Friday, 2016-08-12

*** jvgrant has quit IRC00:00
*** shaohe_feng has quit IRC00:01
*** jwcroppe has quit IRC00:01
*** shaohe_feng has joined #openstack-nova00:01
*** testing1234 has quit IRC00:03
*** kragniz1 has joined #openstack-nova00:03
*** ducttape_ has quit IRC00:05
*** sameo_ has quit IRC00:06
*** diogogmt has quit IRC00:08
*** shaohe_feng has quit IRC00:11
*** shaohe_feng has joined #openstack-nova00:12
*** ducttape_ has joined #openstack-nova00:14
*** hoangcx2 has quit IRC00:16
*** hoangcx has joined #openstack-nova00:16
*** mingyu has joined #openstack-nova00:17
*** baoli has joined #openstack-nova00:17
*** devkulkarni1 has quit IRC00:19
*** Swami has quit IRC00:20
*** devkulkarni1 has joined #openstack-nova00:20
*** shaohe_feng has quit IRC00:21
*** pooja has joined #openstack-nova00:21
*** mingyu has quit IRC00:22
*** shaohe_feng has joined #openstack-nova00:22
*** gouthamr has joined #openstack-nova00:23
poojaHi, I was having an issue when using 'nova rebuild' command as (nova/scheduler/filters/image_props_filter.py)  image_props = spec.get('image', {}).get('properties', {}) fails00:25
oomichipooja: did you submit it on the launchpad?00:26
poojaThe request_spec in scheduler filter_properties has just image id in spec.get('image') but it expects a dict, with properties00:26
poojaHi oomichi, no.. I was not sure if its a product bug00:26
poojaIs rebuild a valid operation for instances that failed to schedule with NoValidHostFound error to begin with?00:27
oomichipooja: it is nice to submit it with the details(logs, config, etc)00:27
poojasure, will file a bug00:27
oomichipooja: thanks00:27
oomichipooja: ah, NoValidHostFound error happens on that?00:28
poojaNo instance failed to spawn with NoValidHostFound and then I did rebuild after removing some constraints00:28
poojaIt failed with this error on rebuild - AttributeError: 'unicode' object has no attribute 'get'00:28
oomichipooja: that seems a bug, and good to report that00:29
poojaSure, will report it00:29
poojaIn the meantime, is there another api to reschedule a failed instance?00:30
*** ducttape_ has quit IRC00:31
*** shaohe_feng has quit IRC00:31
oomichimriedem: I guess https://review.openstack.org/#/c/351960 is easy for you00:32
oomichipooja: I guess the instance never booted at all, so isn't it fine to just "create a server" again?00:34
mriedemoomichi: why would the floating IP have a port_id but the port doesn't exist?00:34
*** shaohe_feng has joined #openstack-nova00:34
*** code-R has joined #openstack-nova00:36
poojaoomichi: I wanted to use the same server uuid since its part of a heat template, so needed a way to reschedule without the availability zone constraint on failure00:36
*** code-R_ has joined #openstack-nova00:37
*** hamzy has joined #openstack-nova00:38
*** code-R has quit IRC00:40
*** raunak has quit IRC00:40
*** shaohe_feng has quit IRC00:42
*** mlavalle has quit IRC00:42
*** shaohe_feng has joined #openstack-nova00:43
gmannalex_xu: on almost all policy tests i saw we use override policy rule with "project:  non_fake"00:44
gmannalex_xu: what exactly we want to tests there, "override rule with any string and check it fail" or "we want to override different project_id and then check if policy fail" ?00:45
*** sdake has joined #openstack-nova00:46
gmannalex_xu: like - https://github.com/openstack/nova/blob/master/nova/tests/unit/api/openstack/compute/test_admin_actions.py#L7800:46
gmannalex_xu: if our motive was to tests later one then we should change project to project_id in all tests00:47
gmannalex_xu: and if first one then current tests are fine00:47
*** devkulkarni2 has joined #openstack-nova00:48
*** devkulkarni1 has quit IRC00:48
openstackgerritGhanshyam Mann proposed openstack/nova: Add tests for user_id policy enforcement on stop  https://review.openstack.org/35333500:48
*** shaohe_feng has quit IRC00:52
*** code-R_ has quit IRC00:53
*** shaohe_feng has joined #openstack-nova00:53
*** code-R has joined #openstack-nova00:53
oomichimriedem: yeah, that was I asked you00:53
oomichimriedem: I wrote my idea on https://review.openstack.org/#/c/35196000:53
*** davidlenwell has quit IRC00:53
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server pause action  https://review.openstack.org/35222100:53
openstackgerritHieu LE proposed openstack/nova: Config options: improve libvirt help text (2)  https://review.openstack.org/31655001:01
*** rfolco has quit IRC01:02
*** shaohe_feng has quit IRC01:02
mriedemgood news, got 1 novaclient test passing on this direct glance change, only like 140+ more to go01:03
*** shaohe_feng has joined #openstack-nova01:03
*** raunak has joined #openstack-nova01:05
*** davidlenwell has joined #openstack-nova01:05
*** diogogmt has joined #openstack-nova01:05
openstackgerritMaciej Szankin proposed openstack/nova: Fix consistency in crypto conf  https://review.openstack.org/35434801:06
*** gouthamr_ has joined #openstack-nova01:06
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Return None in get_instance_id_by_floating_address  https://review.openstack.org/35196001:09
*** gouthamr has quit IRC01:10
*** vishwanathj has quit IRC01:10
oomichimriedem: thanks, ^^^ is updated01:11
oomichimriedem: btw, now novaclient is not only for nova?01:11
*** tonytan4ever has joined #openstack-nova01:11
*** mingyu has joined #openstack-nova01:11
mriedemoomichi: https://review.openstack.org/#/c/354349/01:11
mriedemno, to support 2.36 in novaclient we have to make it not rely on it's own proxy APIs01:11
mriedemwhich 404 in 2.3601:12
oomichimriedem: yeah, I saw that01:12
*** shaohe_feng has quit IRC01:12
*** _diana_ has joined #openstack-nova01:13
*** shaohe_feng has joined #openstack-nova01:13
openstackgerritMaciej Szankin proposed openstack/nova: Improve consistency in libvirt  https://review.openstack.org/35367901:13
*** baoli has quit IRC01:13
*** devkulkarni2 has quit IRC01:14
oomichimriedem: for translating "image name" <-> "image id", novaclient needs to access glance api directly instead of proxying api01:14
*** gouthamr_ is now known as gouthamr01:14
*** zhangjn has quit IRC01:14
oomichimriedem: that means novaclient is not only for nova api01:15
*** jaypipes has quit IRC01:15
oomichimriedem: maybe we(novaclient) will need to take care of glance api also01:15
openstackgerritMaciej Szankin proposed openstack/nova: Improve consistency in libvirt  https://review.openstack.org/35367901:16
*** Apoorva has quit IRC01:16
*** tonytan4ever has quit IRC01:16
*** zhangjn has joined #openstack-nova01:16
oomichimriedem: that seems the burden is moving to client side from nova itself01:16
mriedemwell, the api part of novaclient won't do that01:18
mriedemthat requires an image id01:18
mriedembut the CLI has convenience things for passing in image name and we lookup the id01:18
*** rpodolyaka_ has joined #openstack-nova01:18
mriedembut yeah we really need to just deprecate the CLI and move to osc01:18
mriedemwe're going to have to do the same for networks01:18
oomichimriedem: yeah, the translation is very useful for cli users.01:19
*** Apoorva has joined #openstack-nova01:20
*** rpodolyaka has quit IRC01:21
*** rpodolyaka_ is now known as rpodolyaka01:21
*** shaohe_feng has quit IRC01:23
*** Apoorva has quit IRC01:23
*** wangqun has joined #openstack-nova01:23
*** Apoorva has joined #openstack-nova01:23
oomichimriedem: is openstackclient alternative instead? TBH, I am not good at current situation of clients side01:24
oomichiopenstackclient is for cross-projects and it seems easy to handle this kind of thing01:25
oomichibut it doesn't support microversions anyways..01:25
mriedemright osc doesn't support microversions yet01:25
mriedemwhich is why we haven't deprecated the nova cli01:26
*** shaohe_feng has joined #openstack-nova01:26
mriedemwell, osc doesn't support microversions yet as far as i know, i don't know what the status on that is01:26
oomichimriedem: yeah, according to the code of osc, you are right.01:27
oomichithere is not any code which passes microversions on https://github.com/openstack/python-openstackclient/blob/master/openstackclient/compute/client.py01:28
*** baoli has joined #openstack-nova01:29
*** unicell has quit IRC01:29
*** EinstCrazy has joined #openstack-nova01:30
*** unicell has joined #openstack-nova01:30
* oomichi necessary to bring kids into a swiming pool just before dinner01:30
*** oomichi has quit IRC01:30
*** mingyu has quit IRC01:31
*** vishwanathj has joined #openstack-nova01:32
*** zhangjl has joined #openstack-nova01:33
*** zhangjl has left #openstack-nova01:33
*** shaohe_feng has quit IRC01:33
*** shaohe_feng has joined #openstack-nova01:34
alex_xugmann: when I write those patch, I think of later one01:34
alex_xugmann: non_fake is a project_id01:35
openstackgerritMaciej Szankin proposed openstack/nova: config options: Improve help for SPICE  https://review.openstack.org/31704601:36
gmannalex_xu: humm, may be then we need to cleanup those tests to do like  - "project_id:non_fake"01:38
*** takedakn has joined #openstack-nova01:39
gmannalex_xu: to test the right behavior on project_id things01:39
*** mingyu has joined #openstack-nova01:42
*** takedakn has quit IRC01:43
*** shaohe_feng has quit IRC01:43
*** pooja has quit IRC01:43
*** shaohe_feng has joined #openstack-nova01:44
*** unicell has quit IRC01:46
*** mingyu has quit IRC01:47
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server resize action  https://review.openstack.org/35447801:50
alex_xugmann: emm...sorry I didn't get you. project_id:non_fake already is current test to do?01:52
*** shaohe_feng has quit IRC01:53
gmannalex_xu: i think only in few tests like start stop - https://review.openstack.org/#/c/353335/5/nova/tests/unit/api/openstack/compute/test_server_start_stop.py01:53
gmannalex_xu: for other tests we might need to change project:non_fake - > project_id:non_fake01:54
*** shaohe_feng has joined #openstack-nova01:54
alex_xugmann: ah, I got it. I wrote project_id as project, that is typo mistake....01:55
gmannalex_xu: oh01:55
gmannalex_xu: but m afraid that project_id will not work for all API as we do not enforce project_id at API layer we do in DB as johnthetubaguy pointed01:56
gmannalex_xu: may be those tests will fail then01:56
alex_xugmann: emm....there must be a lot of copy/paste when write those patches...01:56
gmannalex_xu: yea, true01:57
*** _diana_ has quit IRC01:57
*** jwcroppe has joined #openstack-nova01:58
*** jwcroppe has quit IRC01:58
alex_xugmann: actually we fake the get_instance method01:58
*** jwcroppe has joined #openstack-nova01:58
*** jamesden_ has joined #openstack-nova01:59
gmannalex_xu: not for all policy tests in current situation, those are passing as policy gets failed due to 'project: non_fake' string01:59
alex_xugmann: oops, i get your words, yes, without target, that may not works01:59
gmannalex_xu: yes01:59
alex_xugmann: maybe just leave it alone, we have clear plan whether we need pass all the target in the future02:00
*** sdake has quit IRC02:01
alex_xu's/we have clear plan/after we have clear plan/'02:01
gmannalex_xu: ok, but it worth to do for those where we pass user_id and project_id at target now02:01
alex_xugmann: +102:01
gmannalex_xu: other we can cleanup later02:01
alex_xugmann: yea02:01
gmannalex_xu: ok, i will make changes with user_id patch02:01
alex_xugmann: cool02:02
gmannalex_xu: Thanks :)02:02
*** ducttape_ has joined #openstack-nova02:02
alex_xugmann: np, I will try to review them when the patch updated02:02
*** yamahata has quit IRC02:02
gmannalex_xu: Thanks. i will try to finish those by today02:02
gmannalex_xu: ll ping you once all are ready02:02
alex_xugmann: yea, cool02:03
*** shaohe_feng has quit IRC02:04
*** jamesden_ has quit IRC02:04
*** shaohe_feng has joined #openstack-nova02:04
mriedemarmax: if it makes you feel any better about us breaking neutron + LB in the gate today with the os-vif change, we also pulled that into our preprod cloud tonight and completely blew it up too :)02:05
*** ducttape_ has quit IRC02:06
*** yamahata has joined #openstack-nova02:07
*** hoangcx has quit IRC02:07
*** ivase has joined #openstack-nova02:08
alex_xumriedem: is it possiable we implement fallback for those help method https://github.com/openstack/python-novaclient/blob/master/novaclient/utils.py#L277 if this is single point which is used by CLI to access proxy api....02:08
mriedemalex_xu: yes, sdague and i have been working on that all day https://review.openstack.org/#/c/354349/02:09
alex_xumriedem: if we are going to the way get rid of all the proxy API call in the CLI, that sounds like we can't finish this in newton02:09
mriedemi'm working on cleaning up the unit tests right noww02:09
alex_xumriedem:cool!02:09
mriedemalex_xu: right now it's looking like image and network, and just the proxy for GET on name translation02:09
mriedemi.e. image name / net name given, we look up the resource to get the id to pass to the REST API02:10
*** ivase has quit IRC02:10
mriedemwe require volume IDs so no cinder proxying02:10
*** ivase has joined #openstack-nova02:10
mriedemthe network thing is going to be a bit different since it's going to be conditional on the 2.36 microversion02:10
mriedemfor < 2.36 we call the nova proxy api02:10
mriedem>= 2.36 we call neutron02:10
mriedembut sdague and i haven't gotten deep into the details of that yet, need to get the image stuff working first02:11
alex_xumriedem: got it, so the plan is call the neutron API directly?02:11
mriedemyeah02:12
*** sdake has joined #openstack-nova02:13
*** jamesdenton has joined #openstack-nova02:13
*** shaohe_feng has quit IRC02:14
alex_xumriedem: ok, got it, not sure how complex of that, anyway I will read that patch and think about it more.02:14
*** Apoorva_ has joined #openstack-nova02:14
*** shaohe_feng has joined #openstack-nova02:15
*** Apoorva_ has quit IRC02:15
*** raunak has quit IRC02:15
mriedemok, i hope to have unit tests passing before i go to bed,02:16
mriedemi'll push up whatever WIP i have before then anyway02:16
*** gongysh has joined #openstack-nova02:17
*** gyee has quit IRC02:17
*** Apoorva has quit IRC02:17
*** woodster_ has quit IRC02:19
armaxmriedem: not quite, it’s good to know we’re not the only ones experiencing grief :)02:20
*** ljxiash_ has joined #openstack-nova02:21
mriedemthe fix is top of gate queue at least02:22
*** shaohe_feng has quit IRC02:24
*** shaohe_feng has joined #openstack-nova02:25
*** zerda2 has joined #openstack-nova02:28
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for force_delete server  https://review.openstack.org/35448502:29
*** bswartz has quit IRC02:33
*** shaohe_feng has quit IRC02:34
*** sdake has quit IRC02:34
*** shaohe_feng has joined #openstack-nova02:35
*** bswartz has joined #openstack-nova02:37
openstackgerritGhanshyam Mann proposed openstack/nova: Add tests for user_id policy enforcement on stop  https://review.openstack.org/35333502:37
openstackgerritChangBo Guo(gcb) proposed openstack/nova: Don't pass argument sqlite_db in method set_defaults  https://review.openstack.org/35094702:39
*** EinstCrazy has quit IRC02:39
*** hoangcx has joined #openstack-nova02:40
openstackgerritGhanshyam Mann proposed openstack/nova: Add tests for user_id policy enforcement on stop  https://review.openstack.org/35333502:40
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server pause action  https://review.openstack.org/35222102:40
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server resize action  https://review.openstack.org/35447802:41
*** EinstCrazy has joined #openstack-nova02:43
*** shaohe_feng has quit IRC02:45
*** shaohe_feng has joined #openstack-nova02:45
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server resize action  https://review.openstack.org/35447802:48
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for force_delete server  https://review.openstack.org/35448502:48
*** ayoung has joined #openstack-nova02:49
*** sdake has joined #openstack-nova02:54
*** shaohe_feng has quit IRC02:55
*** betherly has quit IRC02:55
*** zhiyan has quit IRC02:55
*** johnsom has quit IRC02:55
*** tdurakov has quit IRC02:56
*** ameade has quit IRC02:56
*** shaohe_feng has joined #openstack-nova02:56
*** ijw has quit IRC02:59
*** betherly has joined #openstack-nova02:59
*** johnsom has joined #openstack-nova03:00
*** tdurakov has joined #openstack-nova03:00
*** amotoki has joined #openstack-nova03:02
*** ducttape_ has joined #openstack-nova03:03
*** hoangcx has quit IRC03:03
*** zhiyan has joined #openstack-nova03:03
*** ameade has joined #openstack-nova03:03
*** diogogmt has quit IRC03:05
*** shaohe_feng has quit IRC03:05
*** shaohe_feng has joined #openstack-nova03:06
*** sudipto has joined #openstack-nova03:06
*** sudipto_ has joined #openstack-nova03:06
*** ducttape_ has quit IRC03:07
*** amotoki has quit IRC03:11
openstackgerritJohn Hua proposed openstack/nova: Use physical utilisation for cached images  https://review.openstack.org/35386203:13
