claudiub|2 has joined #openstack-nova19:50
dansmithmelwitt: mriedem: okay so the bottom few patches of this series are still good to go: https://review.openstack.org/#/c/399710/519:52
dansmithI have -Wd the one where we need to hold off until vladikr fixes the sriov tests19:52
dansmithbut that series is getting a little big for its britches, so would like to see some progress there19:52
*** claudiub has quit IRC19:53
melwittk, will look today19:53
dansmiththanks19:53
dansmithwe need someone else for the fixture anyway19:53
melwittyes19:54
*** jwcroppe has quit IRC19:54
*** WillUK2010 has joined #openstack-nova19:56
*** moshele has joined #openstack-nova19:59
*** rfolco has quit IRC20:01
*** catintheroof has joined #openstack-nova20:02
mriedemi was +2 on the fixture last night20:04
*** catinthe_ has quit IRC20:04
*** kristian__ has quit IRC20:04
mriedembut i see it's all been drastically changed20:04
*** portdirect__ is now known as portdirect_away20:04
dansmithmriedem: drastically fixed you mean20:05
smcginnismriedem: No and yes. ;)20:05
mriedemdrastically awesomatized yes20:05
smcginnismriedem: I only recall seeing one mention in the ML that the bugsmash was coming up.20:05
*** siva_krish has joined #openstack-nova20:06
smcginnismriedem: The only reason I even knew about it this time was I had missed the last two and kind of wanted to participate, especially since I made Ocata mostly bugfixes for Cinder.20:06
smcginnismriedem: I'm hoping by that posting it can get better communicated out for next time.20:06
smcginnismriedem: There was a group working on Nova bugs, but I didn't spend any time at that table.20:07
smcginnismriedem: At a minimum, I think it would be good to alert the cores of the projects that have attendees working on bugs so they can watch out for and review their patches while we can take advantage of their focused attention.20:08
mriedemsmcginnis: they got the timing right this time around re: the release schedule20:08
mriedemso it's not in the 3rd milestone or later20:08
smcginnismriedem: Right!20:08
mriedemi just hadn't heard about it happening until you said you were there last week20:08
*** kristian__ has joined #openstack-nova20:08
*** portdirect_away is now known as portdirect20:10
*** kaisers_ has quit IRC20:11
melwittdiana_clarke, mriedem: I think the problem with the limit/marker thing might be that the servers are BuildRequests and when the 2nd query is made with the marker, the instances query raises MarkerNotFound instead of just returning an empty list? as there are no instances in the instances table, they are build_requests20:12
mriedemmelwitt: it's not that20:12
*** hfu has quit IRC20:12
mriedemmelwitt: we still create the instances in the db from the api right now20:12
mriedemthat hasn't changed20:13
mriedemand this test is waiting for the servers to be active before doing the query20:13
*** kaisers_ has joined #openstack-nova20:13
mriedemmelwitt: but that's what i was thinking of yesterday, that we were actually querying against build requests and there was a bug in that, but i don't think so20:13
melwitthmm20:13
mriedemplus my debug logging in the recreate is showing we're getting down to the point of querying the instances table20:13
openstackgerritMatt Riedemann proposed openstack/nova: Removed unused ComputeNode create/update_inventory methods  https://review.openstack.org/40769220:15
openstackgerritLudovic Beliveau proposed openstack/nova: Resource tracker doesn't free resources on confirm resize  https://review.openstack.org/40557520:16
lbeliveaujaypipes: ^ I have a fix for this RT bug I mentionned to you20:16
diana_clarkemelwitt: I think mriedem added an if statement on build_requests, but I never saw it appear in the logs, so i don't think so...20:16
openstackgerritVasyl Saienko proposed openstack/nova: DO NOT REVIEW  https://review.openstack.org/40769320:16
jaypipeslbeliveau: rock on brother. reviewing now.20:17
mriedemdo i smell some hulkamaniacs?20:17
