Friday, 2015-05-15

*** tjones1 has quit IRC00:00
*** davideagnello has quit IRC00:01
*** andrewbogott is now known as andrewbogott_afk00:02
*** salv-orlando has quit IRC00:03
*** fawadkhaliq has quit IRC00:04
*** kmartin has quit IRC00:04
*** davideagnello has joined #openstack-nova00:07
*** tjones1 has joined #openstack-nova00:07
*** rms_13 has joined #openstack-nova00:08
*** tjones1 has quit IRC00:09
*** tjones1 has joined #openstack-nova00:11
*** Longgeek has quit IRC00:13
*** Longgeek has joined #openstack-nova00:13
*** IanGovett1 has quit IRC00:14
*** ildikov has joined #openstack-nova00:14
*** mestery has quit IRC00:14
*** dims has quit IRC00:15
*** pixelbeat has quit IRC00:15
*** mriedem has joined #openstack-nova00:20
*** dims has joined #openstack-nova00:23
*** promulo_ has joined #openstack-nova00:23
mriedemBobBall: is xenproject ci going wonky again?00:26
*** tjones1 has quit IRC00:26
*** zsais has quit IRC00:27
*** marun has joined #openstack-nova00:27
mriedemBobBall: actually i'm seeing this:00:27
mriedemhttp://logs.openstack.xenproject.org/20/183220/1/check/dsvm-tempest-xen/dfbf491/logs/screen-n-cpu.txt.gz00:27
mriedemhttp://paste.openstack.org/show/223268/00:27
*** armax has quit IRC00:28
mriedemlibvirtError: internal error: libxenlight failed to restore domain 'instance-0000001b'00:28
mriedemjohnthetubaguy: ^00:28
*** promulo__ has quit IRC00:28
mriedemjohnthetubaguy: also, the spec was approved for this https://review.openstack.org/#/c/69047/00:28
rms_13Experts, running onto an issue where changing the MTU of the interface on one of the compute and than restarting nova-compute fails; I think it is timing out on "MessagingTimeout: Timed out waiting for a reply to message "00:28
rms_13I could see that amqp server on the controller has accepted the connection but than I believe something went wrong on the send side00:29
rms_13Question is: Do I need a restart of an rabbitmq-server if I am changing the MTU of the physical interface ?00:30
*** markvoelker has quit IRC00:32
*** ociuhandu has quit IRC00:33
openstackgerritgaryk proposed openstack/nova: libvirt: improve logging  https://review.openstack.org/18245200:34
mriedemgaryk: you're going to love that ^00:38
garykmriedem: :)00:39
garykmriedem: https://www.youtube.com/watch?v=wsbgwNe4Rv400:39
*** zhenguo has joined #openstack-nova00:41
*** otter768 has joined #openstack-nova00:42
*** vishwana_ has quit IRC00:43
*** READ10 has joined #openstack-nova00:43
openstackgerritgaryk proposed openstack/nova: libvirt: improve logging  https://review.openstack.org/18245200:43
garykmriedem: i have gone with the instance name suggestlon like you said.00:44
garykhopefully jay will be open to the change00:44
mriedemgaryk: so just adding the instance= kwarg and adding _LW right?00:44
mriedem+w00:45
mriedemoops00:45
mriedem+200:45
garykyup, thanks00:45
*** vishwanathj has joined #openstack-nova00:46
*** tsekiyam_ has joined #openstack-nova00:46
*** otter768 has quit IRC00:47
*** zzzeek has quit IRC00:48
*** zhangtralon has joined #openstack-nova00:49
*** mtanino has quit IRC00:49
*** tsekiyama has quit IRC00:49
*** tsekiyam_ has quit IRC00:51
*** apuimedo has quit IRC00:53
*** changbl has joined #openstack-nova00:55
*** markvoelker has joined #openstack-nova00:56
*** patrickeast_ has quit IRC00:57
*** harlowja has quit IRC00:58
*** yamahata has joined #openstack-nova00:59
*** sdake has quit IRC01:03
*** marun has quit IRC01:05
*** sdake has joined #openstack-nova01:05
*** markvoelker has quit IRC01:06
*** browne has quit IRC01:06
*** asrangne__ has quit IRC01:07
*** achanda has quit IRC01:07
*** heyongli has joined #openstack-nova01:12
dansmithhmm, seems like grenade is failing01:13
mriedemdansmith: https://review.openstack.org/#/c/182716/01:14
*** gyee has quit IRC01:14
dansmithis that the fix?01:14
*** Jeffrey4l has joined #openstack-nova01:15
*** heyongli has quit IRC01:22
*** alanf-mc has quit IRC01:23
*** hightall has joined #openstack-nova01:25
*** haomaiwa_ has quit IRC01:26
*** claudiub has quit IRC01:26
*** BadCub is now known as BadCub-Hiding01:26
*** haomaiwa_ has joined #openstack-nova01:28
mriedemdansmith: no, but it's a fix01:29