*** raunak has joined #openstack-nova03:13
*** yamahata has quit IRC03:14
*** shaohe_feng has quit IRC03:15
*** shaohe_feng has joined #openstack-nova03:16
*** amotoki has joined #openstack-nova03:22
*** sdake has quit IRC03:24
*** shaohe_feng has quit IRC03:26
*** shaohe_feng has joined #openstack-nova03:26
*** EinstCra_ has joined #openstack-nova03:29
*** amotoki has quit IRC03:31
*** EinstCrazy has quit IRC03:32
*** links has joined #openstack-nova03:33
*** gouthamr has quit IRC03:35
*** amotoki has joined #openstack-nova03:35
*** shaohe_feng has quit IRC03:36
*** shaohe_feng has joined #openstack-nova03:36
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server shelve action  https://review.openstack.org/35449903:37
*** baoli has quit IRC03:43
*** sudipto_ has quit IRC03:43
*** sudipto has quit IRC03:43
*** ayoung has quit IRC03:45
*** shaohe_feng has quit IRC03:46
*** shaohe_feng has joined #openstack-nova03:47
openstackgerritPatrick East proposed openstack/nova: Add release note to warn about os-brick lock dir  https://review.openstack.org/35450203:47
*** raunak has quit IRC03:47
*** liverpooler has quit IRC03:49
*** raunak has joined #openstack-nova03:49
*** hoangcx has joined #openstack-nova03:51
*** sdake has joined #openstack-nova03:54
*** shaohe_feng has quit IRC03:56
*** shaohe_feng has joined #openstack-nova03:57
*** hoangcx has quit IRC03:57
*** amotoki has quit IRC04:01
*** ducttape_ has joined #openstack-nova04:03
*** vishwanathj has quit IRC04:06
*** shaohe_feng has quit IRC04:07
*** shaohe_feng has joined #openstack-nova04:07
*** ducttape_ has quit IRC04:08
*** rkmrhj has joined #openstack-nova04:09
*** Sukhdev has joined #openstack-nova04:09
*** dikonoor has joined #openstack-nova04:09
*** amotoki has joined #openstack-nova04:09
*** amotoki has quit IRC04:10
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Look up image names directly in glance  https://review.openstack.org/35434904:11
mriedemomfg yes ^04:11
*** amotoki has joined #openstack-nova04:11
*** tonytan4ever has joined #openstack-nova04:12
*** fragatina has quit IRC04:16
*** mingyu has joined #openstack-nova04:16
*** tonytan4ever has quit IRC04:17
*** shaohe_feng has quit IRC04:17
*** shaohe_feng has joined #openstack-nova04:17
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Look up image names directly in glance  https://review.openstack.org/35434904:19
*** ljxiash_ has quit IRC04:22
prometheanfireoh god04:22
prometheanfireplease04:22
prometheanfiredo that one04:22
mriedemha04:23
mriedemprometheanfire: now you're reminding me of the neutronclient 5.1.0 issue04:23
prometheanfiremriedem: :(04:24
prometheanfire:P04:24
*** vishwanathj has joined #openstack-nova04:26
*** vishwanathj has quit IRC04:26
*** gongysh has quit IRC04:26
*** mingyu has quit IRC04:26
*** shaohe_feng has quit IRC04:27
mriedemdo you have a link to that again?04:27
mriedemnvm https://review.openstack.org/#/c/353443/04:27
*** shaohe_feng has joined #openstack-nova04:28
mriedemwell, i don't know where the test results were04:29
mriedemis it a separate patch?04:29
mriedemoh nvm04:29
mriedemhttp://logs.openstack.org/43/353443/2/check/gate-cross-nova-python27-db-ubuntu-xenial/2b8780d/04:29
mriedemthis is what broke it https://review.openstack.org/#/c/348096/04:32
*** code-R_ has joined #openstack-nova04:32
prometheanfiresorry, distracted04:32
openstackgerritGhanshyam Mann proposed openstack/nova: Add tests for user_id policy enforcement on trigger_crash_dump  https://review.openstack.org/35451004:32
prometheanfireplaying no mans sky :P04:32
*** jamesdenton has quit IRC04:33
*** code-R_ has quit IRC04:34
*** code-R_ has joined #openstack-nova04:35
*** code-R has quit IRC04:35
*** gongysh has joined #openstack-nova04:36
*** shaohe_feng has quit IRC04:37
*** hshiina has joined #openstack-nova04:37
*** shaohe_feng has joined #openstack-nova04:38
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/35445904:38
*** takedakn has joined #openstack-nova04:39
*** sudipto has joined #openstack-nova04:43
*** sudipto_ has joined #openstack-nova04:43
*** takedakn has quit IRC04:43
*** ratailor has joined #openstack-nova04:45
*** sudipto_ has quit IRC04:46
*** sudipto has quit IRC04:46
*** shaohe_feng has quit IRC04:48
*** shaohe_feng has joined #openstack-nova04:48
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/35445904:53
*** Sukhdev has quit IRC04:54
*** shaohe_feng has quit IRC04:58
*** shaohe_feng has joined #openstack-nova04:58
*** fragatina has joined #openstack-nova04:58
*** fragatina has quit IRC05:03
*** hoangcx has joined #openstack-nova05:03
*** jaosorior has joined #openstack-nova05:04
*** hoangcx has quit IRC05:06
*** shaohe_feng has quit IRC05:08
*** prateek_ has joined #openstack-nova05:10
*** ivase has quit IRC05:10
*** ivase has joined #openstack-nova05:10
*** prateek_ has quit IRC05:11
*** shaohe_feng has joined #openstack-nova05:11
*** prateek has joined #openstack-nova05:11
*** jwcroppe has quit IRC05:11
*** jwcroppe has joined #openstack-nova05:12
rkmrhjmacsz: Sorry for late at night. Have you recheck my bug report(#1587285) that I heard yesterday?05:13
*** mriedem has quit IRC05:14
openstackgerritChangBo Guo(gcb) proposed openstack/nova: Log warning when user set improper config option value  https://review.openstack.org/35453205:15
macszrkmrhj: oh sorry, I forgot :( adding reminder for tomorrow. Sorry! :(05:15
*** chandankumar has joined #openstack-nova05:16
*** jwcroppe has quit IRC05:16
*** shaohe_feng has quit IRC05:18
*** shaohe_feng has joined #openstack-nova05:19
*** shaohe_feng has quit IRC05:29
*** vishwanathj has joined #openstack-nova05:29
*** shaohe_feng has joined #openstack-nova05:29
*** vishwanathj has quit IRC05:29
*** sdake has quit IRC05:32
*** raunak has quit IRC05:33
*** coolsvap has joined #openstack-nova05:34
*** unicell has joined #openstack-nova05:35
rkmrhjmacsz: Thanks a lot. If you can't recheck, please inherit rechecking to next bug skimming duty member.05:37
*** shaohe_feng has quit IRC05:39
*** shaohe_feng has joined #openstack-nova05:39
*** amotoki_ has joined #openstack-nova05:40
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server changePassword action  https://review.openstack.org/35453705:40
*** pbandark has joined #openstack-nova05:41
*** amotoki has quit IRC05:43
*** rama_y has joined #openstack-nova05:44
*** nkrinner_afk is now known as nkrinner05:45
*** unicell has quit IRC05:49
*** unicell has joined #openstack-nova05:49
*** shaohe_feng has quit IRC05:49
openstackgerritRajesh Tailor proposed openstack/nova: Fix race for compute node create record, on compute-start  https://review.openstack.org/30441505:49
*** unicell has quit IRC05:50
*** shaohe_feng has joined #openstack-nova05:50
*** unicell has joined #openstack-nova05:50
*** rcernin has quit IRC05:51
*** code-R has joined #openstack-nova05:53
*** code-R_ has quit IRC05:56
*** shaohe_feng has quit IRC05:59
*** ivase_ has joined #openstack-nova06:00
*** shaohe_feng has joined #openstack-nova06:00
*** ivase has quit IRC06:02
*** coreywright has quit IRC06:02
*** esracelik has joined #openstack-nova06:02
*** esracelik is now known as ecelik06:02
*** armax has quit IRC06:05
*** ducttape_ has joined #openstack-nova06:05
*** shaohe_feng has quit IRC06:10
*** ducttape_ has quit IRC06:10
*** shaohe_feng has joined #openstack-nova06:10
*** code-R has quit IRC06:11
*** code-R has joined #openstack-nova06:11
*** shaohe_feng has quit IRC06:20
*** shaohe_feng has joined #openstack-nova06:21
*** rcernin has joined #openstack-nova06:21
*** oomichi has joined #openstack-nova06:30
*** shaohe_feng has quit IRC06:30
*** shaohe_feng has joined #openstack-nova06:31
*** ljxiash has joined #openstack-nova06:31
*** code-R_ has joined #openstack-nova06:34
*** code-R has quit IRC06:34
*** shaohe_feng has quit IRC06:40
*** nkrinner has quit IRC06:41
*** shaohe_feng has joined #openstack-nova06:41
*** nkrinner has joined #openstack-nova06:42
*** nkrinner has quit IRC06:42
*** nkrinner has joined #openstack-nova06:42
*** coreywright has joined #openstack-nova06:43
*** diga_ has joined #openstack-nova06:44
*** ljxiash_ has joined #openstack-nova06:44
*** ivase_ has quit IRC06:45
*** ivase has joined #openstack-nova06:45
*** ljxiash has quit IRC06:46
*** zhangjn has quit IRC06:48
*** zhangjn has joined #openstack-nova06:49
*** nkrinner is now known as nkrinner_afk06:49
*** ecelik has quit IRC06:50
*** shaohe_feng has quit IRC06:51
*** shaohe_feng has joined #openstack-nova06:51
*** kevinz has joined #openstack-nova06:54
*** GB21 has joined #openstack-nova06:58
*** tesseract- has joined #openstack-nova06:59
*** shaohe_feng has quit IRC07:01
*** shaohe_feng has joined #openstack-nova07:01
*** ljxiash_ has quit IRC07:02
*** esracelik has joined #openstack-nova07:03
*** ljxiash_ has joined #openstack-nova07:04
*** ducttape_ has joined #openstack-nova07:06
*** esracelik is now known as ecelik07:06
*** daemontool has joined #openstack-nova07:06
openstackgerritJohn Hua proposed openstack/nova: Use physical utilisation for cached images  https://review.openstack.org/35386207:06
*** oanson has joined #openstack-nova07:07
*** mingyu has joined #openstack-nova07:08
*** ljxiash_ has quit IRC07:08
*** shaohe_feng has quit IRC07:11
*** shaohe_feng has joined #openstack-nova07:11
*** ducttape_ has quit IRC07:11
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server delete  https://review.openstack.org/35455507:13
*** oomichi has quit IRC07:14
*** rkrum has quit IRC07:16
openstackgerritxhzhf proposed openstack/nova: [sr-iov] pci passthrough whitelist doesn't support mult node  https://review.openstack.org/35455707:17
*** yamahata has joined #openstack-nova07:19
*** shaohe_feng has quit IRC07:21
*** GB21 has quit IRC07:22
*** mingyu has quit IRC07:22
*** shaohe_feng has joined #openstack-nova07:22
*** markus_z has joined #openstack-nova07:24
*** sridharg has joined #openstack-nova07:24
*** ljxiash has joined #openstack-nova07:27
*** RuiChen has joined #openstack-nova07:28
*** nkrinner_afk is now known as nkrinner07:30
*** shaohe_feng has quit IRC07:32
*** f10d4 has quit IRC07:33
*** markvoelker has quit IRC07:34
*** shaohe_feng has joined #openstack-nova07:34
*** jichen has joined #openstack-nova07:36
*** ljxiash has quit IRC07:37
*** jamielennox|away has quit IRC07:38
*** shaohe_feng has quit IRC07:42
*** shaohe_feng has joined #openstack-nova07:43
*** f10d4 has joined #openstack-nova07:46
openstackgerritxhzhf proposed openstack/nova: [vncproxy] log for closing web is misleading  https://review.openstack.org/35380707:46
openstackgerritxhzhf proposed openstack/nova: [sr-iov] pci passthrough whitelist doesn't support multi node  https://review.openstack.org/35455707:47
*** jamielennox|away has joined #openstack-nova07:47
*** jamielennox|away is now known as jamielennox07:48
*** shaohe_feng has quit IRC07:52
*** shaohe_feng has joined #openstack-nova07:52
*** yamahata has quit IRC07:55
*** tonytan4ever has joined #openstack-nova07:59
*** zzzeek has quit IRC08:00
*** zzzeek has joined #openstack-nova08:01
*** shaohe_feng has quit IRC08:02
*** shaohe_feng has joined #openstack-nova08:03
*** tonytan4ever has quit IRC08:03
openstackgerritJohn Hua proposed openstack/nova: Use physical utilisation for cached images  https://review.openstack.org/35386208:06
*** matrohon has joined #openstack-nova08:07
*** ducttape_ has joined #openstack-nova08:08
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: [WIP] Testing latest u-c  https://review.openstack.org/30334508:10
openstackgerritJohn Garbutt proposed openstack/nova: Verified deprecation status for vnc options  https://review.openstack.org/35366608:11
*** ducttape_ has quit IRC08:12
*** shaohe_feng has quit IRC08:13
*** code-R_ has quit IRC08:13
*** shaohe_feng has joined #openstack-nova08:13
*** lucas-dinner is now known as lucasagomes08:15
johnthetubaguyalex_xu: gmann: just looking at the policy stuff08:15
johnthetubaguyI think we can loop back on that servers one later08:15
*** ljxiash has joined #openstack-nova08:16
gmannjohnthetubaguy: did not get08:16
johnthetubaguygmann: I was really meaning, lets just merge the current version of your patch: https://review.openstack.org/#/c/353335/708:17
johnthetubaguyI was talking a bit to sdague yesterday about that one08:17
johnthetubaguylets come back later and fix up the other policy stuff in there, so it only passes the project_id08:17
johnthetubaguygmann: does that make more sense?08:18
alex_xujohnthetubaguy: got it, +108:18
gmannjohnthetubaguy: ah, you mean fix for project_id one08:18
alex_xugmann: I guess johnthetubaguy pointed to my comment?08:19
johnthetubaguyyeah, there are quite a few non-standard policy checks in that file, lets worry about that as follow up bug fixes08:19
johnthetubaguyalex_xu: yeah, I think I had similar comments on an earlier patchset that sdague put up08:19
gmannjohnthetubaguy: alex_xu i see.08:20
*** rmart04 has joined #openstack-nova08:20
alex_xujohnthetubaguy: gmann just +w :)08:20
gmannjohnthetubaguy: alex_xu  yea we can cleanup those later.08:20
johnthetubaguythe deadline is looking real soon, is the summary08:20
alex_xujohnthetubaguy: question about https://review.openstack.org/#/c/353666/3/nova/conf/vnc.py, those 'deprecated_group' already be there one release, should we remove them?08:20
gmannjohnthetubaguy: alex_xu anyways i will last patch to remove user_id from start server which will remove those08:21
*** danpb has joined #openstack-nova08:21
alex_xugmann: cool, thanks08:21
johnthetubaguyalex_xu: basically, trying to do the minimum change at this point08:21
johnthetubaguyalex_xu: I hoped we would have a tool out this cycle to help admins detect when they use deprecated names and groups, so its easier to fix up their config08:21
*** GB21 has joined #openstack-nova08:21
*** oanson has quit IRC08:21
johnthetubaguyalex_xu: its super cheap to leave them there till we get that tool sorted08:22
gmannjohnthetubaguy: alex_xu i left with update and evacuate one, ll post soon08:22
johnthetubaguyalex_xu: turns out oslo can do multiple alias names now, anyways08:22
*** ljxiash has quit IRC08:22
*** shaohe_feng has quit IRC08:23
*** shaohe_feng has joined #openstack-nova08:23
alex_xujohnthetubaguy: ok, got it, the patch title is a little confuse. should we keep the 'needs:check_deprecation_status 'flag'?08:24
johnthetubaguyalex_xu: the plan is to drop all those tags by the end of the cycle08:25
johnthetubaguyalex_xu: mostly its about spotting unused or still things, and adding a TODO to deprecate it next cycle08:25
alex_xujohnthetubaguy: ok, cool, so just follow the plan :)08:25
*** sameo has joined #openstack-nova08:26
johnthetubaguyalex_xu: yeah, changed the plan to reduce the review load, since its this late in the cycle08:26
*** carl_baldwin_ has joined #openstack-nova08:31
*** shaohe_feng has quit IRC08:33
*** shaohe_feng has joined #openstack-nova08:34
*** markvoelker has joined #openstack-nova08:34
*** carl_baldwin has quit IRC08:35
*** carl_baldwin_ is now known as carl_baldwin08:35
*** oanson has joined #openstack-nova08:36
*** markvoelker has quit IRC08:39
*** cdent has joined #openstack-nova08:39
*** ljxiash has joined #openstack-nova08:40
*** shaohe_feng has quit IRC08:43
Dinesh_BhorHi all, is there a documentation on how to configure SSL with nova ? If anyone is having , please share.08:44
*** shaohe_feng has joined #openstack-nova08:44
johnthetubaguyDinesh_Bhor: please ask on #openstack for operator questsions, see channel topic08:45
johnthetubaguyDinesh_Bhor: having said that, I think you want this doc: http://docs.openstack.org/security-guide/api-endpoints.html08:46
*** ihrachys has joined #openstack-nova08:47
Dinesh_Bhorjohnthetubaguy: ok, Thank you.08:48
*** shaohe_feng has quit IRC08:54
*** ljxiash has quit IRC08:54
*** mnestratov|3 has joined #openstack-nova08:54
*** shaohe_feng has joined #openstack-nova08:54
johnthetubaguygmann: not sure about the unit tests in this one: https://review.openstack.org/#/c/354537/108:54
*** mnestratov|3 has quit IRC08:54
*** ljxiash has joined #openstack-nova08:54
*** gszasz has joined #openstack-nova08:54
*** hshiina has quit IRC08:55
gmannjohnthetubaguy: actually i have to use separate req as same project_id is being passed in get_instance stub if we use self.req08:57
gmannjohnthetubaguy: i want to override the project_id in req and diff than instance08:57
gmannjohnthetubaguy: I did that to avoid mock.patch.object in each tests and do stub in setUp08:58
*** ljxiash has quit IRC08:59
johnthetubaguygmann: can't you just update self.req just before passing that into the next method?08:59