jaypipesstep in to a slim jim.20:17
mriedemoooo yyyyeeeaaahhhh20:17
lbeliveau:)20:18
*** kaisers_ has quit IRC20:18
mriedemdansmith: on https://review.openstack.org/#/c/396417/20/nova/rpc.py did oslo.messaging get a release out with the fix for that hack?20:20
mriedemdiana_clarke: bleh still "WHERE instances.uuid = :uuid_1"20:23
mriedemhttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/0c20b61/logs/screen-n-api.txt.gz#_2016-12-06_19_03_40_21220:23
diana_clarkemriedem: yeah, you didn't print .params ;(20:23
mriedeminteresting though20:24
mriedemhttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-neutron-ubuntu-xenial/0c20b61/logs/screen-n-api.txt.gz#_2016-12-06_19_03_40_33320:24
mriedem^ is one the query with marker=None20:24
mriedemthe 2nd query20:24
mriedemthat doesn't show up before the first one20:24
mriedemwhich is https://review.openstack.org/#/c/407204/4/nova/compute/api.py@241020:25
mriedemwait a minute20:25
mriedemare we querying cell0? https://review.openstack.org/#/c/407204/4/nova/compute/api.py@239920:25
*** jwcroppe has joined #openstack-nova20:25
mriedemi bet we are20:26
mriedemupdating that patch to see20:26
*** portdirect is now known as portdirect_away20:26
*** bvanhav__ has quit IRC20:27
*** portdirect_away has quit IRC20:27
openstackgerritMatt Riedemann proposed openstack/nova: Removed unused ComputeNode create/update_inventory methods  https://review.openstack.org/40769220:29
*** slaweq_ has quit IRC20:30
*** sapcc-bot1 has quit IRC20:32
*** sapcc-bot has joined #openstack-nova20:32
*** syjulian_ has quit IRC20:33
*** moshele has quit IRC20:36
vladikrdansmith, the test is passing now. going to pickup the kids ans will send the patch.20:36
openstackgerritMatt Riedemann proposed openstack/nova: DNM: debug novaclient limit/marker failure  https://review.openstack.org/40720420:37
mriedemdiana_clarke: melwitt: sdague: i think ^ is going to show it20:37
mriedemi think we're getting a MarkerNotFound from cell020:37
*** slaweq_ has joined #openstack-nova20:37
mriedemthen when i query the 2nd time, the query to cell0 returns nothing, and then we hit the actual instances table20:38
dansmithvladikr: thanks20:38
dansmithmriedem: ah, forgot about that hack, damn20:38
dansmithmriedem: they merged, but no release It hink20:39
*** ijw_ has joined #openstack-nova20:39
*** ijw_ has quit IRC20:39
*** ijw_ has joined #openstack-nova20:39
*** gouthamr has quit IRC20:39
jaypipeslbeliveau: done20:40
*** prateek has quit IRC20:40
*** Sukhdev has quit IRC20:41
lbeliveaujaypipes: thanks !  will work on it20:41
*** kaisers_ has joined #openstack-nova20:42
jaypipeslbeliveau: no, thank YOU! :)20:42
mriedemdansmith: how much do you want to wait for oslo.messaging release? or just move forward20:42
*** kashyap has joined #openstack-nova20:42
*** kashyap has quit IRC20:43
*** kashyap has joined #openstack-nova20:43
*** daemontool_ has joined #openstack-nova20:44
mriedemnvm i guess harlow is doing one now20:46
*** daemontool has quit IRC20:47
*** diogogmt has quit IRC20:48
*** matrohon has joined #openstack-nova20:48
*** siva_krish has quit IRC20:51
*** david-lyle has quit IRC20:55
dansmithmriedem: there is a slightly less eye-killing hack I can do I think, but if there is a release coming then that's probably better20:56
*** xyang1 has joined #openstack-nova20:56
*** slaweq_ has quit IRC21:01
mriedemdansmith: question in https://review.openstack.org/#/c/396417/20/nova/tests/fixtures.py21:02
*** xinliang has quit IRC21:06
*** catintheroof has quit IRC21:06
openstackgerritEd Leafe proposed openstack/nova: WIP - Cascade deletes of RP aggregate associations  https://review.openstack.org/40770721:06
*** slaweq_ has joined #openstack-nova21:07
*** toure is now known as toure|biab21:07
*** ociuhandu has quit IRC21:07
*** fragatina has joined #openstack-nova21:09
*** huhaoran has joined #openstack-nova21:09
*** dave-mccowan has quit IRC21:10
*** cmartin has joined #openstack-nova21:14
*** huhaoran has quit IRC21:14
dansmithmriedem: replied (in case you didn't see already)21:14