mriedemdansmith: link to the grenade failure?01:29
dansmithit's failing on getting pip or something01:29
mriedemhttp://logs.openstack.org/47/167447/7/gate/gate-grenade-dsvm/fe73dd5/console.html ?01:29
mriedemfatal: read error: Connection reset by peer01:29
mriedemyeah i think that's been spotty all day01:30
mriedemjogo: ^?01:30
dansmithhttp://logs.openstack.org/10/181110/5/check/check-grenade-dsvm-partial-ncpu/a474398/logs/grenade.sh.txt.gz#_2015-05-15_00_09_39_14801:30
dansmithit seems to only affect the grenade jobs01:30
*** okrieg has joined #openstack-nova01:30
dansmithlike, a bunch of mine in a row only fail grenade, both grenade jobs, and like that01:30
dansmiththat's actually getting a 503 from something01:30
mriedemhttp://goo.gl/Gxp1OJ01:31
mriedemshows it spike and drop off so must be fixed01:31
mriedembut according to logstash it's not just grenade jobs01:31
dansmithI just hit it a bit ago01:31
dansmiththat's different from my thing right?01:31
jogoI wonder what caused it01:32
jogoI think infra was looking into it all day01:32
mriedemoh shit nevermind01:32
*** vilobhmm1 has quit IRC01:33
mriedemi was looking at the wrong logs01:33
* mriedem looks at the actual link01:33
*** vilobhmm1 has joined #openstack-nova01:33
dansmithno hits in logstash, or I suck at logstash (which I do)01:33
dansmithhttp://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiU3ludGF4RXJyb3I6XCIgQU5EIGZpbGVuYW1lOlwibG9ncy9ncmVuYWRlLnNoLnR4dC5nelwiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxNDMxNjUzNjUzMDE4LCJtb2RlIjoiIiwiYW5hbHl6ZV9maWVsZCI6IiJ901:34
*** diegows has quit IRC01:34
dansmithah01:34
dansmithhttp://logstash.openstack.org/#eyJzZWFyY2giOiJtZXNzYWdlOlwiU3ludGF4RXJyb3I6XCIgQU5EIGZpbGVuYW1lOlwibG9ncy9ncmVuYWRlLnNoLnR4dFwiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxNDMxNjUzNjc3Nzg2LCJtb2RlIjoiIiwiYW5hbHl6ZV9maWVsZCI6IiJ901:34
dansmithand not just on my changes, so that's good01:35
mriedemhttp://goo.gl/pJJRhX01:36
*** Longgeek has quit IRC01:36
dansmithyeah, same set as my query above01:36
mriedemasking in infra01:36
*** Longgeek has joined #openstack-nova01:37
openstackgerritDan Smith proposed openstack/nova: API support for migration_type  https://review.openstack.org/18320001:38
openstackgerritDan Smith proposed openstack/nova: Make live migration create a migration object record  https://review.openstack.org/18333101:38
openstackgerritDan Smith proposed openstack/nova: Add migration_type to Migration object  https://review.openstack.org/18111001:38
openstackgerritDan Smith proposed openstack/nova: Add migration_type and hidden to Migration database model  https://review.openstack.org/18110901:38
openstackgerritDan Smith proposed openstack/nova: Allow querying for migrations by source_compute only  https://review.openstack.org/18335301:38
openstackgerritDan Smith proposed openstack/nova: Make evacuate leave a record for the source compute host to process  https://review.openstack.org/18335401:38
openstackgerritDan Smith proposed openstack/nova: Set migration_type for existing cold migrations and resizes  https://review.openstack.org/18319901:38
*** Kevin_Zheng has joined #openstack-nova01:39
*** dsanders1 has quit IRC01:45
*** garyk has quit IRC01:47
*** zzzeek has joined #openstack-nova01:48
*** Sukhdev has joined #openstack-nova01:49
openstackgerritYongli He proposed openstack/nova-specs: test cases for third party CI  https://review.openstack.org/14127001:50
*** sdake_ has joined #openstack-nova01:51
mriedemmelwitt: this looks like a novaclient bug report right? https://bugs.launchpad.net/nova/+bug/145436901:53
openstackLaunchpad bug 1454369 in OpenStack Compute (nova) "nova needs to move to using the 'volume' service type" [Low,Confirmed]01:53
mriedemthe volumev2 gorp is in novaclient01:53
* melwitt looks01:54
*** sdake has quit IRC01:55
melwittmriedem: I'm not entirely sure what is going on there, but the volumev2 is something devstack presents similar to the computev21 stuff. novaclient already defaults to the "volume" endpoint. either way, sounds like a client thing01:58