johnthetubaguyi.e. I think line190 just goes to self.req09:00
gmannjohnthetubaguy: i can but sme updated one will be passed to stub method and fake instance get created with that09:00
johnthetubaguygmann: but that has already executed before your test has started09:00
johnthetubaguy(using the old value)09:00
*** code-R has joined #openstack-nova09:01
johnthetubaguyhmm, its a bit nasty either way I guess09:01
gmannjohnthetubaguy: but that callable function is stubbed09:01
johnthetubaguygmann: oh... I see what you mean09:01
*** cdent has quit IRC09:02
johnthetubaguygmann: OK, cools, seems the simplest way to do, miss read that execution order09:02
gmannjohnthetubaguy: yea, we have lot of hoch poch in unit test actually :)09:02
gmanntrying to do minimum changes09:03
johnthetubaguyyeah, agreed09:03
johnthetubaguywe should do something to help this policy stuff as a follow up09:03
* johnthetubaguy has to run off for a little bit09:03
*** shaohe_feng has quit IRC09:04
*** shaohe_feng has joined #openstack-nova09:05
*** code-R has quit IRC09:05
openstackgerritGleb Stepanov proposed openstack/nova: [WIP] Init allocated ports set  https://review.openstack.org/34747909:06
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server changePassword action  https://review.openstack.org/35453709:07
*** ducttape_ has joined #openstack-nova09:09
*** aloga has quit IRC09:10
*** aloga has joined #openstack-nova09:11
openstackgerritGleb Stepanov proposed openstack/nova: Refactor aggregate image properties  https://review.openstack.org/35356809:11
*** ducttape_ has quit IRC09:13
*** shaohe_feng has quit IRC09:14
*** shaohe_feng has joined #openstack-nova09:15
*** cdent has joined #openstack-nova09:17
*** amotoki has joined #openstack-nova09:20
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server update  https://review.openstack.org/35461009:23
*** amotoki_ has quit IRC09:23
gmannjohnthetubaguy: in this update one, i have to do lot of updates in existing tests for project_id, user_id in stub instance - https://review.openstack.org/35461009:24
*** shaohe_feng has quit IRC09:24
openstackgerritAlvaro Lopez Garcia proposed openstack/nova: Ensure that periodic reclaim cleans DB deleted instances  https://review.openstack.org/32325009:25
*** shaohe_feng has joined #openstack-nova09:25
markus_ztdurakov: Got a minute? My live_migration_serial_console test gets executed due to the hook (so far so good) but it fails and would like to see the logs but don't find them.09:25
markus_ztdurakov: The "console.log" says the "n-sproxy.log" is available: http://logs.openstack.org/71/347471/14/check/gate-tempest-dsvm-multinode-live-migration/fc74b36/console.html#_2016-08-11_15_15_53_91376309:27
markus_zbut I don't find them in the test run09:28
markus_zhttp://logs.openstack.org/71/347471/14/check/gate-tempest-dsvm-multinode-live-migration/fc74b36/09:28
tdurakovmarkus_z: hi, let me check09:29
tdurakovI meAn09:29
tdurakovs/ I mean 2 minutes09:29
*** shaohe_feng has quit IRC09:35
*** shaohe_feng has joined #openstack-nova09:36
*** markvoelker has joined #openstack-nova09:36
*** rama_y has quit IRC09:37
openstackgerritRajesh Tailor proposed openstack/nova: Fix race for compute node create record, on compute-start  https://review.openstack.org/30441509:37
*** markvoelker has quit IRC09:40
*** amotoki_ has joined #openstack-nova09:41
openstackgerritxhzhf proposed openstack/nova: [network] nova with neutron can not inject ip address to vm  https://review.openstack.org/35461609:42
openstackgerritRikimaru Honjo proposed openstack/nova: Add checks for preventing unexpected detaching volume  https://review.openstack.org/35461709:42
*** andrey-mp has joined #openstack-nova09:42
*** amotoki has quit IRC09:44
*** ivase has quit IRC09:45
*** shaohe_feng has quit IRC09:45
*** shaohe_feng has joined #openstack-nova09:45
*** ivase has joined #openstack-nova09:45
*** sdague has joined #openstack-nova09:46
*** matrohon has quit IRC09:48
andrey-mphi, can anyone check this patchset - https://review.openstack.org/#/c/352455/2   This reverts quota's behaviour in nova...09:48
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Updated from global requirements  https://review.openstack.org/35445909:49
*** lpetrut has joined #openstack-nova09:50
*** code-R has joined #openstack-nova09:55
*** shaohe_feng has quit IRC09:55
*** shaohe_feng has joined #openstack-nova09:56
*** TRUMP-PENCE has joined #openstack-nova09:58
*** code-R has quit IRC10:00
tdurakovmarkus_z: the thing is that hook actually failed10:01
tdurakovI'll left comment10:01
tdurakovmarkus_z: done10:04
*** EinstCra_ has quit IRC10:05
*** shaohe_feng has quit IRC10:05
*** shaohe_feng has joined #openstack-nova10:06
*** gongysh has quit IRC10:08
*** mnestratov has quit IRC10:09
*** ducttape_ has joined #openstack-nova10:09
*** lpetrut1 has joined #openstack-nova10:09
openstackgerritDavanum Srinivas (dims) proposed openstack/nova: [WIP] Testing latest u-c  https://review.openstack.org/30334510:10
*** mnestratov has joined #openstack-nova10:10
*** lpetrut has quit IRC10:11
*** lpetrut1 is now known as lpetrut10:11
*** rkmrhj has left #openstack-nova10:12
*** coolsvap is now known as _coolsvap_10:13
*** ducttape_ has quit IRC10:14
*** mvk has quit IRC10:14
*** mvk has joined #openstack-nova10:14
*** amotoki_ has quit IRC10:15
*** shaohe_feng has quit IRC10:16
openstackgerritPrateek Arora proposed openstack/nova: Handle ConnectFailure from keystone when neutron is down  https://review.openstack.org/32770610:16
*** shaohe_feng has joined #openstack-nova10:17
*** Rodrigo_BR has quit IRC10:23
*** shaohe_feng has quit IRC10:26
*** shaohe_feng has joined #openstack-nova10:27
*** amotoki has joined #openstack-nova10:27
markus_ztdurakov: The "console.log" says it executed the tempest tests. All of them because I didn't specify the regex to limit the amount of tests.10:29
markus_ztdurakov: IOW, I think I can use "run_tempest" without an regex.10:30
markus_ztdurakov: Therefore I don't understand what you mean with "the hook failed"10:30
hieulqjohnthetubaguy: markus_z can you take a look at this patch https://review.openstack.org/#/c/316550/10:30
openstackgerritSean Dague proposed openstack/python-novaclient: Make novaclient functional tests use pretty tox  https://review.openstack.org/35463810:34
markus_zhieulq: yep, will do that today10:34
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server delete  https://review.openstack.org/35455510:35
*** shaohe_feng has quit IRC10:36
*** shaohe_feng has joined #openstack-nova10:37
*** wangqun has quit IRC10:39
*** daemontool has quit IRC10:39
hieulqmarkus_z: thanks10:40
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server delete  https://review.openstack.org/35455510:40
*** daemontool has joined #openstack-nova10:41
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server update  https://review.openstack.org/35461010:41
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server update  https://review.openstack.org/35461010:41
*** shaohe_feng has quit IRC10:46
*** daemontool_ has joined #openstack-nova10:47
*** shaohe_feng has joined #openstack-nova10:47
*** daemontool has quit IRC10:47
*** ivase has quit IRC10:48
gmannjohnthetubaguy: sdague alex_xu : we want to enforce user_id on evacuate also right? as this is already admin only10:49
*** code-R has joined #openstack-nova10:49
sdaguegmann: yes, it's on the list10:49
gmannsdague: ok10:49
sdaguegmann: remember, this only takes effect if people are changing their policy file10:49
sdagueso it being admin only in policy doesn't mean anything10:49
gmannsdague: but it need admin password also for evacuation10:50
sdaguegmann: link?10:50
gmannsdague:  https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/evacuate.py#L5110:51
gmannsdague: ah, if no admin password then we generate one and proceed10:51
*** kragniz1 has quit IRC10:52
sdagueright, that's setting the adminPass on the guest10:53
*** kragniz1 has joined #openstack-nova10:53
sdagueit's not that we're checking against it10:53
*** code-R has quit IRC10:53
gmannsdague: yup, on new rebuilt instance10:53
gmannsdague: make sense. doing for that..10:53
*** shaohe_feng has quit IRC10:57
*** shaohe_feng has joined #openstack-nova10:57
*** kragniz1 has quit IRC10:59
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: live-migration test hook for serial console  https://review.openstack.org/34747111:04
*** mnestratov has quit IRC11:05
*** JoseMello has joined #openstack-nova11:06
*** shaohe_feng has quit IRC11:07
*** shaohe_feng has joined #openstack-nova11:07
*** ducttape_ has joined #openstack-nova11:10
openstackgerritChen Fan proposed openstack/nova: context: change the name 'rule' to 'action' in context.can  https://review.openstack.org/34765111:13
*** ducttape_ has quit IRC11:14
*** kragniz1 has joined #openstack-nova11:15
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server evacuate  https://review.openstack.org/35465611:16
*** shaohe_feng has quit IRC11:17
*** kragniz1 has quit IRC11:18
*** kragniz1 has joined #openstack-nova11:18
*** shaohe_feng has joined #openstack-nova11:20
*** RuiChen has quit IRC11:25
*** shaohe_feng has quit IRC11:27
*** shaohe_feng has joined #openstack-nova11:28
*** GB21 has quit IRC11:28
*** jaosorior has quit IRC11:29
*** manous has joined #openstack-nova11:29
*** jaosorior has joined #openstack-nova11:30
*** amotoki has quit IRC11:30
manoushi11:31
manousplease how can is solve this on mitaka ? http://paste.openstack.org/show/555955/11:32
openstackgerritGhanshyam Mann proposed openstack/nova: Remove user_id policy enforcement from server start  https://review.openstack.org/35466211:32
*** kragniz is now known as Guest6666611:36
*** kragniz1 is now known as kragniz11:36
*** markvoelker has joined #openstack-nova11:37
*** shaohe_feng has quit IRC11:38
*** GB21 has joined #openstack-nova11:38
*** shaohe_feng has joined #openstack-nova11:38
*** jistr has quit IRC11:39
prateekmanous, can you please turn on the debug logs and then try11:40
*** f10d4 has quit IRC11:40
*** f10d4 has joined #openstack-nova11:40
manousoka11:40
*** dansmith has quit IRC11:40
prateekmanous, pastebin the glance logs after that11:40
*** StevenK has quit IRC11:41
gmannsdague: johnthetubaguy apart from action listed in spec we have user_id things in 3 more place11:41
*** amotoki has joined #openstack-nova11:41
gmannsdague: johnthetubaguy : unlock_override and instance_action-> index and show11:41
manousglance image-list work fine11:41
*** dansmith has joined #openstack-nova11:41
*** dansmith is now known as Guest1395711:42
*** markvoelker has quit IRC11:42
tdurakovmarkus_z: it looks like job just hasn't finished in time and was terminated11:42
*** StevenK has joined #openstack-nova11:42
gmannsdague: johnthetubaguy : https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/lock_server.py#L5111:42
gmannsdague: johnthetubaguy https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/instance_actions.py#L66-L7611:42
markus_ztdurakov: ohhh, so it didn't have the chance to collect the data?11:43
*** zerda2 has quit IRC11:43
prateekmanous, start with the following logs, nova-api.log, nova-conductor.log and then glance logs, find the exception and paste bin them11:43
*** code-R has joined #openstack-nova11:43
markus_ztdurakov: I pushed a new ps which hopefully gives me more insight why things are failing. It also has the regex you suggested.11:43
tdurakovmarkus_z: possible,  http://logs.openstack.org/71/347471/14/check/gate-tempest-dsvm-multinode-live-migration/fc74b36/console.html#_2016-08-11_15_18_49_967318 - tests are failed, so that's the reason I think we should limit tests only live-migration11:44
*** ftersin has joined #openstack-nova11:44
tdurakovyeah, saw it, let's just wait for results11:44
ftersinmdbooth: hi, are you around?11:45
markus_ztdurakov: agreed, executing a subset should be a good trade-off between finding issues and burning CI resources.11:45
* markus_z waits patiently11:46
*** snikitin has quit IRC11:47
*** shaohe_feng has quit IRC11:48
*** gouthamr has joined #openstack-nova11:48
*** code-R has quit IRC11:48
*** shaohe_feng has joined #openstack-nova11:49
*** mnestratov has joined #openstack-nova11:56
*** ratailor has quit IRC11:57
*** tblakes has joined #openstack-nova11:58
*** shaohe_feng has quit IRC11:58
*** gongysh has joined #openstack-nova11:59
*** tonytan4ever has joined #openstack-nova12:00
*** lucasagomes is now known as lucas-hungry12:00
*** shaohe_feng has joined #openstack-nova12:01
*** jaypipes has joined #openstack-nova12:02
*** tonytan4ever has quit IRC12:05
*** diga_ has quit IRC12:05
*** alaski is now known as lascii12:05
jroll/b 17912:07
jrolloops12:07
*** GB21 has quit IRC12:08
*** shaohe_feng has quit IRC12:08
*** prateek has quit IRC12:08
*** shaohe_feng has joined #openstack-nova12:09
*** amotoki has quit IRC12:10
*** ducttape_ has joined #openstack-nova12:11
openstackgerritGhanshyam Mann proposed openstack/nova: Fix policy tests for project_id enforcement  https://review.openstack.org/35469012:11
gmannsdague: johnthetubaguy alex_xu : all patches for policy things are up - https://review.openstack.org/#/q/project:openstack/nova+branch:master+topic:bp/user_id_based_policy_enforcement12:13
*** xyang1 has joined #openstack-nova12:14
gmannsdague: johnthetubaguy alex_xu : if needed we can do for  these 3 place also (at least passing user_id, project_id explicitly in target)- unlock_override and instance_action-> index and show12:14
gmannsdague: johnthetubaguy alex_xu : currently complete instance is being passed there12:15
*** ducttape_ has quit IRC12:15
*** amotoki has joined #openstack-nova12:16
*** ducttape_ has joined #openstack-nova12:16
*** shaohe_feng has quit IRC12:19
*** shaohe_feng has joined #openstack-nova12:19
mdboothftersin: Yup, I am now12:21
*** wangqun has joined #openstack-nova12:21
johnthetubaguygmann: will take a peak at the updated series right now12:23
johnthetubaguygmann: for stuff that passes the instance as a target, the fix should be only passing the project_id, I believe12:23
gmannjohnthetubaguy: not user_id also?12:24
*** jistr has joined #openstack-nova12:24
ftersinmdbooth: thank you for review of https://review.openstack.org/#/c/33460912:24
johnthetubaguygmann: nope, we are only adding user_id in that very small minimal list of places, it shouldn't be anywhere else12:24
ftersinmdbooth: could you also take a look at other pathes in the series?12:25
johnthetubaguygmann: thats why I didn't want us to use the helper function12:25
gmannjohnthetubaguy: ok12:25
ftersinmdbooth: and probably you should look at the 3rd version of this patch, because it changes some your recent code12:25
mdboothftersin: I took a brief look at the other 2. There was nothing which scared me :)12:26
* mdbooth will look closer.12:26
*** jwcroppe has joined #openstack-nova12:26
ftersinmdbooth: you commented the 2nd version, but in the 3rd version i merged my code with your changes of disk_info in tests12:28
*** nolwenn has joined #openstack-nova12:29
mdboothftersin: Ah, fix for volume-backed?12:29
openstackgerritLudovic Beliveau proposed openstack/nova: Update binding:profile for SR-IOV ports on resize-revert  https://review.openstack.org/34744412:29
*** shaohe_feng has quit IRC12:29
mdboothSorry, I thought you were talking about the third patch in the series12:29
*** ivase has joined #openstack-nova12:29
*** amotoki has quit IRC12:29
*** shaohe_feng has joined #openstack-nova12:30
ftersinmdbooth: the 3rd patch is interesting also. it completely reverts some old code. But at the first patch i fixes this code. And the way i do it raises questions.12:30
*** jistr has quit IRC12:31
ftersinmdbooth: i mean the part with strange equation len(disk_infos) > int(not booted_from_volume)12:31
mdboothftersin: The third patch removes that logic I can't parse :)12:31
mdboothYeah12:31
*** amotoki has joined #openstack-nova12:31
ftersinmdbooth: so that i think if core reviewers approve the removal, i can reorder the patch set to remove that logic at the first12:32
mdboothftersin: That would be better, I think.12:32
*** jistr has joined #openstack-nova12:32
* mdbooth should look in more detail, though12:33
*** gongysh has quit IRC12:33
*** adreznec has quit IRC12:33
ftersinmdbooth: what do you think about chances of that old code to be removed? my concern is that core reviewers do not like to remove something.12:34
*** markvoelker has joined #openstack-nova12:35
ftersinmdbooth: but the main point of the patch set is to fix the bug, i.e. i need the first review to be merged12:35
mdboothI've managed to delete a bunch lately. I think deleting cruft is extremely valuable, personally.12:35
*** adreznec has joined #openstack-nova12:35
*** amotoki has quit IRC12:35
mdboothWell, I *really* don't like that has_local_disks logic. It scares the hell out of me, and this code is scary enough already.12:35