cmartinmdbooth, Hey if you are there I'd like to have a discussion about this bug https://bugs.launchpad.net/nova/+bug/150180821:15
openstackLaunchpad bug 1501808 in OpenStack Compute (nova) "Enabling soft-deletes opens a DOS on compute hosts" [High,In progress] - Assigned to Chris Martin (cm876n)21:15
mriedemdansmith: yeah i think a fixme is more clear21:15
*** xinliang has joined #openstack-nova21:18
*** cdelatte has quit IRC21:18
*** kristia__ has joined #openstack-nova21:22
*** kristian__ has quit IRC21:22
mriedemdiana_clarke: ha success http://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-identity-v3-only-ubuntu-xenial-nv/61e9a05/console.html21:27
mriedemhttp://logs.openstack.org/05/407205/1/check/gate-novaclient-dsvm-functional-identity-v3-only-ubuntu-xenial-nv/61e9a05/logs/screen-n-api.txt.gz#_2016-12-06_20_58_54_41721:27
mriedemit was cell021:27
mriedemdirty old cell021:27
openstackgerritChris Dent proposed openstack/nova: Support filtering resource providers by aggregates  https://review.openstack.org/40762921:29
mriedemmelwitt: dansmith: do you know if the functional tests setup cell0 for us automatically?21:30
mriedemi'd rather fix this with a functional test than a comptue api unit test using mock21:31
*** Jeffrey4l has quit IRC21:32
*** Jeffrey4l has joined #openstack-nova21:34
mriedemlooks like the database fixture only handles main and api dbs, but i'd think that could be changed to be main or cell021:36
mriedemor api21:36
mriedemb/c main and cell0 setup the same schema21:36
*** ihrachys has quit IRC21:36
dansmithmriedem: they do as of my base setup patch21:37
dansmithmriedem: otherwise no21:37
mriedemso something like this:21:38
mriedemfix = nova_fixtures.CellDatabases()21:38
mriedemfix.add_cell_database('main')21:38
mriedemfix.add_cell_database('cell0')21:39
mriedemself.useFixture(fix)21:39
mriedem?21:39
*** READ10 has quit IRC21:39
mriedemand then create the cell0 mapping in the api db i think21:39
dansmithmriedem: that won't make cell0 work, that will give you two cell databases21:39
dansmithyeah21:39
*** hamzy has quit IRC21:39
dansmithor just merge my patch that does this for any db-using test :)21:39
mriedemdude21:39
dansmithwell, once we get vladikr's fix21:39
mriedemi +2d the bottom fixture patch21:39
dansmithI know, we can't even merge the test base patch until the rpc one, which depends on o.msg, and vladikr's fix :)21:40
dansmithbut.. UH21:40
mriedemwhy can't we merge the bottom change yet?21:40
mriedemnothing is using it yet21:40
dansmiththe bottom one we can yeah21:41
dansmithlet me clarify with links21:41
dansmithmriedem: this is what I meant: https://review.openstack.org/#/c/406380/21:42
mriedemoh yeah that's what i want21:43
mriedemthat's what i NEED21:43
dansmithyou know I'll give you what you need21:43
dansmithyou KNOW it21:43
mriedemwell in the interest of time i'm probably just going to unit test this with a mock21:44
mriedemb/c novaclient is blocked21:44
dansmithsure21:46
diana_clarkemriedem: nice work! It's still unclear to me which tables to expect in the cell0 database vs the others, but that can be my homework ;)21:48
mriedemdiana_clarke: cell0 schema is the same as the normal nova db schema21:48
mriedemthe non-api db21:49
dansmithright21:49
diana_clarkemriedem: ah... thanks!21:49
mriedemwe just don't populate cell0 yet21:49
mriedemcell0 is the place where failed instance builds and their records go to die21:49
dansmithkind of like florida for old people21:49
* mriedem thinks how to make a jay joke w/o offending him21:50
*** alexpilo_ has joined #openstack-nova21:51
*** smatzek has quit IRC21:52
*** rtheis has quit IRC21:53
*** whenry has quit IRC21:53
*** alexpilotti has quit IRC21:54
*** daemontool_ has quit IRC21:55
*** daemontool_ has joined #openstack-nova21:55
*** matrohon has quit IRC22:02
mriedemdansmith: why do we need cell0 again if we know which cell an instance failed to build in?22:03