mriedemok02:00
*** zzzeek has quit IRC02:00
*** diegows has joined #openstack-nova02:00
melwittmriedem: devstack has "volume" as the v1 url, and he disabled v1, so it fails02:00
melwittis what I'm guessing02:00
mriedemo02:01
mriedemi'll admit the service catalog stuff involved here is over my head02:01
mriedembknudson says that it changes all the time02:02
*** Kennan2 has joined #openstack-nova02:02
melwittit's devstack stuff really. it makes an endpoint for each version of every service as different names to "help you" and it causes a lot of confusion02:02
*** Kennan has quit IRC02:03
*** Kennan2 is now known as Kennan02:04
mriedemright, so i thought there was a push at one point to stop helping us in devstack02:06
mriedemand just use version discovery02:06
*** mtanino has joined #openstack-nova02:09
*** signed8bit has quit IRC02:10
mriedemhttps://review.openstack.org/#/c/145613/02:10
*** Sukhdev has quit IRC02:10
mriedemam i remembering correctly that our novaclient volume CLI talks to cinder api directly rather than going through nova volume apis which talks to cinderclient?02:10
mriedemmelwitt: ^?02:10
mriedemb/c if cinderclient has version discovery per https://review.openstack.org/#/c/145613/, then nova api talking to cinderclient wouldn't have to worry about this as long as novaclient was talking to nova api02:11
melwittversion discovery would be ideal. I didn't know if any progress has been done yet for that, other than the link you just showed02:11
melwittand yes, novaclient does talk directly to cinder api02:11
mriedemoh barf02:12
mriedemso...how hard would it be to change novaclient to talk to nova apis for volume thingies?02:12
mriedemor is there some ugly reason we can't just do that?02:12
melwittbecause as I found out, the volumes proxy in nova doesn't have the same features, volume types and snapshots I think02:13
mriedemso novaclient is basically duplicating cinderclient02:13
mriedemdouble barf02:13
melwittI tried to do it, but it would require augmenting our volumes proxy first02:13
melwittyeah :(02:13
mriedemcan we....deprecate volume CLIs in novaclient and say using cinderclient?02:13
mriedemshit, we don't create subnets in neutron via novaclient02:14
melwittyeah, I'm not sure how this one was allowed as an okay thing :P02:14
*** vilobhmm1 has quit IRC02:14
mriedemlet's blame Vek02:14
melwittheh02:15
mriedemOR02:16
mriedemomg02:16
mriedemlet's just import cinderclient into novaclient like an sdk and proxy the shell calls to cinderclient, that would be terribad02:17
mriedemeffectively making a mutated version of openstackclient02:17
openstackgerritYongli He proposed openstack/nova-specs: test cases for third party CI  https://review.openstack.org/14127002:18
*** vilobhmm1 has joined #openstack-nova02:18
melwittyeah, I was thinking similar02:19
melwittthe monster other openstackclient02:19
melwittnovaclient is my favorite client. don't want to use anything else02:20
mriedemit does images, volumes, the works02:20
melwitthaha the works, yes!02:21
*** vilobhmm1 has quit IRC02:21
* melwitt commutes02:23
openstackgerritJoe Gordon proposed openstack/nova: Update docs layout  https://review.openstack.org/18322102:25
openstackgerritJoe Gordon proposed openstack/nova: Prune development.environment doc  https://review.openstack.org/18288402:25
openstackgerritJoe Gordon proposed openstack/nova: Remove jenkins, launchpad and gerrit docs  https://review.openstack.org/18334202:25
openstackgerritJoe Gordon proposed openstack/nova: Make blueprints doc a reference for nova blueprints  https://review.openstack.org/18334302:25
*** leifz has quit IRC02:27
*** ZZelle has joined #openstack-nova02:28
*** artom has quit IRC02:29
*** leifz has joined #openstack-nova02:29
openstackgerritChangBo Guo(gcb) proposed openstack/nova: Recover instance's attributes host and node when evacuate failure  https://review.openstack.org/14337202:29
mriedemhttp://lists.openstack.org/pipermail/openstack-dev/2015-May/064210.html02:30