*** wangqun has quit IRC12:36
*** baoli has joined #openstack-nova12:36
mdboothI can well believe that you have a proof of its correctness, but it's still incomprehensible to regular mortals.12:36
sdaguegmann: cool12:36
openstackgerritGhanshyam Mann proposed openstack/nova: Remove user_id policy enforcement from server start  https://review.openstack.org/35466212:37
gmannsdague: johnthetubaguy like this right - https://review.openstack.org/35466212:37
gmannsdague: johnthetubaguy passing user_id from context12:37
*** code-R has joined #openstack-nova12:37
*** baoli_ has joined #openstack-nova12:37
ftersinmdbooth: i could rewrite it with a set of temporary variables, named pretty well to understand the logic12:37
ansiwenis someone taking care of ec2api? the default value of bucket_path is set to $pybasedir, which is pointless, because it's not writeable. Not sure if this is a packaging or a module problem...12:37
johnthetubaguygmann: should just pass project_id, I think12:38
nolwennhello ! I made a patch to clean up the second ephemeral disk. Would it be possible that people make a review to have some returns?12:38
nolwennhttps://review.openstack.org/#/c/341008/12:38
andrey-mpansiwen: i care - which version do you use?12:38
johnthetubaguygmann: because start is not on that list: http://specs.openstack.org/openstack/nova-specs/specs/newton/approved/user_id_based_policy_enforcement.html12:38
gmannjohnthetubaguy: yea, but user_id is from context not from instance12:39
ansiwenandrey-mp: awesome... :-)12:39
johnthetubaguygmann: but why?12:39
gmannjohnthetubaguy: i think we need to pass both for oslo policy not?12:39
*** shaohe_feng has quit IRC12:39
johnthetubaguygmann: really?12:39
gmannjohnthetubaguy: am not sure, i remember the bug you pointed out earlier but not sure12:40
ansiwenandrey-mp: I deploy with puppet-ec2api, how can I find out which version it is?12:40
*** shaohe_feng has joined #openstack-nova12:40
johnthetubaguygmann: what was project_id, because of the default policy12:40
ftersinmdbooth: but i'd prefer to remove it at the first. The author of the code said me that it was a mistake to do that. Moreover now he proposes to revert it: https://review.openstack.org/#/c/337056/12:40
andrey-mpansiwen: which oepnstack version I mean12:40
ansiwenandrey-mp: master12:40
*** baoli has quit IRC12:41
ansiwenandrey-mp: https://review.openstack.org/#/c/352552/12:41
*** jistr has quit IRC12:41
ansiwenandrey-mp: this is how the resulting config looks like: http://logs.openstack.org/52/352552/3/check/gate-puppet-ec2api-puppet-beaker-rspec-centos-7/347faee/logs/etc/ec2api/ec2api.conf.txt.gz12:41
ftersinmdbooth: what do you think: are there enough reasons for core reviewers to approve the removal?12:41
openstackgerritTomi Juvonen proposed openstack/nova-specs: Maintenance reason to server  https://review.openstack.org/31051012:42
*** code-R has quit IRC12:42
*** ducttape_ has quit IRC12:42
ansiwenandrey-mp: it fails, because it tries to create /usr/lib/python2.7/site-packages/buckets, which fails of course12:42
mdboothftersin: I hadn't appreciated the importance of the third patch and had only looked at it briefly.12:42
ftersinmdbooth: and the last question - what core should i ping to approve this?12:42
mdboothftersin: I'm going to spend some time right now looking at this thoroughly, then I'll get back to you.12:43
*** pbandark has quit IRC12:43
ftersinmdbooth: thanks12:43
andrey-mpansiwen: s3 implementation in ec2api stores bucket to some data directory - I think that you should configure this directory (let me see key name)12:44
*** rkrum has joined #openstack-nova12:45
ansiwenandrey-mp: but the default value should be something reasonable I guess... not pybasedir I assume?12:45
cdentjaypipes: do you have an opinion on what the python interface should be for setting/adding/removing aggregates from a resource provider? There was code in the past (since removed) that used an aggregates field on the resource provider but given how we are managing inventories we may prefer to just use methods. Your opinion?12:45
andrey-mpansiwen: this was copyied from nova - and we didn't change this behaviour12:46
*** rlrossit has joined #openstack-nova12:46
andrey-mpthis string in the config file - '#buckets_path=$state_path/buckets'12:47
andrey-mpansiwen: ^12:47
*** rfolco has joined #openstack-nova12:47
andrey-mpansiwen: so all projects have to configure state_path in config file - in this case bucket will be placed there12:47
*** amotoki has joined #openstack-nova12:47
ansiwenandrey-mp: so it's wrong to create a config file with no state_path defined? you can't rely on the default value?12:48
ansiwenandrey-mp: and this line is commented out, will a value state_path still be picked up?12:48
ansiwenandrey-mp: I have to fix this problem, and I'm just trying to figure our the correct approach to do this12:49
*** shaohe_feng has quit IRC12:49
ansiwenandrey-mp: so any suggestions are highly appreciated :-)12:50
andrey-mpansiwen: yeah, it's wrong. nova has same default value for state_path12:50
*** shaohe_feng has joined #openstack-nova12:50
openstackgerritJohn Garbutt proposed openstack/nova: Remove user_id policy enforcement from server start  https://review.openstack.org/35466212:50
andrey-mpansiwen: check puppets for nova - how they configure state_path value.12:50
ansiwenandrey-mp: I see, will do that, thanks a bunch!12:50
*** rkrum has quit IRC12:50
*** edleafe is now known as figleaf12:51
openstackgerritJohn Garbutt proposed openstack/nova: Allow authorization by user_id for server changePassword action  https://review.openstack.org/35453712:52
ftersinansiwen: hi. am i right that you setup s3 server from ec2-api by puppets? since this server is not intended to be in production, may i ask you - why do you set up it?12:52
openstackgerritjichenjc proposed openstack/nova: WIP: not use cert and private key in glance client  https://review.openstack.org/35470612:52
*** rlrossit has quit IRC12:53
*** links has quit IRC12:53
ansiwenftersin: I have no idea :-) let me explain my situation:12:53
openstackgerritjichenjc proposed openstack/nova: refactory query user_id code for quota-sets  https://review.openstack.org/28016512:54
gmannjohnthetubaguy: if user override policy with "user_id:%(user_id)s" and  if we do not pass user_id=context.user_id in target then, policy will fail on start server12:54
johnthetubaguygmann: thats exactly our intent12:55
mdboothftersin: Are you planning to backport this, btw?12:55
johnthetubaguygmann: we don't want anyone using that policy, it breaks the API semantics12:55
gmannjohnthetubaguy: but in all other place where user_id and project_id a re default of context it pass even operator override polivy with "user_id:%(user_id)s12:55
* mdbooth assumes yes12:55
johnthetubaguygmann: true, but that will "work" by accident12:56
ansiwenftersin: I have to integrate puppet-ec2api into tripleO, and at the moment puppet-ec2api doesn't work, because it tries to create a bucket directory in a non-writeable directory. so you say, puppet-ec2api shouldn't deplay the s3 server in the first place?12:56
*** rlrossit has joined #openstack-nova12:56
johnthetubaguygmann: I want that code to have the default target of {} ASAP12:56
ansiwenftersin: see here: https://review.openstack.org/#/c/352552/12:56
andrey-mpansiwen: production deployment must use swift/swift3 for S3 implementation12:56
openstackgerritjichenjc proposed openstack/nova: Verify os-aggregates.inc on sample files  https://review.openstack.org/33311412:57
gmannjohnthetubaguy: so we always want to say wrong policy if they do "user_id:%(user_id)s for non user_id honors API action12:57
ftersinmdbooth: frankly i have no thought about it. I need this fix for ScaleIO ephemeral driver, to contribute it to upstream. I have an alternative solution for previous releases. And this bug is fixed in the solution. So that i do not need to backport it.12:57
andrey-mpansiwen: s3 implementation from ec2api is only for testing - this implementation doesn't have authorization at all for example12:58
ftersinmdbooth: but perhaps someone else wants it once12:58
ansiwenandrey-mp, ftersin: so you would say that this line is wrong: https://review.openstack.org/gitweb?p=openstack/puppet-ec2api.git;a=blob;f=spec/acceptance/basic_ec2api_spec.rb;h=147349408d5e5a659d86255c67334967bf974785;hb=26c34eae13aa66a5c295b8ec848d0f0834b0c9e9#l3812:59
mdboothftersin: I was just considering the edict for 'no new mox', which https://review.openstack.org/#/c/335542/ technically falls foul of.12:59
gmannjohnthetubaguy: i got it "user_id:%(user_id)s is always for user_id enforcement which we do not want to pass for default cases.12:59
mdboothHowever, if you planned to backport it, it wouldn't make sense to add a bunch of almost entirely irrelevant additional test churn.12:59
gmannjohnthetubaguy: and yea default target {} will make all API consistent on that12:59
ansiwenandrey-mp, ftersin: so ::ec2api::s3 shouldn't be deployed in testing, right?13:00
*** shaohe_feng has quit IRC13:00
*** zul has quit IRC13:00
ansiwenandrey-mp, ftersin: could you give me a hint, which components of puppet-ec2api are "supported" and can be tested against?13:00
*** lucas-hungry is now known as lucasagomes13:00
*** shaohe_feng has joined #openstack-nova13:00
andrey-mpansiwen: ec2api::s3 - it can be deployed in testing. and it shouldn't be deployed in production13:01
andrey-mpansiwen: but all components are supported and should be tested )13:01
ftersinansiwen: you can see our thoughts about s3 server at the end of install section of ec2-api's readme (https://github.com/openstack/ec2-api#installation)13:01
*** tonytan4ever has joined #openstack-nova13:01
*** baoli_ has quit IRC13:02
ansiwenandrey-mp: ok, so we should still test it in puppet-ec2api and fix the state_path thing, ok...13:02
andrey-mpansiwen: yes, it's true13:02
*** jistr has joined #openstack-nova13:03
ftersinmdbooth: since that patch reverts some changes, it restores old code...13:04
*** daemontool_ has quit IRC13:05
*** prateek has joined #openstack-nova13:05
*** tonytan4ever has quit IRC13:06
*** zul has joined #openstack-nova13:07
ftersinmdbooth: it is not so hard to change "new" mox to mock. but in this case the change will not be a pure reverting. do not know if it is important13:08
mdboothftersin: I'd hope not. I'd certainly flag this as a pragmatic exception.13:08
jaypipescdent: I'd prefer to just use methods, yes.13:10
*** shaohe_feng has quit IRC13:10
jaypipesand good morning cdent. :)13:10
*** jaypipes is now known as leakypipes13:10
*** kevinz has quit IRC13:10
cdentjaypipes: cool and good afternoon13:10
cdentthanks13:10
*** jwcroppe has quit IRC13:11
cdentthat makes things easier13:11
*** jistr has quit IRC13:13
*** shaohe_feng has joined #openstack-nova13:13
*** manous has quit IRC13:13
*** kylek3h has joined #openstack-nova13:13
*** fragatina has joined #openstack-nova13:13
*** sdake has joined #openstack-nova13:14
*** edmondsw has joined #openstack-nova13:14
mdboothftersin: +1 from me on the revert, and I'd support re-ordering the patches.13:16
*** pwnall has joined #openstack-nova13:16
ftersinansiwen: i hope this ec2api::s3 puppet does not intended for production usage as well? It should be used for some kind of test deployments only. Am i right? Who can i ask more about this puppet purposes?13:16
ftersinansiwen: does -> is13:17
*** mdrabe has joined #openstack-nova13:17
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: live-migration test hook for serial console  https://review.openstack.org/34747113:18
*** kylek3h has quit IRC13:18
*** fragatina has quit IRC13:18
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server delete  https://review.openstack.org/35455513:18
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server update  https://review.openstack.org/35461013:19
openstackgerritGhanshyam Mann proposed openstack/nova: Allow authorization by user_id for server evacuate  https://review.openstack.org/35465613:19
openstackgerritGhanshyam Mann proposed openstack/nova: Remove user_id policy enforcement from server start  https://review.openstack.org/35466213:19
ansiwenftersin: I have nothing to do with puppet-ec2api so far, I just have the task to integrate it into tripleO, so operators can deploy a ec2api. It sounds like you are suggesting we should not expose the s3 service to an installer, right?13:19
ftersinmdbooth: ok, thanks. before to reorder the series i'd like to approve the revert with a +2er. who of them can help me with that?13:19
openstackgerritGhanshyam Mann proposed openstack/nova: Fix policy tests for project_id enforcement  https://review.openstack.org/35469013:20
gmannjohnthetubaguy: rebased the series13:20
gmannjohnthetubaguy: and for pass tests in https://review.openstack.org/#/c/354690/13:20
*** shaohe_feng has quit IRC13:20
gmannjohnthetubaguy: existing tests go with admin  or  "project_id:%(project_id)s" and that tests explicitly tests  "project_id:%(project_id)s" only13:21
*** shaohe_feng has joined #openstack-nova13:21
ansiwenftersin: will it ever be usable for production usage?13:21
ftersinansiwen: correct13:21
andrey-mpDoes anybody can approve this review https://review.openstack.org/#/c/352455/ ? It reverts quota's behaviour. due to this bug I can't install devstack with fake virt driver.13:21
ftersinansiwen: never13:21
ansiwenftersin: I see13:22
*** tonytan4ever has joined #openstack-nova13:22
ansiwenftersin: ok, then it's really the question why a puppet module should test it. it can be tested in ec2api itself...13:22
ftersinansiwen: it has a lot of omissions: authentification, acl, multinode ability13:22
*** pwnall has quit IRC13:23
mdboothftersin: The advantage of just doing it first is that then the revert becomes a dependency of a bugfix13:24
mdboothAnd bugfixes are generally higher priority13:24
andrey-mpansiwen: as I understand - puppet module tests own resources, right?13:24
mdboothI don't even think that's gaming anything, because it really does improve the bugfix13:24
ansiwenandrey-mp: it does functional tests as well...13:25
ftersinansiwen: it is tested in ec2-api, it is used in CI to verify some cases, using lightweight (w/o Swift) deployment13:25
andrey-mpansiwen: I think that it's better to remove s3 manifest at all from this puppet13:25
sdaguehmmm no dansmith yet?13:26
ftersinmdbooth: hmm, sounds good13:26
*** ababich1 has quit IRC13:26
ansiwenandrey-mp: ok... I will propose it13:26
*** jwcroppe has joined #openstack-nova13:27
*** jamesdenton has joined #openstack-nova13:29
*** rama_y has joined #openstack-nova13:29
*** shaohe_feng has quit IRC13:30
*** shaohe_feng has joined #openstack-nova13:31
*** code-R has joined #openstack-nova13:31
*** chandankumar has quit IRC13:31
*** code-R has quit IRC13:36
*** s1061123 has joined #openstack-nova13:37
*** s1061123_ has quit IRC13:37
openstackgerritGhanshyam Mann proposed openstack/nova: Remove user_id policy enforcement from instance_actions  https://review.openstack.org/35472613:37
*** pwnall has joined #openstack-nova13:38
*** rodrigods has quit IRC13:39
*** rodrigods has joined #openstack-nova13:40
*** mriedem has joined #openstack-nova13:40
*** shaohe_feng has quit IRC13:41
*** shaohe_feng has joined #openstack-nova13:41
mriedemsdague: did you look over the changes to https://review.openstack.org/#/c/354349/ ?13:42
*** ducttape_ has joined #openstack-nova13:43
sdaguemriedem: yeh, I was trying to work through the mitaka test issue13:43
sdaguebecause we unconditionally test microversions, even if the cloud we are hitting doesn't support them13:43
sdaguewhich should be easy enough to change to skips13:44
sdaguebut in doing so, I discovered that a default stable/mitaka devstack has no flavors13:44
openstackgerritDan Smith proposed openstack/nova: Make simple_cell_setup idempotent  https://review.openstack.org/35473413:44
mriedemsdague: i created a stable/mitaka devstack the other day and it had flavors13:45
sdaguethe tempest ones13:45
sdaguewe took out the flavor part of the db migrations13:46
sdaguebut forgot to backport this - https://review.openstack.org/#/c/354710/13:46
*** sdake has quit IRC13:46
mriedemsdague: i have them all13:46
sdaguemriedem: ... hmmm13:46
mriedemi'm at dcb218d63e044a7b88ce2b5b0eae6b6cd166396113:46
sdagueoh, gdi13:47
mriedemdid you mean the special sdague custom ones?13:47
*** ducttape_ has quit IRC13:48
sdagueI forgot how flakey the branch changing in devstack is13:48
mriedemok, i just git cloned devstack and git checkout stable/mitaka13:48
mriedemeasy peasy13:48
sdagueyeh, well, this box had previously done devstack master13:48