dansmithmriedem: for instances that failed to schedule22:03
mriedemso it doesn't land in a cell22:04
mriedemNoValidHost?22:04
mriedemi.e. all scheduler filters rejected the build request22:04
mriedem / request spec22:04
mriedemok22:04
mriedemi need to be reminded quarterly22:04
dansmithyeah, and for things that we discover are hosed after we return from the api call,22:04
dansmithwhich could be things other than scheduling, but mostly scheduling22:04
*** whenry has joined #openstack-nova22:06
*** krtaylor has quit IRC22:07
jaypipesmriedem: nothing you could say about FL would offend me. :)22:07
*** thorst has quit IRC22:09
*** thorst has joined #openstack-nova22:09
*** WillUK2010 has quit IRC22:14
openstackgerritMatt Riedemann proposed openstack/nova: Handle MarkerNotFound from cell0 database  https://review.openstack.org/40720422:14
mriedemmelwitt: diana_clarke: andreykurilin: tdurakov: ^22:14
mriedemooo i think alaski is still core :)22:15
* mriedem fishes22:15
*** Sukhdev has joined #openstack-nova22:15
dansmithespecially since that's his code, IIRC22:16
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: [DO NOT MERGE] Check nova's revert  https://review.openstack.org/40638822:16
mriedemi'm pretty sure i was +2 or +W on the change that added that code to the compute api though22:16
mriedemandreykurilin: you don't need that anymore22:17
dansmithyeah, probably me too :)22:17
mriedemandreykurilin: https://review.openstack.org/407204 is the fix22:17
andreykurilin))22:17
mriedemandreykurilin: see https://review.openstack.org/#/c/407205/22:17
mriedemtests passed22:17
openstackgerritVladik Romanovsky proposed openstack/nova: tests: avoid starting compute service twice in sriov functional test  https://review.openstack.org/40757222:17
openstackgerritVladik Romanovsky proposed openstack/nova: tests: generate correct pci addresses for fake pci devices  https://review.openstack.org/40773022:17
andreykurilinoh22:18
andreykurilinok, will abandon my patch)22:18
*** thorst has quit IRC22:18
*** whenry has quit IRC22:18
dansmithvladikr: do we have tests for fakelibvirt?22:19
dansmithvladikr: nova/tests/unit/virt/libvirt/test_fakelibvirt.py22:19
vladikrdansmith, ah.. ok, no idea what does that do.. I was thinking about the functional tests as the test for the fakelibvirt22:20
vladikrdansmith, I can add tests to the test_fakelibvirt.py - it's just a lot of things to add22:21
dansmithvladikr: looks like that file has some basic tests, probably from good-intentioned original authors22:21
*** kristia__ has quit IRC22:21
dansmithvladikr: might be good just to make sure there's a little coverage of this path at least22:21
*** kristian__ has joined #openstack-nova22:21
vladikrdansmith, I'll look into it..22:22
vladikrdansmith, have to drive the kids to a class, will be online in 20 mins22:22
cfriesenmriedem: do you have a proposal for https://bugs.launchpad.net/nova/+bug/1527623 ?  Without the fix we can end up orphaning cinder volumes, which will require manual cleanup.22:22
openstackLaunchpad bug 1527623 in OpenStack Compute (nova) "Nova might orphan volumes when it's racing to delete a volume-backed instance" [Medium,In progress] - Assigned to ChangBo Guo(gcb) (glongwave)22:22
tdurakovmriedem: question on that fix22:22
tdurakovmriedem: is it a rootcause https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/api.py#L2240?22:22
dansmithvladikr: okay, it's late there anyway, not a huge rush, but would be cool to get these merged tomorrow22:22
tdurakovif it is, why not to fix that instead?22:23
vladikrdansmith, sure22:23
tdurakovdansmith: could you take a look also?22:23
dansmithvladikr: thanks!22:23
dansmithtdurakov: we have to raise that exception somehow22:24
mriedemtdurakov: cell0 is the same behavior as a normal cell database,22:24
*** david-lyle has joined #openstack-nova22:24
mriedemthe bug is we're looking for an instance in there by marker which doesn't exist, when listing instances22:24
mriedemif it's not in cell0, it's in the main cell db22:25