*** unicell has quit IRC02:33
*** dave-mccowan has quit IRC02:37
*** baoli has joined #openstack-nova02:40
*** otter768 has joined #openstack-nova02:43
*** dave-mccowan has joined #openstack-nova02:44
*** okrieg has quit IRC02:46
*** otter768 has quit IRC02:47
*** okrieg has joined #openstack-nova02:50
*** yamahata has quit IRC02:50
*** okrieg has quit IRC02:52
*** mitz has quit IRC02:52
*** READ10 has quit IRC02:54
openstackgerritMonty Taylor proposed openstack/python-novaclient: Use clouds.yaml for functional test credentials  https://review.openstack.org/18310602:55
openstackgerritEli Qiao proposed openstack/nova: Neutron: Check port binding status  https://review.openstack.org/15611203:00
*** mtanino has quit IRC03:00
*** comstud has quit IRC03:02
*** browne has joined #openstack-nova03:04
*** krtaylor has joined #openstack-nova03:05
openstackgerritzhangtralon proposed openstack/nova: add warning when using force_hosts in the ChanceScheduler  https://review.openstack.org/17195103:11
*** dsanders has joined #openstack-nova03:11
*** dims has quit IRC03:12
*** vilobhmm1 has joined #openstack-nova03:13
*** pczesno has quit IRC03:13
*** pczesno has joined #openstack-nova03:13
openstackgerritzhangtralon proposed openstack/nova: Ironic:check the configuration item api_max_retries  https://review.openstack.org/12331903:16
*** achanda has joined #openstack-nova03:21
*** achanda has quit IRC03:27
openstackgerritMonty Taylor proposed openstack/python-novaclient: Move from in-tree clouds.yaml to oslotest  https://review.openstack.org/18338803:27
*** baoli has quit IRC03:29
*** garyk has joined #openstack-nova03:29
*** baoli has joined #openstack-nova03:29
openstackgerritMonty Taylor proposed openstack/python-novaclient: Move from in-tree clouds.yaml to oslotest  https://review.openstack.org/18338803:30
*** sdake has joined #openstack-nova03:33
*** alexpilotti has quit IRC03:35
*** sdake_ has quit IRC03:36
*** baoli has quit IRC03:37
openstackgerritRui Chen proposed openstack/nova-specs: CreateVM supports subnet specified  https://review.openstack.org/16963803:37
*** sdake_ has joined #openstack-nova03:38
openstackgerritMonty Taylor proposed openstack/python-novaclient: Move from in-tree clouds.yaml to oslotest  https://review.openstack.org/18338803:39
*** sdake has quit IRC03:41
*** armax has joined #openstack-nova03:42
*** yamahata has joined #openstack-nova03:43
*** vilobhmm1 has quit IRC03:43
*** vilobhmm1 has joined #openstack-nova03:44
*** tobe4333 has joined #openstack-nova03:48
*** armax has quit IRC03:48
*** HenryG has quit IRC03:51
*** hightall has quit IRC03:52
*** Sukhdev has joined #openstack-nova03:55
*** tojuvone has joined #openstack-nova03:58
openstackgerritChris Friesen proposed openstack/nova: Fix race between resource audit and cpu pinning  https://review.openstack.org/18276604:06
*** slong has joined #openstack-nova04:09
*** coolsvap|afk is now known as coolsvap04:12
*** tobe4333 has quit IRC04:20
*** achanda has joined #openstack-nova04:22
*** Marga_ has quit IRC04:24
*** pradipta has joined #openstack-nova04:26
*** ijw has joined #openstack-nova04:31
*** vilobhmm1 has quit IRC04:34
*** ijw_ has joined #openstack-nova04:35
*** sdake_ has quit IRC04:35
*** vishwana_ has joined #openstack-nova04:35
*** ijw has quit IRC04:36
*** vishwanathj has quit IRC04:38
*** vilobhmm1 has joined #openstack-nova04:38
*** otter768 has joined #openstack-nova04:44
*** otter768 has quit IRC04:49
openstackgerritQin Zhao proposed openstack/nova: Refactor show_port() in neutron api  https://review.openstack.org/13502004:50
openstackgerritFei Long Wang proposed openstack/nova: Fix nova backup for volume-backed instance  https://review.openstack.org/16449404:52