openstackgerritGhanshyam Mann proposed openstack/nova: Remove user_id policy enforcement from unlock_override  https://review.openstack.org/35473713:49
lasciiGuest13957: johnthetubaguy left a comment on the original review suggesting something like https://review.openstack.org/#/c/354734. I never really expected the command to be run more than once though.13:49
*** Guest13957 is now known as dansmith13:50
*** dansmith is now known as Guest3789013:50
Guest37890lascii: well, I'm actually not sure it's going to be specifically needed here, but I expect puppet people would appreciate just being able to run it13:50
Guest37890gdi13:50
*** nkrinner is now known as nkrinner_afk13:50
*** Guest37890 is now known as dansmith13:51
*** shaohe_feng has quit IRC13:51
johnthetubaguyseems easier to make it idempotent, stops nasty surprizes13:51
dansmithyeah13:51
*** shaohe_feng has joined #openstack-nova13:51
lasciiyeah, idempotence is good. But I don't want folks to feel the need to run it multiple times. it's intended to be a fire and forget thing13:51
*** Dinesh_Bhor has quit IRC13:52
dansmithwell, I print a thing13:52
*** ecelik has quit IRC13:52
openstackgerritAndrew Laski proposed openstack/nova: Add ObjectListBase concat methods  https://review.openstack.org/35258713:52
openstackgerritAndrew Laski proposed openstack/nova: Fix migration list + MigrationList operation  https://review.openstack.org/35369413:52
*** thorst_ has joined #openstack-nova13:53
lasciimy concern really stems from wanting deployers to use something like https://review.openstack.org/#/c/328491 for cell maintenance and not abuse the setup command13:53
lasciiI'm not complaining though. not failing in the setup command is good13:54
dansmithbut this command is before that command, right?13:54
dansmithyou have to do the initial cell setup once and then you can add hosts over and over13:55
dansmithlet me see if it's actually required here and then we can discuss the meta question if you want13:55
lasciisure13:55
dansmithat a minimum it should not explode with a stack trace like it does today :)13:56
sdaguemriedem: did you see my email about os-brick locks?13:56
lascii+113:56
*** vishwanathj has joined #openstack-nova13:56
*** eharney has joined #openstack-nova13:56
*** sdake has joined #openstack-nova13:56
mriedemsdague: already replied, twice13:57
mriedemsdague: did you see my email about get me a network schema changes?13:57
*** thorst_ has quit IRC13:57
*** dansmith is now known as superdan13:57
lasciidansmith: I'm just wondering how we incentivize folks to use the add_host command when the setup command will also do what they want but is The Wrong Thing(tm). anyways, just overthinking it for the moment13:58
*** TravT_ has joined #openstack-nova13:59
superdanlascii: the simple command only does anything the first time you run it13:59
sdaguemriedem: I saw that it existed, I didn't digest it13:59
superdanwith my mchange13:59
*** diogogmt has joined #openstack-nova13:59
superdanlascii: don't they still have to run that once before they use add_host?14:00
sdagueI was trying to get these novaclient changes passing stable (and completely screwing up my stable setup in the process apparently)14:00
*** anush_ has joined #openstack-nova14:00
*** kylek3h has joined #openstack-nova14:01
*** shaohe_feng has quit IRC14:01
lasciisuperdan: yeah, the process should be run the simple command once and then run add_host anytime a host or hosts are added14:01
mriedemsdague: i'm a little less concerned about getting the current state of novaclient passing on stable/mitaka, but more worried about 2.3614:01
mriedemsince that's the breaking change14:02
*** shaohe_feng has joined #openstack-nova14:02
mriedembut yeah it looks like it's really just the functional tests are not capping the api version based on what's available from the server14:02
lasciisuperdan: but the map_cell_and_hosts within the simple command will also end up adding hosts I believe. but I think I'm overly worrying about it14:02
*** TravT has quit IRC14:02
cdentsdague: if you get a chance today to restamp the bottom three starting at https://review.openstack.org/#/c/352573/ that would be great. Once those three are in I can parallelize those that follow and avoid some of the "depends on merge that failed" problem that was happening last night ought to be less of an issue14:02
*** rhagarty_ has joined #openstack-nova14:03
sdaguemriedem: ok, which bits about 2.36?14:03
sdaguemriedem: or more importantly, what are the outstanding bits you think we're missing14:04
*** rhagarty has quit IRC14:04
openstackgerritJay Pipes proposed openstack/nova-specs: Resource providers: dynamic resource classes  https://review.openstack.org/31269614:04
*** pradk has quit IRC14:05
mriedemsdague: we still have to do the network name translation part14:05
mriedemi'm in meetings for like the next 1.5 hours14:05
*** baoli has joined #openstack-nova14:05
sdagueok, I'll get back a master stack on a different NUC and run that in parallel14:05
*** jaugustine_ has joined #openstack-nova14:05
ansiwenandrey-mp: https://review.openstack.org/#/c/354732/14:06
sdaguemriedem: you have a couple of commands that fail that should pass?14:06
sdagueI knew the image ones off the top of my head14:06
mriedemsdague: nova boot --nic net-name=public14:07
*** bnemec is now known as beekneemech14:07
*** pradk has joined #openstack-nova14:07
*** rhagarty_ has quit IRC14:08
*** burgerk has joined #openstack-nova14:09
sdaguemriedem: ok, cool, I'll poke on that14:10
mriedemsdague: i think we're ok with the image patch...i feel weird about it given the amount of unit testing that i had to change,14:10
mriedembut the only thing that really needed to change in the actual shell code was the image delete stuff14:11
mriedemi could add functional tests for those if we have reservations, or i'll just pull it down and test that stuff out locally a bit14:11
*** shaohe_feng has quit IRC14:11
andrey-mpansiwen: good )14:12
sdagueis there a volume name issue out there as well?14:12
*** jwcroppe has quit IRC14:12
*** jaugustine_ has quit IRC14:12
*** thorst_ has joined #openstack-nova14:14
*** shaohe_feng has joined #openstack-nova14:14
*** _coolsvap_ is now known as coolsvap14:16
*** zenoway has joined #openstack-nova14:19
mriedemsdague: pretty sure there isn't14:21
mriedemyou have to pass volume id to the cli14:21
mriedemi know that because we completely removed the cs.volumes/snapshots/types api code14:21
mriedemso the shell wouldn't have anything to call for that translation14:21
sdaguemriedem: ok, good14:21
*** shaohe_feng has quit IRC14:22
*** shaohe_feng has joined #openstack-nova14:23
*** raunak has joined #openstack-nova14:26
sdaguemriedem: you want to kick this in as well - https://review.openstack.org/#/c/354638/ - just makes debugging these novaclient tests easier14:31
*** ducttape_ has joined #openstack-nova14:32
*** shaohe_feng has quit IRC14:32
*** hemanthm|afk is now known as hemanthm14:33
*** mlavalle has joined #openstack-nova14:33
*** whenry has joined #openstack-nova14:33
markus_zjohnthetubaguy: Maybe you could have a look at https://review.openstack.org/#/c/352455/214:33
*** vladikr has joined #openstack-nova14:33
mriedemsdague: done14:34
mriedemsdague: in that vein, i'm pretty proud of this change i made last night https://review.openstack.org/#/c/354349/4/novaclient/tests/unit/fakes.py14:34
mriedemafter banging my head against the wall14:34
mriedemfor several hours14:34
*** devkulkarni1 has joined #openstack-nova14:34
*** code-R has joined #openstack-nova14:34
mriedemi could pull that out as a separate change, but whatever14:34
*** shaohe_feng has joined #openstack-nova14:35
*** baoli has quit IRC14:36
*** baoli has joined #openstack-nova14:36
*** tanvirt has joined #openstack-nova14:36
*** jistr has joined #openstack-nova14:39
*** zenoway has quit IRC14:40
*** zenoway has joined #openstack-nova14:40
*** rhagarty has joined #openstack-nova14:41
*** devkulkarni1 has quit IRC14:41
openstackgerritMarkus Zoeller (markus_z) proposed openstack/nova: live-migration test hook for serial console  https://review.openstack.org/34747114:41
*** shaohe_feng has quit IRC14:42
*** raunak has quit IRC14:43
*** shaohe_feng has joined #openstack-nova14:43
*** zenoway has quit IRC14:45
*** woodster_ has joined #openstack-nova14:45
*** zz_dimtruck is now known as dimtruck14:46
EmilienMhi, I'm not sure if it's a bug but I noticed since recently, nova-scheduler and nova-consoleauth (maybe more services) are failing to start if AMQP is not ready (see logs). I thought it would retry until it works14:47
EmilienMhttp://logs.openstack.org/55/354355/4/check/gate-puppet-openstack-integration-4-scenario003-tempest-ubuntu-xenial-nv/dda4cc6/logs/nova/nova-consoleauth.txt.gz#_2016-08-12_12_11_08_71214:47
EmilienMit sounds like it try during 16 seconds and then the service stops14:47
EmilienMwe are running trunk14:48
*** devkulkarni1 has joined #openstack-nova14:48
sdagueEmilienM: could it be an oslo.message change?14:51
sdaguebecause at this point I think all that logic is outside of nova14:51
superdanlascii: sorry, call.. so my change to the simple command makes it completely not do anything if you call it a second time.. it just fails and bails.. it will not do the second step14:51
EmilienMsdague: I checked recent commits in oslo.messaging and oslo.service and I haven't noticed anything about that14:52
*** mdrabe has quit IRC14:52
*** shaohe_feng has quit IRC14:52
*** devkulkarni1 has quit IRC14:52
*** rlrossit_ has joined #openstack-nova14:52
lasciisuperdan: it's https://review.openstack.org/#/c/354734/1/nova/cmd/manage.py@1193 that does somewhat the same thing as add_hosts14:53
*** shaohe_feng has joined #openstack-nova14:53
superdanlascii: okay, can I put that after the cell0 mapping to make it not happen?14:54
mriedemno recent kombu or amqp changes in upper-constraints14:54
*** devkulkarni has joined #openstack-nova14:54
superdanlascii: whatever, I think we need to make this idempotent one way or the other.. my day is stacking up fast, so let me get the grenade thing working, and we can discuss what to do about this later, okay?14:54
EmilienMof course we could ensure that rabbit is ready before nova tries to start in puppet14:55
EmilienMbut I figured it would be useful to share this info with you14:55
lasciisuperdan: yeah. no argument on the idempotence. switching the two should work. but worrying about it later is fine with me14:55
*** zenoway has joined #openstack-nova14:55
superdanlascii: we can worry about it before merge, I mean14:56
*** rlrossit has quit IRC14:56
*** rrecio has joined #openstack-nova14:56
lasciisuperdan: yeah, once it's proven to be working we can discuss it14:57
*** rrecio_ has joined #openstack-nova14:58
*** mdrabe has joined #openstack-nova14:59
*** _ducttape_ has joined #openstack-nova14:59
*** ijw has joined #openstack-nova15:00
sdaguemriedem: hmmm... os-networks is returning a 200 at 2.35?15:00
*** jdurgin1 has joined #openstack-nova15:00
*** EinstCrazy has joined #openstack-nova15:00
mriedemsdague: do you have a nova-network devstack or neutron?15:01
sdagueneutron15:01
*** rrecio has quit IRC15:01
openstackgerritSarafraj Singh proposed openstack/nova: Improve help text for glance options  https://review.openstack.org/32400715:01
mriedemoh yeah the proxy will work either way for that one15:01
mriedemsdague: 2.36 is the breaking api15:01
mriedem2.35 is fine15:01
sdagueoh, we've forced the fallback15:02
johnthetubaguymarkus_z: dang it, I thought I -1ed a few patches for that, but I see it was only some of the options in that file that are correct15:02
openstackgerritjichenjc proposed openstack/nova: not use cert and private key in glance client  https://review.openstack.org/35470615:02
*** shaohe_feng has quit IRC15:03
*** ducttape_ has quit IRC15:03
*** shaohe_feng has joined #openstack-nova15:03
*** abhishekk has quit IRC15:04
mriedemsdague: not yet15:04
mriedemsdague: are you on vanila novaclient?15:05
mriedemhttps://review.openstack.org/#/c/347514/ does the fallback to 2.3515:05
sdagueno, this is with the image patch15:05
mriedemok, that shouldn't fallback to 2.3515:05
mriedemhttps://review.openstack.org/#/c/347514/ adds the 2.36 support, so you shouldn't even get to 2.36 yet15:05
mriedemhttps://review.openstack.org/#/c/347514/6/novaclient/__init__.py15:05
openstackgerritJohn Garbutt proposed openstack/nova: Allow Nova Quotas to be Disabled  https://review.openstack.org/35245515:06
*** sheel has joined #openstack-nova15:06
sdaguemriedem: ok, though apparently I'll need to do something else to my config to make this work15:06
sdagueExternalNetworkAttachForbidden: It is not allowed to create an interface on external network e9bdb773-08de-4454-9505-393a5ee1c1a815:06
mriedemsdague: which nic are you trying to attach to?15:07
markus_zjohnthetubaguy: thanks for https://github.com/openstack/openstack-doc-tools/commit/4f0433f btw, looks really nice.15:08
mriedemtry: nova boot --flavor 42 --nic net-name=private --image ? --poll test15:08
mriedemi'd think you could attach to the 'public' network though, not sure what's up there15:08
*** armax has joined #openstack-nova15:08
johnthetubaguymarkus_z: no worries, I hate the patch, but it does what we needed :)15:08
mriedemi have my own local.conf for neutron, i haven't tried the defaults yet15:08
sdagueok, private seems to be getting further15:08
johnthetubaguymarkus_z: we have a few bugs there still, because of how we use "*" and because we sometimes indent I think15:08
markus_zjohnthetubaguy: yeah, I had that on my todo list for ages :(15:09
sdaguemriedem: pastebin your config?15:09
johnthetubaguymarkus_z: going to go through the config reference with a fine tooth comb once we get back feature freeze15:09
mriedemsdague: http://paste.openstack.org/show/556116/15:09
johnthetubaguymarkus_z: got some of the OSIC docs folks on hand to help review and edit those changes15:09
markus_zjohnthetubaguy: ah, explains why some get rendered differently15:09
sdaguemriedem: also... this is what the defaults currently look like - https://review.openstack.org/#/c/354658/ - was trying to push a new devstack doc update daily15:09
mriedemfull disclosure, it's rtheis' config15:09
johnthetubaguymarkus_z: the bug is ln465 I think, or its a bug with how we currently have help text, depending on how you look at it15:10
*** ayoung has joined #openstack-nova15:10
mriedemi don't use 75% of that local.conf really15:11
johnthetubaguymarkus_z: I am thinking of writing a pep8 check to catch all that formatting stuff, we actually need to wrap the lines early, else the sample file goes silly as well15:11
markus_zjohnthetubaguy: I'm trying to run tox for the manuals project but it doesn't use the code I have locally15:11
sdaguemriedem: other than the service changes, that's all what I've got15:11
markus_zjohnthetubaguy: Did you do something to test that locally?15:11
*** xyang1 has quit IRC15:11
johnthetubaguymarkus_z: yeah, its nasty actually, there is a README somewhere, let me find it...15:11
sdaguemost of that is commented out15:12
johnthetubaguymarkus_z: https://github.com/openstack/openstack-doc-tools/blob/master/autogenerate_config_docs/README.rst15:12
markus_zjohnthetubaguy: I think I remember a python package which checks plain text for rst format15:12
*** xyang1 has joined #openstack-nova15:12
johnthetubaguymarkus_z: honestly, I think we should only use the subset that works with the tooling, just "*" and "``" I think15:12
openstackgerritjichenjc proposed openstack/nova: Add log for instance with host field set  https://review.openstack.org/35478115:12
mriedemsdague: and?15:13
mriedemstack@newton:~$ nova network-list15:13
mriedem+--------------------------------------+---------+------+15:13
mriedem| ID                                   | Label   | Cidr |15:13
mriedem+--------------------------------------+---------+------+15:13
mriedem| 3d724099-b8f6-464f-ad43-bd092e8e8320 | public  | -    |15:13
mriedem| fcdad99f-3b6d-4376-bbd9-e8f9190390c4 | private | -    |15:13
mriedem+--------------------------------------+---------+------+15:13
markus_zjohnthetubaguy: yeah, fair enough, that's all we should need IMO15:13
mriedemsame networks?15:13
*** shaohe_feng has quit IRC15:13
sdagueyeh15:13
*** rmart04 has quit IRC15:13
mriedemsdague: oh, router:external           | True15:13
mriedemfor public15:13
mriedemyou can't use that15:13
mriedemso you have to use the private nic15:13
mriedemthat's why you're hitting ExternalNetworkAttachForbidden15:14
sdagueok, well, at least it's broken in a known way15:14
mriedemthat's working as designed bro!15:14
sdaguethough, it's a little sad that we don't detect that until n-cpu15:14
mriedemyeah, we could probably move that to validate_networks15:14
johnthetubaguymarkus_z: cool, I think we can get away with just that15:14