andreykurilinmriedem: btw, rally gates has a lot of servers in "ERROR" state now. I think server creation start failure at the end of last week. Rally did not change anything in own code. I did not noticed failures since rally gates were broken by keystone...22:25
mriedemwe know things might not exist in cell022:25
dansmithtdurakov: go -1 my patch that looks in more than one cell for not handling that properly22:25
dansmithtdurakov: but in the current master code, what mriedem has is right22:25
mriedemwe could move cell0 queries until after the main cell db,22:25
dansmithtdurakov: https://review.openstack.org/#/c/396775/9/nova/compute/api.py@250222:26
mriedemi'm not entirely sure why laski had the order this way, except that once we have multiple cells that mult-cell listing gets more expensive22:26
dansmithmriedem: no we can't22:26
mriedembut i think we also agreed this order at the newton midcycle22:26
*** kristian__ has quit IRC22:26
dansmithmriedem: we decided that sorting would always be build requests, cell0, then cells22:26
dansmithright22:26
mriedemyeah22:26
mriedemi just didn't want to go to the trouble of looking that up in the etherpad or ML if challenged22:27
mriedem:P22:27
dansmithheh22:27
mriedemdansmith: i left a comment in there as a reminder22:28
dansmithstolen from the hands of tdurakov !22:28
mriedemi didn't score22:28
mriedemhe can -1 it22:28
mriedempulled a classic alaski no vote22:29
dansmithtdurakov: I'd be pissed if I were you22:29
bauzasjaypipes: okay for your comment about using POST /resource_provider22:29
bauzasI just want to not use GET22:29
mriedemcfriesen: did you look at https://review.openstack.org/#/c/260339/ ?22:29
mriedemcfriesen: also that was a year ago so i've lost a lot of context on that race bug since then22:30
*** jheroux has quit IRC22:30
tdurakovdansmith: that's why I'm not going to vote on that:p what's the point if you are already know about that22:30
mriedemwould have to dig into the bug report details to remember22:30
*** lpetrut has joined #openstack-nova22:30
mriedemtdurakov: dansmith needs -1s to remember22:30
mriedemand he enjoys the pain22:30
dansmithtdurakov: as a reminder! :)22:30
dansmithright ^22:30
dansmithtdurakov: my wife has to -1 me to remind me to take out the trash and clean the dishes22:30
tdurakovlol22:31
*** eharney has quit IRC22:31
tdurakovdansmith: done, is that good?22:31
*** dave-mccowan has joined #openstack-nova22:32
dansmithtdurakov: ahhhhhh, ohhh yeahhh..22:32
mriedemhaha22:32
mriedemcleanup in aisle dan22:32
dansmithhah22:32
*** daemontool_ has quit IRC22:33
*** daemontool_ has joined #openstack-nova22:34
*** dave-mcc_ has joined #openstack-nova22:34
tdurakov<dansmith> tdurakov: ahhhhhh, ohhh yeahhh.. - is it common reaction for -1?)22:36
dansmithtdurakov: common reaction for a masochist22:36
*** ijw_ has quit IRC22:37
*** dave-mccowan has quit IRC22:37
*** ijw has joined #openstack-nova22:37
*** esberglu has quit IRC22:38
*** thorst_ has joined #openstack-nova22:38
*** thorst_ has quit IRC22:38
openstackgerritSylvain Bauza proposed openstack/nova: WIP: Expose a REST API for a specific list of RPs  https://review.openstack.org/39256922:40
openstackgerritSylvain Bauza proposed openstack/nova: Support multiple uuids and names for RP list  https://review.openstack.org/40763022:40
bauzasjaypipes, edleafe: amended my patch to use POST /requests instead of claims :)22:40
openstackgerritChris Dent proposed openstack/nova: rt: pass the nodename to public methods  https://review.openstack.org/39847022:40
openstackgerritChris Dent proposed openstack/nova: rt: remove fluff from test_resource_tracker  https://review.openstack.org/39847122:40
openstackgerritChris Dent proposed openstack/nova: rt: pass nodename to internal methods  https://review.openstack.org/39846922:40
openstackgerritChris Dent proposed openstack/nova: placement: create aggregate map in report client  https://review.openstack.org/40730922:40