*** Mike_D_laptop has joined #openstack-nova04:57
*** hightall has joined #openstack-nova04:58
*** Marga_ has joined #openstack-nova05:00
*** ildikov has quit IRC05:05
*** markvoelker has joined #openstack-nova05:11
*** irenab has joined #openstack-nova05:12
*** Mike_D_laptop has quit IRC05:12
*** armax has joined #openstack-nova05:13
*** armax has quit IRC05:15
*** ijw_ has quit IRC05:16
*** ijw has joined #openstack-nova05:16
*** ildikov has joined #openstack-nova05:18
*** bkopilov has quit IRC05:20
*** ZZelle has quit IRC05:20
*** bkopilov has joined #openstack-nova05:22
*** Mike_D_laptop has joined #openstack-nova05:27
*** irenab has quit IRC05:27
*** dulek has joined #openstack-nova05:29
*** gokrokve has joined #openstack-nova05:31
*** Mike_D_laptop has quit IRC05:32
*** HenryG has joined #openstack-nova05:38
*** pece has joined #openstack-nova05:42
*** armax has joined #openstack-nova05:42
*** mestery has joined #openstack-nova05:44
*** afazekas has joined #openstack-nova05:46
*** markvoelker has quit IRC05:47
*** sdake has joined #openstack-nova05:52
*** mestery has quit IRC05:53
*** mestery has joined #openstack-nova05:53
*** sdake has quit IRC05:53
*** sdake has joined #openstack-nova05:54
*** unicell has joined #openstack-nova05:54
*** Sukhdev has quit IRC05:55
*** gokrokve has quit IRC05:56
*** gokrokve_ has joined #openstack-nova05:56
*** sdake has quit IRC05:58
*** ddutta has joined #openstack-nova05:59
*** Jeffrey4l has quit IRC06:01
*** gokrokve_ has quit IRC06:01
*** diegows has quit IRC06:01
*** nihilifer has joined #openstack-nova06:03
*** vilobhmm1 has quit IRC06:04
*** Jeffrey4l has joined #openstack-nova06:04
*** Marga_ has quit IRC06:12
*** slong has quit IRC06:12
openstackgerritYongli He proposed openstack/nova: resource tracker style pci resource management  https://review.openstack.org/13132106:12
*** mestery has quit IRC06:13
*** sdake has joined #openstack-nova06:16
*** ajo has joined #openstack-nova06:21
*** lpetrut has joined #openstack-nova06:21
*** achanda has quit IRC06:23
*** sudipto has joined #openstack-nova06:24
openstackgerritOpenStack Proposal Bot proposed openstack/nova: Imported Translations from Transifex  https://review.openstack.org/17963806:25
*** Murali has joined #openstack-nova06:29
*** vilobhmm1 has joined #openstack-nova06:31
*** pradipta has quit IRC06:32
*** pczesno has quit IRC06:38
*** pczesno has joined #openstack-nova06:39
*** Kennan2 has joined #openstack-nova06:41
*** Kennan has quit IRC06:41
*** Marga_ has joined #openstack-nova06:43
*** unicell1 has joined #openstack-nova06:43
*** armax has quit IRC06:43
*** otter768 has joined #openstack-nova06:45
*** unicell has quit IRC06:45
*** Marga_ has quit IRC06:48
*** Sukhdev has joined #openstack-nova06:49
*** otter768 has quit IRC06:50
*** haypo has joined #openstack-nova06:50
*** vilobhmm11 has joined #openstack-nova06:52
*** vilobhmm1 has quit IRC06:54
openstackgerritRoman Bogorodskiy proposed openstack/nova-specs: User-controlled SR-IOV ports allocation  https://review.openstack.org/18224206:59
*** cfriesen has quit IRC07:00
*** jlanoux has joined #openstack-nova07:03
*** ildikov has quit IRC07:04
*** sdake_ has joined #openstack-nova07:05
*** ZZelle_ has quit IRC07:08
*** sdake has quit IRC07:08
*** ZZelle_ has joined #openstack-nova07:09
*** aluria has joined #openstack-nova07:12
*** eglynn has joined #openstack-nova07:14
*** ildikov has joined #openstack-nova07:17
*** eglynn has quit IRC07:20
gilliardMorning Nova!07:26
gilliardFriday is it?07:26
*** gilliard is now known as gilllliard07:26
eliqiao1gilliard: yeah :)07:27
*** jcoufal has joined #openstack-nova07:27
*** zhangtralon has quit IRC07:28
gilllliardHello qli07:28
gilllliarderr... Hello eliqiao107:29
*** pask81 has joined #openstack-nova07:31
*** jistr has joined #openstack-nova07:33
*** pask81 has quit IRC07:36
*** vilobhmm11 has quit IRC07:37
*** bfic has joined #openstack-nova07:42
*** Marga_ has joined #openstack-nova07:44
*** tobe4333 has joined #openstack-nova07:44
*** pixelb has joined #openstack-nova07:47
*** Marga_ has quit IRC07:49
*** dsanders has quit IRC07:51
*** pask81 has joined #openstack-nova07:51
*** redbeard has joined #openstack-nova07:56
*** ociuhandu has joined #openstack-nova08:00
*** dsanders has joined #openstack-nova08:01
*** mpaolino has joined #openstack-nova08:02
*** Kennan2 is now known as Kennan08:04
*** Jeffrey4l has quit IRC08:05
*** tobe4333 has quit IRC08:07
*** lpetrut has quit IRC08:07
*** apuimedo has joined #openstack-nova08:10
*** derekh has joined #openstack-nova08:13
*** Sukhdev has quit IRC08:15
*** browne has quit IRC08:15
*** browne has joined #openstack-nova08:15
*** bjornar has joined #openstack-nova08:16
*** pczesno has quit IRC08:17
openstackgerritQin Zhao proposed openstack/nova: Make resize api of compute manager to send flavor object  https://review.openstack.org/16186008:18