mriedemb/c we have the same code to find the available network15:14
sdague"yep totally works!" "ERROR"15:14
mriedemwe just aren't checking the external thing until n-cpu15:15
*** eharney has quit IRC15:15
*** eharney_ has joined #openstack-nova15:15
*** jaosorior has quit IRC15:15
mriedemi think i was talking about this with someone awhile back too, b/c there is a policy check in the neutron code in nova around this too15:15
mriedemwhich is really weird15:15
sdagueok, well this will be my afternoon, but I'll end up taking a break soon to go hit lunch with mtreinish15:16
*** nolwenn has quit IRC15:16
*** shaohe_feng has joined #openstack-nova15:16
openstackgerritAndrey Shestakov proposed openstack/nova: Add support for detaching interfaces in ironic  https://review.openstack.org/34741015:16
mriedemok, i'm going to work on getting a nv functional novaclient + neutron job going15:16
mriedemposted a thread to the ML about how to configure that since i have someone here that can work on it i think15:16
*** eharney_ is now known as eharney15:16
mriedembut not sure about the best way to do the config/check for neutron15:16
sdaguenetwork in the service catalog15:17
sdaguejust run with that15:17
mriedemok, that's my preferred approach15:17
mriedemcool15:17
*** code-R has quit IRC15:18
markus_zjohnthetubaguy: Is "$sentinel$" some rst magic?15:18
*** rhagarty_ has joined #openstack-nova15:18
johnthetubaguymarkus_z: i substitute that later on with real RST15:18
markus_zjohnthetubaguy: oh, ok, simpler than I thought :)15:19
*** andrey-mp has left #openstack-nova15:19
johnthetubaguymarkus_z: :) the nasty bit is the number of spaces in that substituted string has to match the indent in the template15:20
*** rhagarty has quit IRC15:20
*** yamahata has joined #openstack-nova15:20
*** devkulkarni has quit IRC15:20
prateekleakypipes, regarding your comment on the patch https://review.openstack.org/#/c/327706/, i can see that glance does and cinder also handle conn failures the same way, its just that neutron was missing a wrapper in client15:20
prateekleakypipes, can you please help me understand if that is wrong so that i can further work on it in line with your comment ?15:21
*** devkulkarni has joined #openstack-nova15:23
*** thorst_ has quit IRC15:23
*** shaohe_feng has quit IRC15:23
*** pcaruana has quit IRC15:23
*** EinstCrazy has quit IRC15:23
*** jistr is now known as jistr|call15:23
*** thorst_ has joined #openstack-nova15:24
*** shaohe_feng has joined #openstack-nova15:24
*** eharney_ has joined #openstack-nova15:27
*** TravT_ has quit IRC15:27
*** thorst_ has quit IRC15:28
*** eharney has quit IRC15:29
mriedemsdague: for your post-lunch reading, this is the thread on that external network policy check thing http://lists.openstack.org/pipermail/openstack-dev/2016-June/096919.html15:29
*** danpb has quit IRC15:29
*** beagles is now known as beagles_brb15:29
*** sridharg has quit IRC15:30
*** sulmer has joined #openstack-nova15:30
*** haplo37_ has joined #openstack-nova15:30
openstackgerritjichenjc proposed openstack/nova: prevent non-admin user search soft-delete server  https://review.openstack.org/25847215:31
*** dikonoor has quit IRC15:31
*** thorst_ has joined #openstack-nova15:31
*** markus_z has quit IRC15:32
*** TravT has joined #openstack-nova15:33
*** ayoung has quit IRC15:33
*** shaohe_feng has quit IRC15:33
*** shaohe_feng has joined #openstack-nova15:34
*** thorst__ has joined #openstack-nova15:35
openstackgerritjichenjc proposed openstack/nova: Add comment about how status field changed  https://review.openstack.org/35411915:35
*** thorst_ has quit IRC15:36
*** Cassowary has quit IRC15:37
*** Cassowary has joined #openstack-nova15:37
*** Vek has quit IRC15:38
*** sdake_ has joined #openstack-nova15:39
*** pbandark has joined #openstack-nova15:39
*** thorst__ has quit IRC15:40
johnthetubaguydoes anyone remember why tunneled live-migration was on by default, was that a security thing?15:40
johnthetubaguyhttps://review.openstack.org/#/c/35048015:41
*** jichen has quit IRC15:41
*** sdake has quit IRC15:41
*** jichen has joined #openstack-nova15:42
*** jistr|call is now known as jistr15:42
*** shaohe_feng has quit IRC15:44
sdagueheh, neat, neutron is using file extensions in addition to mime types - http://10.42.0.53:9696/v2.0/networks.json?fields=id&name=private15:45
sdagueok, heading out to get food, but the switcheroo shouldn't be too bad here15:46
*** jichen has quit IRC15:46
*** jwcroppe has joined #openstack-nova15:47
*** shaohe_feng has joined #openstack-nova15:47
*** rcernin has quit IRC15:47
*** ivase has quit IRC15:48
*** mdrabe has quit IRC15:49
*** mdrabe has joined #openstack-nova15:50
*** mdrabe has quit IRC15:52
*** mdrabe has joined #openstack-nova15:52
*** thorst_ has joined #openstack-nova15:53
*** _ducttape_ has quit IRC15:53
*** shaohe_feng has quit IRC15:54
*** ducttape_ has joined #openstack-nova15:54
*** vladikr has quit IRC15:55
*** vladikr has joined #openstack-nova15:55
*** thorst_ has quit IRC15:55
*** thorst_ has joined #openstack-nova15:56
*** shaohe_feng has joined #openstack-nova15:57
*** beagles_brb is now known as beagles15:57
superdanlascii: http://logs.openstack.org/65/354265/6/check/gate-grenade-dsvm-multinode/09c6cce/logs/grenade.sh.txt.gz#_2016-08-12_15_42_05_22415:57
*** krtaylor has quit IRC15:59
*** Apoorva has joined #openstack-nova15:59
*** sdake_ has quit IRC15:59
*** haplo37_ has quit IRC16:00
*** eharney_ is now known as eharney16:00
*** yamahata has quit IRC16:03
*** shaohe_feng has quit IRC16:04
openstackgerritDan Smith proposed openstack/nova: Make simple_cell_setup idempotent  https://review.openstack.org/35473416:05
*** shaohe_feng has joined #openstack-nova16:05
openstackgerritIldiko Vancsa proposed openstack/nova: Replace mox with mock in test_validate_bdm  https://review.openstack.org/35363816:06
openstackgerritIldiko Vancsa proposed openstack/nova: Remove check_attach  https://review.openstack.org/33535816:06
openstackgerritIldiko Vancsa proposed openstack/nova: Replace mox with mock in test_cinder  https://review.openstack.org/35257216:06
*** TRUMP-PENCE has quit IRC16:07
*** raunak has joined #openstack-nova16:09
prometheanfiremriedem: hi :D16:10
*** TRUMP-PENCE has joined #openstack-nova16:10
*** annegentle has joined #openstack-nova16:10
prometheanfiremriedem: here's the reminder16:10
mriedemprometheanfire: heh16:11
mriedemwell, i opened that bug last night16:11
mriedembut i'm doing about 10 other things atm which trump that16:11
prometheanfireya, saw that16:11
*** oanson has quit IRC16:11
prometheanfirethat's fine, it's just a reminder :P16:11
*** pbandark has quit IRC16:12
*** prateek_ has joined #openstack-nova16:13
*** code-R has joined #openstack-nova16:13
*** code-R_ has joined #openstack-nova16:13
*** shaohe_feng has quit IRC16:14
*** prateek has quit IRC16:16
mriedemhttps://bugs.launchpad.net/nova/+bug/1612494 in case someone else wants to hack on this16:16
openstackLaunchpad bug 1612494 in OpenStack Compute (nova) "neutronv2 unit tests fail with python-neutronclient 5.1.0" [High,Confirmed]16:16
*** code-R has quit IRC16:17
*** shaohe_feng has joined #openstack-nova16:17
*** clenimar_ has joined #openstack-nova16:20
*** oanson has joined #openstack-nova16:24
*** parora has joined #openstack-nova16:24
*** shaohe_feng has quit IRC16:25
*** mingyu has joined #openstack-nova16:25
*** mingyu has quit IRC16:25
*** shaohe_feng has joined #openstack-nova16:25
*** clenimar_ has quit IRC16:26
*** yamahata has joined #openstack-nova16:26
*** mingyu has joined #openstack-nova16:27
*** mingyu has quit IRC16:28
*** prateek_ has quit IRC16:28
*** daidv has joined #openstack-nova16:28
*** fragatina has joined #openstack-nova16:30
*** fragatin_ has joined #openstack-nova16:31
*** fragatina has quit IRC16:35
*** shaohe_feng has quit IRC16:35
*** tesseract- has quit IRC16:35
*** shaohe_feng has joined #openstack-nova16:35
*** EinstCrazy has joined #openstack-nova16:36
*** rrecio_ has quit IRC16:36
lasciisuperdan: woot!16:40
superdanlascii: reordered the nova code to fail first and re-used your devstack function in grenade16:41
superdanso if that works I think we're good16:42
*** stewie925 has joined #openstack-nova16:44
*** shaohe_feng has quit IRC16:45
*** shaohe_feng has joined #openstack-nova16:46
openstackgerritAlberto Murillo proposed openstack/nova: Create additional hosts file for dnsmasq before using it.  https://review.openstack.org/21062716:46
stewie925hi guys, am working on ConfigOption cleanup but I notice that http://45.55.105.55:8082/config-options.html  is no longer updated16:48
raj_singhstewie925: Sometimes it takes a little while to reflect changes in that chart16:49
stewie925hi raj - I see that "ephemeral_storage.py" frst patch (removing the needs: tag) was committed in July  (https://review.openstack.org/#/c/348487/1) but its still showing To Do today16:52
*** rlrossit_ has quit IRC16:54
*** rlrossit has joined #openstack-nova16:54
*** sulmer has quit IRC16:55
*** lucasagomes is now known as lucas-dinner16:55
*** unicell has quit IRC16:55
*** shaohe_feng has quit IRC16:55
*** unicell has joined #openstack-nova16:56
*** shaohe_feng has joined #openstack-nova16:56
raj_singhstewie925: must be some issue with linking of that specific file, I will take a look next week.16:58
stewie925thank you raj_singh - no problem16:58
kevinbentonmriedem: you coming to the neutron mid-cycle? :)16:59
*** rlrossit has quit IRC16:59
*** anush_ has quit IRC17:00
*** shaohe_feng has quit IRC17:06
*** shaohe_feng has joined #openstack-nova17:06
*** dikonoor has joined #openstack-nova17:08
*** oomichi has joined #openstack-nova17:08
*** woodster_ has quit IRC17:09
openstackgerritMichael Glaser proposed openstack/nova: Check opt consistency for neutron.py  https://review.openstack.org/34960517:12
*** ayoung has joined #openstack-nova17:13
*** shaohe_feng has quit IRC17:16
*** devkulkarni has quit IRC17:16
*** shaohe_feng has joined #openstack-nova17:16
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Return None in get_instance_id_by_floating_address  https://review.openstack.org/35196017:17
*** baoli has quit IRC17:19
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Remove unnecessary debug logs of normal API ops  https://review.openstack.org/35196217:19
*** tri2sing has joined #openstack-nova17:21
*** coolsvap has quit IRC17:21
cfriesen_Where in the code does CONF.upgrade_levels actually get set?17:23
*** dikonoor has quit IRC17:23
*** krtaylor has joined #openstack-nova17:23
*** prateek_ has joined #openstack-nova17:24
*** code-R_ has quit IRC17:25
*** sheel has quit IRC17:26
*** shaohe_feng has quit IRC17:26
*** parora has quit IRC17:27
*** shaohe_feng has joined #openstack-nova17:27
*** mingyu has joined #openstack-nova17:28
*** jdurgin1 has quit IRC17:29
*** jwcroppe has quit IRC17:29
*** rlrossit has joined #openstack-nova17:29
superdanlascii: grenade-multinode passed and the logs look legit17:31
superdanit will be a while before the rest finish17:31
superdanbut I think we're good17:31
lasciiawesome17:31
superdanwhen sdague comes back I17:32
superdanwill see if he can look at the grenade bits17:32
openstackgerritPatrick East proposed openstack/nova: Add release note to warn about os-brick lock dir  https://review.openstack.org/35450217:32
lasciicool. The nova stuff all looked good to me17:32
*** mingyu has quit IRC17:32
superdanthe nova stuff goes first, so we can move on that at any point when we have test results17:33
*** rlrossit has quit IRC17:34
*** shaohe_feng has quit IRC17:36
*** Vek has joined #openstack-nova17:37
superdanlascii: yes, we'll ask sdague for a grenadeyness assessment, exactly17:38
*** shaohe_feng has joined #openstack-nova17:39
*** rrecio has joined #openstack-nova17:40
*** ijw has quit IRC17:40
lasciiI had to debate that term a bit, grenadic?, grenadonic?, ahh grenadey17:40
superdanlascii: I believe the proper term is "grenadeliness"17:41
*** rrecio_ has joined #openstack-nova17:41
superdan"this code, I must say, possesses excellent grenadeliness"17:41
lasciiyeah, I missed the mark there17:42
superdanheh17:42
openstackgerritFeodor Tersin proposed openstack/nova: libvirt: Fix wrong _is_booted_from_volume usages  https://review.openstack.org/33460917:42
openstackgerritFeodor Tersin proposed openstack/nova: Delete duplicate check when live-migrating  https://review.openstack.org/33705617:42
openstackgerritFeodor Tersin proposed openstack/nova: libvirt: Simplify internal usage of get_instance_disk_info  https://review.openstack.org/33490117:42
*** rrecio has quit IRC17:44
*** ihrachys has quit IRC17:46
*** Swami has joined #openstack-nova17:46
armaxsdague, superdan: ping re https://review.openstack.org/#/c/351876/17:46
armaxanything I can do to help? I am not quite sure I am fully up to speed on the latest eents17:46
armaxevents17:46
*** oanson has quit IRC17:46
*** shaohe_feng has quit IRC17:47
superdanarmax: you can abandon17:47
superdanarmax: I have proper fixes proposed17:47
superdanarmax: https://review.openstack.org/#/c/353670/17:47
armaxsuperdan: you are indeed super then! it’s not just the nick17:47
*** shaohe_feng has joined #openstack-nova17:47
superdanarmax: only on fridays17:48
*** zenoway has quit IRC17:48
armaxsuperdan: still, that’s quite a feat!17:48
superdanheh17:48
*** zenoway has joined #openstack-nova17:48
*** EinstCrazy has quit IRC17:48
kevinbentonsuperdan: armax doesn't allow us to have fun nicknames on fridays :(17:49
kevinbenton:P17:49
superdankevinbenton: what a tyrant17:49
armaxme?17:49
armaxkevinbenton finds every opportunity to be mean to me :’(17:49
*** tonytan4ever has quit IRC17:49
superdanheh17:49
*** sdague has quit IRC17:51
kevinbentonsuperdan: i see here that https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L256-L257 is not privileged if get_client is called17:51
kevinbentondo you happen to know if a privileged client is ever passed in?17:52
superdankevinbenton: no, I'd have to grep17:52
armaxjohnthetubaguy: ^17:52
armaxjohnthetubaguy: I brought in the big guns17:52
*** zenoway has quit IRC17:53
kevinbentoni can dig into it if nobody knows off the top of their head17:53
johnthetubaguyarmax: :)17:53
armaxkevinbenton: see? I am only making compliments17:53
johnthetubaguyso I think a privileged client is passed in17:53
johnthetubaguywhen port_binding is enabled17:53
johnthetubaguybut I would have to go double check17:53
johnthetubaguyoh wait... its simpler17:54
*** devananda has quit IRC17:54
*** krtaylor has quit IRC17:54
kevinbentonjohnthetubaguy: can we always use an unprivileged client in this particular chunk of code?17:54
johnthetubaguyhmm, no its not... I was going to say we call it in a periodic task with admin, but I am not sure we hit that side of the else clause in that case17:54
johnthetubaguykevinbenton: so I have a plan to split this into two methods17:55
johnthetubaguykevinbenton: one for when we have a list of net_ids, the other when we don't17:55
johnthetubaguykevinbenton: once we do that, it should be much clearer17:55
kevinbentonjohnthetubaguy: ok, well the issue is that the net listings should be done as a tenant17:55
*** sulmer has joined #openstack-nova17:55
johnthetubaguykevinbenton: so we can't in the periodic task to update the network_info (at the moment, at least)17:56
johnthetubaguykevinbenton: but I think we are really just fetching a specific list in that case, but I could be remembering wrongly17:56
*** shaohe_feng has quit IRC17:57
kevinbentonjohnthetubaguy: if it's fetching IDs, that's cool17:57
kevinbentonjohnthetubaguy: the problem is that listing networks as an admin doesn't really get you all of the networks that a tenant might be able to access17:57
johnthetubaguykevinbenton: yeah, not 100% on that, will need to trace to double check17:57
*** shaohe_feng has joined #openstack-nova17:57
kevinbentonjohnthetubaguy: RBAC (added in Liberty) allows a network to be shared to a specific tenant17:57
kevinbentonjohnthetubaguy: and it won't show 'shared' = True to the admin17:58
johnthetubaguykevinbenton: I think if we split that in two, we should be able to list shared and unshared in the single call I guess17:58
johnthetubaguykevinbenton: ah... gotcha17:58
johnthetubaguykevinbenton17:58
johnthetubaguyoops17:58
johnthetubaguykevinbenton: I guess we probably shouldn't "fix" that, as it will break existing clients17:58
johnthetubaguyknowing it works with the user tenant is good info, we should be able to work with17:59
kevinbentonjohnthetubaguy: how would it break clients?17:59
*** oanson has joined #openstack-nova17:59
johnthetubaguykevinbenton: if they rely on that not returning the shared networks, they will get a nasty surpize17:59
kevinbentonjohnthetubaguy: but it does return shared networks18:00