openstackgerritChris Dent proposed openstack/nova: Support filtering resource providers by aggregates  https://review.openstack.org/40762922:40
openstackgerritChris Dent proposed openstack/nova: placement: correct improper test case inheritance  https://review.openstack.org/40636322:40
openstackgerritChris Dent proposed openstack/nova: rt: explicitly pass compute node to _update()  https://review.openstack.org/39847222:40
openstackgerritChris Dent proposed openstack/nova: rt: use a single ResourceTracker object instance  https://review.openstack.org/39847322:40
openstackgerritChris Dent proposed openstack/nova: HTTP interface for resource providers by aggregates  https://review.openstack.org/40774222:40
*** david-lyle has quit IRC22:41
*** david-lyle has joined #openstack-nova22:41
*** ijw has quit IRC22:42
*** harlowja has quit IRC22:43
*** mriedem has quit IRC22:43
cdentjaypipes: that stack ^ got rebased that stack to pull in the 1.2 microversion (for resource classes) because I was making a 1.3 microversion22:43
*** harlowja has joined #openstack-nova22:43
*** cmartin has quit IRC22:45
*** david-lyle has quit IRC22:46
*** cdent has quit IRC22:53
andreykurilinmriedem: http://logs.openstack.org/13/407713/1/check/gate-rally-dsvm-rally-cinder/6225c7a/logs/screen-n-cpu.txt.gz?level=ERROR it appeared last week22:54
*** tblakes has quit IRC22:55
STEWIE925hi guys, has anyone installed devstack with ubuntu 16.04?22:56
*** Swami has quit IRC22:57
STEWIE925its my first with Ubuntu 16.04, and I am getting "could not determine host ip address.  See local.conf for suggestions on setting HOST_IP22:57
STEWIE925I copied the local.conf from my working devstack - the local.conf file didnt have the host_ip line but its stacking fine.22:58
andreykurilinmriedem: also, "marker" error presents in rally gates too...22:58
STEWIE925oh, and I am installing the new devstack instance in virtualbox22:59
*** ijw has joined #openstack-nova23:01
*** jamesdenton has quit IRC23:02
*** rcernin has quit IRC23:02
*** mdrabe has quit IRC23:04
*** jamielennox is now known as jamielennox|away23:05
*** jamielennox|away is now known as jamielennox23:06
*** nic has left #openstack-nova23:07
*** nic has joined #openstack-nova23:07
*** dave-mcc_ has quit IRC23:08
*** sdague has quit IRC23:11
*** xyang1 has quit IRC23:13
*** ducttape_ has quit IRC23:14
*** jwcroppe has quit IRC23:26
*** kfarr has quit IRC23:26
*** jwcroppe has joined #openstack-nova23:26
*** jklare has quit IRC23:27
*** krtaylor has joined #openstack-nova23:30
*** jklare has joined #openstack-nova23:30
*** jwcroppe has quit IRC23:31
dansmithtonyb: ohai. can haz? https://review.openstack.org/#/c/407750/23:35
tonybdansmith: pass jenkins it must hmmm then look good it does23:36
* dansmith runs that through google translate23:37
dansmithoh, wait for jenkins?23:37
*** masber has quit IRC23:37
tonybdansmith: I'm sad that my Yoda impression didn't work :(23:37
dansmithtonyb: can haz when 'murica sleeps and jenkins is +1?23:37
tonybdansmith: Yup.23:38
dansmithtonyb: thanks23:38
*** thorst_ has joined #openstack-nova23:39
*** lpetrut has quit IRC23:42
openstackgerritSylvain Bauza proposed openstack/nova: Extend get_all_by_filters to support resource criteria  https://review.openstack.org/38624223:46
openstackgerritSylvain Bauza proposed openstack/nova: Expose a REST API for a specific list of RPs  https://review.openstack.org/39256923:46
openstackgerritSylvain Bauza proposed openstack/nova: Support multiple uuids and names for RP list  https://review.openstack.org/40763023:46
bauzasdansmith: jaypipes: edleafe: last call for https://review.openstack.org/392569, that's no longer a WIP23:47
bauzasI wrote the gabbi tests23:47
*** thorst_ has quit IRC23:48
*** ducttape_ has joined #openstack-nova23:51
openstackgerritSylvain Bauza proposed openstack/nova: Expose a REST API for a specific list of RPs  https://review.openstack.org/39256923:58
*** catintheroof has joined #openstack-nova23:58

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