*** pczesno has joined #openstack-nova08:18
*** eglynn has joined #openstack-nova08:19
*** dsanders has quit IRC08:22
openstackgerritRui Chen proposed openstack/nova-specs: CreateVM supports subnet specified  https://review.openstack.org/16963808:23
*** pask81 has quit IRC08:25
*** pask81 has joined #openstack-nova08:25
*** yamahata has quit IRC08:33
*** Nikolay_St has joined #openstack-nova08:34
*** tojuvone has quit IRC08:37
openstackgerritRui Chen proposed openstack/nova: CreateVM supports subnet specified  https://review.openstack.org/18296908:39
openstackgerritChangBo Guo(gcb) proposed openstack/nova: Recover instance's attributes host and node when evacuate failure  https://review.openstack.org/14337208:42
*** Marga_ has joined #openstack-nova08:44
*** otter768 has joined #openstack-nova08:46
*** Marga_ has quit IRC08:50
*** otter768 has quit IRC08:50
*** afazekas has quit IRC08:52
johnthetubaguymriedem: with the spec merged, I approved the blueprint, and so removed the requested -2 https://blueprints.launchpad.net/nova/+spec/db2-database08:53
johnthetubaguymriedem: any chance you could clear out the whiteboard, so its super clear what patches are left to review please? https://blueprints.launchpad.net/nova/+spec/db2-database08:53
johnthetubaguyalaski: I couldn't find a blueprint for this one, so I created it, let me know if thats correct: https://blueprints.launchpad.net/nova/+spec/persist-request-spec08:56
BobBallmriedem: Interesting question... It's not hitting a large number of tests, but could easily be a libvirt+xen specific issue; I'll ask anthonyper if he's got any thoughts on it.08:57
*** lucasagomes has joined #openstack-nova08:57
BobBallmriedem: It's quite possible that https://review.openstack.org/#/c/182680/ is needed to track this one08:57
*** danpb has joined #openstack-nova08:57
*** inara has quit IRC08:59
*** ijw has quit IRC09:00
openstackgerritJohn Garbutt proposed openstack/nova: Consolidate the APIs for getting consoles  https://review.openstack.org/14850909:01
*** ppal has joined #openstack-nova09:06
*** ildikov has quit IRC09:07
*** inara has joined #openstack-nova09:09
*** afazekas has joined #openstack-nova09:14
*** ildikov has joined #openstack-nova09:16
*** ildikov is now known as ildikov_afk09:16
*** gszasz has joined #openstack-nova09:20
*** ildikov_afk has quit IRC09:25
*** hightall has quit IRC09:30
*** RuiChen has quit IRC09:35
johnthetubaguydanpb: so bumping the min_version of libvirt and qemu, whats your take on that?09:37
*** ildikov_afk has joined #openstack-nova09:38
*** ddutta has quit IRC09:39
*** pixelb has quit IRC09:39
*** pixelb has joined #openstack-nova09:41
*** lpetrut has joined #openstack-nova09:42
danpbjohnthetubaguy: best if i reply on list, as i think this is quite a complicated issue really09:44
johnthetubaguydanpb: OK09:44
johnthetubaguydanpb: we would drop quite a lot of (mostly untested) code right, if we bump the min_version?09:45
*** dims has joined #openstack-nova09:45
*** Marga_ has joined #openstack-nova09:45
eliqiao1 johnthetubaguy: hi GM john, can you help to review again  https://review.openstack.org/#/c/177671/ ? it's quiet similar with https://review.openstack.org/#/c/16583809:45
*** alexpilotti has joined #openstack-nova09:46
*** Marga_ has quit IRC09:50
johnthetubaguyeliqiao1: agreed they are similar, would it be possible to work with sahid in implementing https://review.openstack.org/#/c/165838 ? they seem to be both hitting on the same issue, its just https://review.openstack.org/#/c/165838 was proposed first, and I think that was the closest to getting merged when I last looked at it09:53