*** sulmer has quit IRC18:00
johnthetubaguykevinbenton: well, when your not admin only I guess?18:00
*** lpetrut has quit IRC18:00
*** sdague has joined #openstack-nova18:00
johnthetubaguykevinbenton: I may just be confused, its 7pm on a friday for me18:01
kevinbentonjohnthetubaguy: sorry, let me clarify. Right now this function is attempting to get all networks a tenant can access, shared and owned by the tenant18:01
johnthetubaguyack18:01
kevinbentonjohnthetubaguy: but it's doing this with admin creds, so querying on the shared flag won't show ones shared to specific tenants18:01
*** TRUMP-PENCE has quit IRC18:02
*** fragatin_ has quit IRC18:02
kevinbentonjohnthetubaguy: however, when a tenant does a net-list, they will see networks shared to them as 'shared', whether it's via RBAC or globally shared18:02
kevinbentonjohnthetubaguy: so basically right now this method is just missing some networks because it doesn't have the right perspective18:02
johnthetubaguykevinbenton: ah... I missed the RBAC vs global thing18:02
johnthetubaguykevinbenton: if we drop the shared:false flag, would we just get all the shared ones (assuming its not admin with the above bug)18:03
*** lpetrut has joined #openstack-nova18:03
kevinbentonjohnthetubaguy: yep, one single query with the tenant would get you all networks that tenant is allowed to access18:03
*** TRUMP-PENCE has joined #openstack-nova18:03
johnthetubaguykevinbenton: awesome, so that does clear things up18:03
johnthetubaguykevinbenton: I am creating a big list of TODOs for next cycle, so this really clears that one up nicely :)18:04
kevinbentonjohnthetubaguy: cool18:04
kevinbentonjohnthetubaguy: technically this will even work now if a regular VM boot does these listing operations unprivileged18:05
johnthetubaguykevinbenton: this is the stuff, I will add some comments from your info, so I don't forget them over the weekend! https://review.openstack.org/#/c/353982/18:05
johnthetubaguykevinbenton: cool, yeah, I hope we can get away with doing just that18:05
kevinbentonjohnthetubaguy: awesome18:06
kevinbentonjohnthetubaguy: i always forget about the auto net listing looking for things to latch onto18:06
*** shaohe_feng has quit IRC18:07
johnthetubaguykevinbenton: there is always one more thing in that code18:07
johnthetubaguykevinbenton: I added a comment here, for me to tidy up those notes: https://review.openstack.org/#/c/353982/2/specs/ocata/approved/neutron-interaction-refresh-ocata.rst@19618:08
*** lpetrut has quit IRC18:08
*** shaohe_feng has joined #openstack-nova18:08
johnthetubaguyarmax: kevinbenton: thanks for helping clear that up, sounds promising18:08
kevinbentonjohnthetubaguy: no prob. ultimately i would like to get nova to not use the admin client for anything other than background information refresh like you mentioned18:10
kevinbentonjohnthetubaguy: and setting port binding details18:10
johnthetubaguykevinbenton: yeah, we might actually be quite close to that already, (kinda by accident)18:11
johnthetubaguykevinbenton: well, we do update_port with the admin client, which does loads of things including port binding, but thats more an efficiency thing18:12
kevinbentonexcellent. it's just always something that worries me because it means nova has to approximate everything the policy engine in neutron would do18:12
johnthetubaguykevinbenton: +118:12
*** ftersin has left #openstack-nova18:13
johnthetubaguykevinbenton: we might actually want some observer role for that background refresh, longer term18:13
*** rama_y has quit IRC18:13
*** parora has joined #openstack-nova18:14
* johnthetubaguy goes to get some food18:14
kevinbentonjohnthetubaguy: yeah, that makes sense18:15
kevinbentonjohnthetubaguy: also getting good makes sense18:15
kevinbentonjohnthetubaguy: ttyl :)18:15
*** baoli has joined #openstack-nova18:17
*** shaohe_feng has quit IRC18:17
*** prateek_ has quit IRC18:18
*** shaohe_feng has joined #openstack-nova18:18
*** sdake has joined #openstack-nova18:21
*** jichen has joined #openstack-nova18:25
*** devananda has joined #openstack-nova18:26
openstackgerritJohn Garbutt proposed openstack/nova: Improve help text for interval_opts  https://review.openstack.org/33176118:27
*** shaohe_feng has quit IRC18:28
*** shaohe_feng has joined #openstack-nova18:28
*** rlrossit has joined #openstack-nova18:30
*** rlrossit has quit IRC18:35
*** rlrossit has joined #openstack-nova18:37
*** shaohe_feng has quit IRC18:38
*** shaohe_feng has joined #openstack-nova18:38
*** krtaylor has joined #openstack-nova18:39
*** devkulkarni1 has joined #openstack-nova18:40
*** amotoki has quit IRC18:42
*** rlrossit has quit IRC18:42
*** lpetrut has joined #openstack-nova18:43
*** lpetrut has quit IRC18:43
*** rama_y has joined #openstack-nova18:44
*** sdake has quit IRC18:45
*** wasmum has quit IRC18:47
*** agireud has quit IRC18:47
*** shaohe_feng has quit IRC18:48
*** sulmer has joined #openstack-nova18:48
*** rlrossit_ has joined #openstack-nova18:48
*** shaohe_feng has joined #openstack-nova18:48
*** wasmum has joined #openstack-nova18:49
*** tonytan4ever has joined #openstack-nova18:50
mriedemkevinbenton: neutron midcycle?18:53
*** mgoddard_ has joined #openstack-nova18:54
*** agireud has joined #openstack-nova18:55
*** tonytan4ever has quit IRC18:55
mriedemheh ireland next week?18:55
mriedemnope18:55
*** daidv has quit IRC18:55
superdanmriedem: you wanna slam some easy patches for the cells upgrade testing?18:56
mriedemlike, summer slam?18:57
mriedem'9518:57
mriedem?18:57
superdanyes exactly like that18:57
superdanhttps://review.openstack.org/#/c/354261/18:57
superdanalthough leakypipes still has a -1 on the bottom, but I think he'll probably acquiesce18:57
mriedemjaybird is -118:57
*** mgoddard has quit IRC18:57
*** shaohe_feng has quit IRC18:58
*** shaohe_feng has joined #openstack-nova18:59
*** mgoddard_ has quit IRC18:59
mriedemwill take a look18:59
superdanneed to land those before I can hassle sdague about the grenade ones18:59
sdaguemriedem: ... notable thing, the service catalog 3 card monty trick here does mean we're sending all the same headers to the other services that we would send to nova19:00
sdaguewhich, shouldn't be a problem19:00
sdaguebut I did notice the mv headers are there19:00
mriedemsdague: you lost me19:00
mriedemand have me thinking of mordred19:00
sdagueno, I'm talking to you19:01
mriedemoh the context manager19:01
sdaguethis is the neutron call we make19:01
sdagueDEBUG (session:337) REQ: curl -g -i -X GET http://10.42.0.53:9696/v2.0/networks?name=private -H "OpenStack-API-Version: compute 2.35" -H "User-Agent: python-novaclient" -H "Accept: application/json" -H "X-OpenStack-Nova-API-Version: 2.35" -H "X-Auth-Token: {SHA1}8510eb704356f855f2c028fdb9b3c5c6dc44da83"19:01
mriedemmv = microversion, not move, got it :)19:01
mriedemyeah, ok19:01
sdagueah, yeh, my bad19:01
*** fragatina has joined #openstack-nova19:01
mriedemi mean, we could filter those out at some point if we cared19:01
sdagueright19:01
sdagueI don't think they matter19:01
mriedemshouldn't19:02
sdaguebut I'm at least telling someone else they exist19:02
sdagueso it's not just a thing I figured out, then forgot later19:02
mriedemwatch out for buses19:02
*** fifieldt has quit IRC19:03
*** kevinz has joined #openstack-nova19:06
openstackgerritSean Dague proposed openstack/python-novaclient: WIP: use neutron for network name -> id resolution  https://review.openstack.org/35498119:06
sdaguemriedem: ok, that's the possitive through case for network lookup19:07
sdagueit needs the version switching logic19:07
*** krtaylor has quit IRC19:07
sdaguebut that seems to work fine on my devstack19:07
mriedemsuperdan: lascii: leakypipes: comments inline https://review.openstack.org/#/c/354261/19:08
mriedemit's ok, but some usability comments inline19:08
*** krtaylor has joined #openstack-nova19:08
*** shaohe_feng has quit IRC19:09
superdanmriedem: thanks.. what's the translation policy? not everything in that file is marked19:09
*** code-R has joined #openstack-nova19:09
superdanso I assumed not19:09
*** shaohe_feng has joined #openstack-nova19:09
*** tonytan4ever has joined #openstack-nova19:10
mriedemit's user facing so i assume it's supposed to be translated19:10
*** code-R_ has joined #openstack-nova19:10
*** kevinz has quit IRC19:11
superdanit's not really intended to be, but fine19:13
*** code-R has quit IRC19:13
mriedemi'm not -1 on that19:14
mriedemwas just saing19:14
openstackgerritAndreas Jaeger proposed openstack/python-novaclient: Move other-requirements.txt to bindep.txt  https://review.openstack.org/35501419:14
mriedem*saying19:14
superdanyeah, it's fine I just didn't know19:14
*** sulmer has quit IRC19:15
*** sulmer has joined #openstack-nova19:15
cfriesen_does anyone know what the situation is with cpu hotplug for x86-64 in qemu?19:16
cfriesen_I suppose that's offtopic, so never mind. :)19:17
*** raunak has quit IRC19:17
*** fifieldt has joined #openstack-nova19:18
*** shaohe_feng has quit IRC19:19
*** shaohe_feng has joined #openstack-nova19:20
sdaguemriedem: does the net-name currently work with neutron backends?19:24
mriedemyeah19:24
sdagueah, I lost a return19:27
sdaguemy bad19:27
*** thorst_ has quit IRC19:27
superdansdague: I know you're busy but if you could skim those grenade patches for obvious "I will never let you do this" things I'd appreciate it19:27
*** devkulkarni2 has joined #openstack-nova19:28
superdanmriedem: I was doing that thing you just said19:28
openstackgerritDan Smith proposed openstack/nova: Make simple_cell_setup idempotent  https://review.openstack.org/35473419:28
openstackgerritDan Smith proposed openstack/nova: Add cell_v2 verify_instance command  https://review.openstack.org/35426119:28
*** fragatina has quit IRC19:28
superdanwhich is here ^19:28
*** devkulkarni1 has quit IRC19:29
*** fragatina has joined #openstack-nova19:29
*** fragatina has quit IRC19:29
openstackgerritSean Dague proposed openstack/python-novaclient: WIP: use neutron for network name -> id resolution  https://review.openstack.org/35498119:29
*** shaohe_feng has quit IRC19:29
sdaguesuperdan: yeh, I think I got that novaclient stuff as far as I'm going today19:29
sdaguemriedem: the novaclient patch there has things working for real, I'm sure it blows unit tests all over the place19:30
superdansdague: okay, starts here: https://review.openstack.org/#/c/354264/19:30
*** shaohe_feng has joined #openstack-nova19:30
mriedemsdague: yeah i'll pick it up from here19:30
mriedemthanks19:30
superdansdague: mriedem, lascii, and leakypipes are beating me up over the nova manage thing you made me add, in case it gets me any sympathy points19:30
lasciiI prefer to call it "polishing"19:31
superdanwell, it does involve rubbing19:32
mriedemtsk tsk19:32
mriedemalignment brotha19:32
superdanwat19:32
superdangah19:33
openstackgerritDan Smith proposed openstack/nova: Make simple_cell_setup idempotent  https://review.openstack.org/35473419:33
superdanshit19:33
sdaguesuperdan: ok, patch one seems functionally fine. I think we should just make sure that pre/post is the names we want to live forever as this is pretty static interface. (I'm wondering if old/new or base/target are better / worse given existing terminology)19:34
mriedemwe could also validate the uuid is a uuid, but i'm really nit picking now19:34
sdaguehttps://review.openstack.org/#/c/354265/7/projects/60_nova/upgrade.sh ...19:34
sdaguethe create_cell function has be worried19:34
sdaguethat a nova-manage command?19:34
openstackgerritDan Smith proposed openstack/nova: Make simple_cell_setup idempotent  https://review.openstack.org/35473419:34
openstackgerritDan Smith proposed openstack/nova: Add cell_v2 verify_instance command  https://review.openstack.org/35426119:34
superdansdague: no that's in lascii's devstack patch19:34
lasciihttps://review.openstack.org/#/c/322311/4/lib/nova19:34
superdansdague: also, the logs show it works properly, just FYI19:35
superdansdague: fine if you want to nit over pre/post, just tell me what to make them19:35
sdagueok, I would rather put that nova-manage command inline in grenade19:35
superdansdague: really? I had it that way and was SURE you'd make me change it :)19:35
*** thorst_ has joined #openstack-nova19:35
sdagueno, on the upgrade side I really want things explicit for everything except start19:36
superdanokay19:36
*** devkulkarni2 has quit IRC19:37
sdaguethe upgrade scripts try to be as unbound to devstack as we can, so that other tools can glean what we are doing for upgrade more easily19:37
superdansdague: this sets off a big set of dependent patches when I push, so.. tell me what you want the pre/post to be and I'll push it19:37
*** devkulkarni1 has joined #openstack-nova19:38
superdanbase/target?19:38
sdagueyeh, give me a sec, I saw these for the first time in code 5 minutes ago :)19:38
superdanI demand an answer now19:38
*** shaohe_feng has quit IRC19:39
sdagueok, how about we go for explicit and do pre-upgrade / post-upgrade, which means if we need to add other phases later we're not doing things like pre-pre19:40
*** shaohe_feng has joined #openstack-nova19:40
superdanokay19:40
*** thorst_ has quit IRC19:40
sdagueso, if you do that change, and unwind the create_cell call in the grenade script, I'm good with those 219:40
sdagueand will +2 them19:41
superdanmoments19:41
*** oanson has quit IRC19:41
*** sulmer has quit IRC19:41
*** devkulkarni1 has quit IRC19:41
*** devkulkarni1 has joined #openstack-nova19:41
*** sulmer has joined #openstack-nova19:41
superdansdague: done.. you don't have to +2 now, I just wanted some early smoke signals19:43
superdanwe have to land the two (tiny) nova patches first anyway19:43
*** sulmer has quit IRC19:46
*** matrohon has joined #openstack-nova19:49
*** liusheng has quit IRC19:50
*** shaohe_feng has quit IRC19:50
*** sulmer has joined #openstack-nova19:50
*** liusheng has joined #openstack-nova19:50
*** shaohe_feng has joined #openstack-nova19:50
*** Apoorva has quit IRC19:52
mriedemi'm +2 on the bottom nova change19:53
*** mvk has quit IRC19:54
superdanthanks19:56
*** niloofar has joined #openstack-nova19:56
niloofarHi all19:57
niloofarIs anyone know How can I stop nova from writing log to local disk?19:57
niloofarI have rsylog server setup, and I dont like to keep the log on local server19:58
*** gouthamr has quit IRC19:59
*** shaohe_feng has quit IRC20:00
paroraniloofar, you could try changing the log file destination in the nova.conf20:00
*** shaohe_feng has joined #openstack-nova20:01
niloofarbut it will still in local disk20:01
sdaguemriedem: ok, I'm fading out for the week. I'll look back in the state of patches on monday, and see where we are at from getting this stuff landed.20:01
*** devkulkarni1 has quit IRC20:01
mriedemyup i just pulled it down20:01
mriedemwill hack on this a bit20:01
niloofarI like to log written locally on the server20:01
paroraniloofar, how about putting it as /dev/null20:01
parora?20:01
superdansdague: thanks for hitting those20:01
openstackgerritPushkar Umaranikar proposed openstack/nova: Improve help text for compute manager options  https://review.openstack.org/32963920:02
niloofaryou mean in nova.conf for log path I change it to logdir=/dev/null  ?20:03
paroraniloofar, yeah, try it on a test server first but redirecting your logs to /dev/null should leave you log less20:04
*** mnestratov has quit IRC20:04
niloofarThank you very much, I will git it atry20:04
paroraniloofar, ywc20:04
*** raunak has joined #openstack-nova20:05
mriedemmordred: did you say that you can get network details in neutron via name or id?20:05
mriedemmordred: was wondering if we can just do a /networks/name here https://review.openstack.org/#/c/354981/2/novaclient/v2/networks.py@5420:06
* mordred looks20:06
paroraoomichi, can you please look into https://review.openstack.org/#/c/327706/  when you have time, i have tried to address your review comments20:06
niloofarI just tried20:07
niloofarI changed logdir tp /var/null20:07
niloofarI restarted my nova-compute few times20:07
parora*/dev/null20:07
niloofarsorry typo I meant /dev/null20:08
*** xyang1 has quit IRC20:08
niloofarbut I still see the logs20:08
niloofarany idea?20:09
mriedemniloofar: so, this isn't a support channel20:10
mriedemniloofar: these are questions better asked in the #openstack channel, or ask.openstack.org, or the general openstack mailing list20:10
*** shaohe_feng has quit IRC20:10
*** shaohe_feng has joined #openstack-nova20:11
niloofarI see, I already ask there as well20:11
niloofarI was hopping there is a setting I can use in nova.conf file20:12
niloofarwhich a developer might know20:12
openstackgerritPushkar Umaranikar proposed openstack/nova: Removing duplicates from columns_to_join list  https://review.openstack.org/34229920:15