*** akosiaris has left #openstack-nova09:53
johnthetubaguyeliqiao1: I fear I will not be able to review your new revision before the summit at this point, there are many specs I still need to get to for the first time, who have been waiting a little while now09:53
*** ppal_ has joined #openstack-nova09:54
peceif I cycle power (off/on) on virtual instance is nova-scheduler involved then? (particulary: scheduler filters)?09:54
openstackgerritFeng Xi Yan proposed openstack/nova: Make VMWare Driver support third party distributed switch.  https://review.openstack.org/15200609:54
johnthetubaguypece: no its a direct message to the compute node09:54
eliqiao1johnthetubaguy: thanks, I am agree to work them together , I don't notice it before you point it to me.. i'ts little different, I don't want to deprate memcache at this moment..09:54
eliqiao1johnthetubaguy:  sure, thanks.09:55
*** ppal__ has joined #openstack-nova09:55
*** dulek_ has joined #openstack-nova09:55
*** ppal_ has quit IRC09:55
johnthetubaguyeliqiao1: yeah, I think we need to get you both together, agree a direction, and work out how to split up the work09:55
*** pask81 has quit IRC09:56
johnthetubaguyeliqiao1: maybe you could start an ML thread about the two specs, and agree some common goals, etc?09:56
eliqiao1johnthetubaguy:  thanks.. but I don't think there will be too much work to do.. I will put my comments on that spec too.09:56
johnthetubaguyeliqiao1: awesome, that works09:56
eliqiao1johnthetubaguy:  okay , thanks for the advice , I will do that :)09:56
johnthetubaguyeliqiao1: well, if we can get agreement on the spec, that might be easier than going to the ML09:57
*** ppal__ has quit IRC09:57
*** ppal has quit IRC09:57
eliqiao1 johnthetubaguy:  okay , I will put my comments on #165838 first ..09:57
pecejohnthetubaguy: thx... so filters are called only when instance is created and provisioned, right? I'm looking for simple way to control if user can start particular instance.09:58
*** pask81 has joined #openstack-nova09:58
*** dulek has quit IRC09:58
johnthetubaguyeliqiao1: thank you, do catch me after the summit if there is no agreement, and I can try get folks together09:58
johnthetubaguypece: there is a lock instance call that means an admin stops a user doing any actions on an instance?09:58
eliqiao1 johnthetubaguy: nice, thanks09:58
johnthetubaguypece: the scheduler is basically involved whenever you need to pick a host, resize, migrate, live-migrate, create, etc10:00
pecejohnthetubaguy: hmm, is this call available in Juno? Is it an API call?10:00
*** pask811 has joined #openstack-nova10:00
johnthetubaguypece: unsure to be honest, I think the admin lock was quite recent10:00
*** IanGovett has joined #openstack-nova10:00
*** dulek_ has quit IRC10:01
pecejohnthetubaguy: thank you, I'll look up on admin lock10:01
johnthetubaguypece: seems like the general lock is there in juno: https://github.com/openstack/nova/blob/stable/juno/nova/api/openstack/compute/contrib/admin_actions.py#L20910:02
johnthetubaguypece: I think with the policy you could just restrict lock and unlock to some group of "administrators" although I don't think thats the default policy, hope that helps10:03
tdurakovhi, is it possible in nova to resize down server?10:04
johnthetubaguypece: looks like juno has everything you need: https://github.com/openstack/nova/blob/stable/juno/nova/compute/api.py#L286610:04
johnthetubaguytdurakov: it depends on what driver you use10:04
johnthetubaguytdurakov: I think only XenAPI supports that at this point10:04
*** pask81 has quit IRC10:04
johnthetubaguytdurakov: I wouldn't recommend trying to implement resize down though, its really not a very reliable operation, because of whats required with changing the size of the root file system10:05
tdurakovjohnthetubaguy: I've noticed there is no notification for op, when driver doesn't allow to resize down server:http://paste.openstack.org/show/223636/ http://paste.openstack.org/show/223635/10:05
johnthetubaguytdurakov: now if you don't change the size of the root disk, I suspect all drivers support that10:06