*** lpetrut has joined #openstack-nova20:16
mordredmriedem: hrm. I apparently may have lied - it seems possible now that the magic happens in neutronclient20:16
mordredmriedem: I just did some testing using a direct requests session20:16
mriedemok20:17
mordred>>> c.get('/v2.0/networks/6d6357ac-0f70-4afa-8bd7-c274cc4ea235').json()20:17
mordredworks20:17
mordred>>> c.get('/v2.0/networks?name=public').json()20:17
mordredworks20:17
mordredbut >>> c.get('/v2.0/networks?name=6d6357ac-0f70-4afa-8bd7-c274cc4ea235').json()20:17
mordredreturns []20:17
*** gouthamr has joined #openstack-nova20:17
mordredas does >>> c.get('/v2.0/networks?id=public').json()20:18
mordredso - sorry for the misinformation20:18
*** shaohe_feng has quit IRC20:20
*** shaohe_feng has joined #openstack-nova20:21
*** xyang1 has joined #openstack-nova20:22
mriedemnp, thanks for looking it up for me :)20:23
mriedemi'm lazy20:23
superdanmordred: you can read that as "I never trust anything you say so I am not surprised"20:24
* superdan decodes mriedem speak for a small fee20:24
*** yamahata has quit IRC20:25
mordredsuperdan: :)20:26
*** tri2sing has quit IRC20:26
mordredsuperdan: to be clear, I also never trust anything I say20:26
mordredit's one of the topics on which mriedem and I agree20:26
superdanheh20:27
*** Apoorva has joined #openstack-nova20:27
*** shaohe_feng has quit IRC20:31
*** shaohe_feng has joined #openstack-nova20:31
*** sdake has joined #openstack-nova20:32
*** ihrachys has joined #openstack-nova20:33
*** niloofar has quit IRC20:39
*** haplo37_ has joined #openstack-nova20:39
*** shaohe_feng has quit IRC20:41
*** shaohe_feng has joined #openstack-nova20:42
*** parora has quit IRC20:42
openstackgerritjichenjc proposed openstack/nova: not use cert and private key in glance client  https://review.openstack.org/35470620:42
*** mingyu has joined #openstack-nova20:43
*** tblakes has quit IRC20:44
*** rfolco has quit IRC20:46
*** mingyu has quit IRC20:47
*** agireud has quit IRC20:48
*** jichen has quit IRC20:49
-openstackstatus- NOTICE: The Gerrit service on review.openstack.org is restarting for a scheduled upgrade, but should return to service momentarily: http://lists.openstack.org/pipermail/openstack-dev/2016-August/101394.html20:50
*** shaohe_feng has quit IRC20:51
mriedemwtf, where is nova.network.model.Route.interface set?20:51
mriedemfor ovs runs?20:51
*** oomichi has quit IRC20:51
*** oomichi has joined #openstack-nova20:52
*** shaohe_feng has joined #openstack-nova20:52
openstackgerritMichael Glaser proposed openstack/nova: Check opt consistency for neutron.py  https://review.openstack.org/34960520:53
*** agireud has joined #openstack-nova20:55
*** gouthamr has quit IRC20:58
*** ducttape_ has quit IRC20:58
-openstackstatus- NOTICE: The Mediawiki service at wiki.openstack.org will be offline from 21:00 UTC until approximately 23:00 UTC for a planned upgrade http://lists.openstack.org/pipermail/openstack-dev/2016-August/101395.html20:58
*** ChanServ changes topic to "The Mediawiki service at wiki.openstack.org will be offline from 21:00 UTC until approximately 23:00 UTC for a planned upgrade http://lists.openstack.org/pipermail/openstack-dev/2016-August/101395.html"20:58
*** shaohe_feng has quit IRC21:01
mriedemsean-k-mooney: leakypipes: you around?21:02
*** shaohe_feng has joined #openstack-nova21:02
mriedemwe're blowing up on not setting this in the os-vif code in nova https://github.com/openstack/os-vif/blob/master/os_vif/objects/route.py#L2821:03
mriedemhttps://bugs.launchpad.net/nova/+bug/161281221:03
openstackLaunchpad bug 1612812 in OpenStack Compute (nova) "OS-VIF: Interface missing from Route Object. (OVS)" [Undecided,New]21:03
*** burgerk has quit IRC21:03
patrickeastmriedem: hey, i added a comment on https://review.openstack.org/#/c/354502/ about the os-brick lock dir thingy21:03
mriedemfor neutron the route gets added here https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L215821:03
mriedembut interface isn't set21:04
patrickeastmriedem: let me know what you'd like to have happen there, where its more info or a bug or whatever21:04
mriedemhttps://github.com/openstack/nova/blob/master/nova/network/model.py#L22321:04
mriedemset on the os-vif route object here https://github.com/openstack/nova/blob/master/nova/network/os_vif_util.py#L13921:04
mriedembut we don't ever set nova.network.model.Route.interface anywhere in nova21:04
mriedemAND21:04
mriedemi don't see anything in the neutron data model for Route.interface21:04
mriedemhttps://github.com/openstack/neutron/blob/master/neutron/objects/subnet.py#L6721:05
mriedemarmax: HenryG: ^21:05
mriedemam i missing something?21:05
doffmos-vif object with presumably non-optional interface field here https://github.com/openstack/os-vif/blob/master/os_vif/objects/route.py#L2821:05
mriedemor was Route.interface just added to nova's nw model but never used?21:05
mriedemapparently we don't fail on this in the openstack CI b/c we aren't using subnets with routes21:06
mriedemhttp://logs.openstack.org/99/340599/2/gate/gate-tempest-dsvm-neutron-full-ubuntu-xenial/3128550/logs/screen-n-cpu.txt.gz#_2016-08-12_17_08_55_24821:06
mriedem'routes': []21:06
*** lpetrut has quit IRC21:07
*** pwnall has quit IRC21:07
mriedemso looks to me like a lie was propagated to os-vif21:07
mriedemand the interface field should be nullable=True21:07
* mriedem wishes some os-vif peeps were around21:08
*** pwnall has joined #openstack-nova21:08
*** pwnall1 has joined #openstack-nova21:10
*** pwnall has quit IRC21:10
*** fragatina has joined #openstack-nova21:11
*** shaohe_feng has quit IRC21:12
*** shaohe_feng has joined #openstack-nova21:12
*** rfolco has joined #openstack-nova21:16
*** tblakes has joined #openstack-nova21:16
*** yamahata has joined #openstack-nova21:16
*** rfolco has quit IRC21:16
*** rlrossit_ has quit IRC21:16
openstackgerritAlberto Murillo proposed openstack/nova: linux_net.py adds --addn-hosts parameter to dnsmasq when using nova network in multi_host mode.  https://review.openstack.org/21062721:18
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Use neutron for network name -> id resolution  https://review.openstack.org/35498121:18
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Look up image names directly in glance  https://review.openstack.org/35434921:18
*** fragatina has quit IRC21:19
*** sulmer has quit IRC21:21
*** edmondsw has quit IRC21:22
*** shaohe_feng has quit IRC21:22
*** shaohe_feng has joined #openstack-nova21:23
*** sdake has quit IRC21:23
*** lascii is now known as alaski21:23
*** lpetrut has joined #openstack-nova21:27
*** oomichi has quit IRC21:28
openstackgerritMatt Riedemann proposed openstack/python-novaclient: Use neutron for network name -> id resolution  https://review.openstack.org/35498121:28
*** oomichi has joined #openstack-nova21:29
*** dimtruck is now known as zz_dimtruck21:29
*** tonytan4ever has quit IRC21:30
*** rama_y has quit IRC21:31
*** Hosam has joined #openstack-nova21:32
*** shaohe_feng has quit IRC21:32
*** shaohe_feng has joined #openstack-nova21:32
openstackgerritmelanie witt proposed openstack/nova: Make req spec and RT aware of volume-backed instances  https://review.openstack.org/35509121:34
*** leakypipes has quit IRC21:34
openstackgerritMatt Riedemann proposed openstack/os-vif: Make Route.interface nullable  https://review.openstack.org/35509321:35
mriedemsean-k-mooney: ^21:35
armaxmriedem: what are you missing in the RouterInterface model?21:39
*** ihrachys has quit IRC21:39
mriedemarmax: this is Route.interface21:40
mriedemsubnet.host_routes21:40
mriedemis a list of routes21:40
mriedembut there is no 'interface' field on those routes21:40
mriedembut nova's nw info model assumes there is21:40
mriedemand os-vif expects it to be non-None21:40
mriedemhttps://github.com/openstack/os-vif/blob/master/os_vif/objects/route.py#L28 is non-nullable21:40
armaxum21:40
mriedemnova sets it here https://github.com/openstack/nova/blob/master/nova/network/os_vif_util.py#L13921:41
*** jamesdenton has quit IRC21:41
mriedembut nova net sets that field https://github.com/openstack/nova/blob/master/nova/network/neutronv2/api.py#L215821:41
armaxwe never had an interface attribute on router interfaces21:41
mriedemso it's always None21:41
mriedemarmax: ha, yeah :)21:41
armaxever21:41
mriedemcool21:41
mriedemcan you ack that os-vif patch?21:41
mriedemi'm doing a workaround in nova for now21:41
* armax checks21:41
mriedemmaybe some out of tree enablement BS at some point21:41
mriedemor another bp that never materialized21:42
mriedembut if you have a subnet with host_routes set, you're busted right now21:42
*** gszasz has quit IRC21:42
armaxquestion is21:42
armaxdo you think we should have one?21:42
*** shaohe_feng has quit IRC21:42
*** lpetrut has quit IRC21:42
*** shaohe_feng has joined #openstack-nova21:43
mriedemno21:43
mriedemwell, nova never tries to set one21:43
*** Hosam_ has joined #openstack-nova21:43
mriedemsuperdan: field=None isn't the same as field not set right?21:43
mriedemso if i want to avoid "ValueError: Fieldinterface' cannot be None" i can just not set the field, right?21:44
superdanmriedem: correct, field=None is "set..to None"21:44
superdanmriedem: you can yes21:44
mriedemok21:44
armaxmriedem: are you sure you don’t want to use the interface field to carry over a port id or a subnet id?21:44
doffminterface="FIXME:Workaround"21:44
armaxmriedem: but to be fair21:44
armaxmriedem: if you do do need that, you’d rather name it after what’s exposed by neutron no?21:45
*** sameo_ has joined #openstack-nova21:45
*** Hosam has quit IRC21:46
*** sameo has quit IRC21:46
*** Hosam has joined #openstack-nova21:46
*** jistr has quit IRC21:47
mriedemarmax: i don't know we'd all of a sudden need a thing that's never been set in nova before21:48
mriedemand isn't modeled in neutron, at least not in that way21:48
armaxmriedem: agreed21:48
*** fragatina has joined #openstack-nova21:48
*** fragatina has quit IRC21:50
*** Hosam_ has quit IRC21:50
openstackgerritPushkar Umaranikar proposed openstack/nova: Api-ref: Improve os-migrateLive input parameters  https://review.openstack.org/34565221:50
*** haplo37_ has quit IRC21:50
*** fragatina has joined #openstack-nova21:50
*** jistr has joined #openstack-nova21:51
openstackgerritMatt Riedemann proposed openstack/nova: os-vif: do not set Route.interface if None  https://review.openstack.org/35509821:51
mriedemsuperdan: alaski: melwitt: can you take a quick look at ^ please? this is blowing up our preprod cloud internally, it's pretty obviously wrong21:51
mriedemi'd bug danpb or jaypipes but they aren't around21:52
melwittsure21:52
*** xyang1 has quit IRC21:52
*** shaohe_feng has quit IRC21:53
superdanmriedem: seems like you're not testing that path21:53
*** shaohe_feng has joined #openstack-nova21:53
mriedemwhich one?21:54
*** thorst_ has joined #openstack-nova21:54
superdanthe "if is None" pth21:54
mriedemit does here https://review.openstack.org/#/c/355098/1/nova/tests/unit/network/test_os_vif_util.py21:55
superdanand I'm not sure why you're removing that interface21:55
mriedemnova.network.model.Route defaults interface to None21:55
superdanoh I see21:55
mriedemtest_nova_to_osvif_routes tests with 2 routes, one set and one not21:55
mriedemsetting it to None is the bug21:55
superdanokay21:55
mriedemalso, os-vif doesn't use the Route.interface anywhere either21:55
mriedemso it's just totally vestigial21:55
mriedemi didn't want to remoe the nova.network.model stuff until danpb could confirm that on monday though21:56
mriedemhence the FIXME21:56
superdanyeah21:56
openstackgerritKen'ichi Ohmichi proposed openstack/nova: Change hypervisor_type to actual value  https://review.openstack.org/35509921:56
alaskilooks like melwitt and superdan are on it, but I can ack if needed. The patch looks reasonable to me21:57
mriedemthanks, i have to head out to get maya21:58
mriedemi'll get https://review.openstack.org/#/c/354734/ later tonight21:59
*** mriedem has quit IRC21:59
*** tanvirt has quit IRC21:59
*** pradk has quit IRC22:00
*** ijw has joined #openstack-nova22:00
*** jistr has quit IRC22:00
*** Hosam has quit IRC22:02
*** shaohe_feng has quit IRC22:03
*** shaohe_feng has joined #openstack-nova22:04
openstackgerritPushkar Umaranikar proposed openstack/nova: Api-ref: Improve os-migrateLive input parameters  https://review.openstack.org/34565222:05
openstackgerritmelanie witt proposed openstack/nova: Make req spec and RT aware of volume-backed instances  https://review.openstack.org/35509122:06
*** ducttape_ has joined #openstack-nova22:07
*** sdague has quit IRC22:07
*** jistr has joined #openstack-nova22:09
*** tomaz_ has quit IRC22:10
*** mdrabe has quit IRC22:11
*** beekneemech is now known as bnemec-pto22:12
*** shaohe_feng has quit IRC22:13
*** cdent has quit IRC22:13
*** shaohe_feng has joined #openstack-nova22:14
*** kylek3h has quit IRC22:16
*** jistr has quit IRC22:19
*** TravT has quit IRC22:21
*** tblakes has quit IRC22:21
*** shaohe_feng has quit IRC22:23
*** tonytan4ever has joined #openstack-nova22:25
*** TravT has joined #openstack-nova22:25
*** jistr has joined #openstack-nova22:25
*** shaohe_feng has joined #openstack-nova22:26
*** rama_y has joined #openstack-nova22:27
*** ijw has quit IRC22:28
*** shaohe_feng has quit IRC22:34
*** shaohe_feng has joined #openstack-nova22:34
*** ducttape_ has quit IRC22:35
*** ducttape_ has joined #openstack-nova22:36
*** armax has quit IRC22:40
*** ducttape_ has quit IRC22:40
*** palexster has quit IRC22:42
*** annegentle has quit IRC22:43
*** shaohe_feng has quit IRC22:44
*** shaohe_feng has joined #openstack-nova22:44
*** code-R_ has quit IRC22:50
*** markvoelker has quit IRC22:53
*** shaohe_feng has quit IRC22:54
*** shaohe_feng has joined #openstack-nova22:54
*** figleaf is now known as edleafe22:56
-openstackstatus- NOTICE: ok https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong23:00
*** ChanServ changes topic to "ok https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong"23:00
*** vishwanathj has quit IRC23:02
*** armax has joined #openstack-nova23:04
*** shaohe_feng has quit IRC23:04
*** shaohe_feng has joined #openstack-nova23:05
*** ChanServ changes topic to "This channel is for Nova development. For support of Nova deployments, please use #openstack. Please see: https://wiki.openstack.org/wiki/Nova/Newton_Release_Schedule"23:06
-openstackstatus- NOTICE: https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong23:06
*** jlk has joined #openstack-nova23:11
*** sameo_ has quit IRC23:12
*** sameo has joined #openstack-nova23:12
*** shaohe_feng has quit IRC23:15
*** shaohe_feng has joined #openstack-nova23:15
jlkHow crazy would I be to try to implement my own scheduler_driver that would sit above the filter_scheduler in order to send some requests through the regular host_manager and other requests through the ironic host manager? (obviously with different filters used in each of those managers).23:15
jlkI'm trying to figure out a way to implement VMs and bare metal in the same cloud without going cells v123:15
*** thorst_ has quit IRC23:17
*** shaohe_feng has quit IRC23:25
*** shaohe_feng has joined #openstack-nova23:25
*** thorst_ has joined #openstack-nova23:26
*** jwcroppe has joined #openstack-nova23:26
*** thorst_ has quit IRC23:30
*** rrecio_ has quit IRC23:31
*** raunak has quit IRC23:32
*** rama_y has quit IRC23:33
*** sdake has joined #openstack-nova23:33
*** rama_y has joined #openstack-nova23:33
*** ijw has joined #openstack-nova23:34
*** shaohe_feng has quit IRC23:35
*** shaohe_feng has joined #openstack-nova23:36
*** ijw has quit IRC23:41
*** agireud has quit IRC23:44
*** shaohe_feng has quit IRC23:45
*** ijw has joined #openstack-nova23:46
*** shaohe_feng has joined #openstack-nova23:46
*** rama_y has quit IRC23:47
stewie925hey guys, can anyone share how to do gerrit search on a file path like 'nova/conf/xenserver.py'?23:50
*** ducttape_ has joined #openstack-nova23:53
*** agireud has joined #openstack-nova23:53
*** markvoelker has joined #openstack-nova23:54
*** Swami has quit IRC23:55
*** shaohe_feng has quit IRC23:56
raj_singhstewie925: https://review.openstack.org/#/q/project:openstack/nova+file:xenserver.py+status:open refer this23:56
*** baoli has quit IRC23:56
*** shaohe_feng has joined #openstack-nova23:57
stewie925nahh that wouldnt work23:57
stewie925but yes Ive tried that23:57
stewie925so far the only one that works is using a partial regex  like  "file:^.*/conf/xenserver.py"23:57
stewie925but its okay, I'll try asking in the gerrit room on Monday23:58
stewie925raj_singh: but thanks :)23:58
*** markvoelker has quit IRC23:59
*** gouthamr has joined #openstack-nova23:59
*** yamahata has quit IRC23:59

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