johnthetubaguytdurakov: notifications are totally untested and largely unmaintained at this point, although I hope to fix that10:06
tdurakovjohnthetubaguy, ok thx. I'll file bug for that and try to fix it10:07
johnthetubaguytdurakov: that would be great, appreciated10:07
johnthetubaguytdurakov: I have a summit session planned to try and work out the best way forward for "error handling"10:08
tdurakovjohnthetubaguy, could you provide sessio results then?:) It would be interesting for me10:09
johnthetubaguytdurakov: I hope to write everything up, starting to do something here: https://review.openstack.org/#/c/178635/10:09
tdurakovjohnthetubaguy: as i understood in my case such error couldn't be held by design10:11
tdurakovthere is cast rpc call10:11
johnthetubaguytdurakov: so, you could have a policy for resize down of a disk, and enforce that in the API, so the API can return an error in that case?10:13
johnthetubaguytdurakov: we could add a call to a compute node to check if resize is possible, we do something like that for live-migrate10:14
tdurakovjohnthetubaguy: good point, let me check it:)10:14
*** pask811 has quit IRC10:15
johnthetubaguytdurakov: it might be worth a nova-spec to get agreement on the approach, but I would try fix the bug first, as that should be less paperwork10:15
*** pask81 has joined #openstack-nova10:15
*** pask81 has quit IRC10:15
tdurakovbtw, there is more confusing point: http://paste.openstack.org/show/223659/10:15
tdurakovjohnthetubaguy: compute logs give me the same picture at that moment, op has illusion that everything is OK10:17
*** takedakn has joined #openstack-nova10:17
johnthetubaguytdurakov: yeah, thats a client issue I think, the API will show the instance as Active rather than Resize_confirm, and as such, the resize didn't happen10:18
johnthetubaguytdurakov: its an issue with have with the API right now, we have plans for a "tasks" API, where it would be much clearer that your API request "failed", but the instance is not in an "Error" state10:18
johnthetubaguytdurakov: if you look at instance_actions, it should list the failed resize, i think10:19
johnthetubaguytdurakov: a pyhton-novaclient bug + patch around this would be awesome10:19
johnthetubaguy(and docs I guess)10:19
tdurakov2 bug: 1 in nova, 1 in client and some fix for now10:20
openstackbug 1 in Ubuntu Malaysia LoCo Team "Microsoft has a majority market share" [Critical,In progress] https://launchpad.net/bugs/1 - Assigned to MFauzilkamil Zainuddin (apogee)10:20
*** aix has joined #openstack-nova10:21
*** tobe4333 has joined #openstack-nova10:27
*** VW_ has quit IRC10:28
lpetrutHi guys, is there something wrong w/ the Xen CI? It keeps failing but I can't access the logs10:28
BobBalllpetrut: Example change?10:28
lpetruthttps://review.openstack.org/#/c/127159/10:28
BobBalllpetrut: There was a failure last night that also hit the -infra CI with git.openstack.org failing to clone10:28
*** VW has joined #openstack-nova10:28
BobBallYup - times match10:29
lpetrutgreat, thanks! Is it fixed at the moment? I was thinking about a recheck10:29
lpetrutalso, it would be great if the logs would be accessible10:29
BobBallYes, please do a "recheck xen" and it'll requeue it10:30
BobBall(or at least it should! Would be added to http://zuul.openstack.xenproject.org/)10:31
*** VW has quit IRC10:31
*** VW has joined #openstack-nova10:32
*** takedakn has quit IRC10:32
*** takedakn has joined #openstack-nova10:32
*** ajo has quit IRC10:33
*** bfic has quit IRC10:34
*** VW has quit IRC10:37
*** VW has joined #openstack-nova10:38
BobBalllpetrut: Actually it hopefully wouldn't requeue it... Another change we made yesterday is to ignore changes that can't affect libvirt+xen! I've manually requeued.10:39
*** VW has quit IRC10:39
*** ociuhandu has quit IRC10:39
*** VW has joined #openstack-nova10:40
*** VW has quit IRC10:42
*** VW has joined #openstack-nova10:43
*** Marga_ has joined #openstack-nova10:45
*** VW has quit IRC10:45
*** VW has joined #openstack-nova10:46
sdaguetonyb: if you push a patch up to flip on utopic and vivid for devstack, I'll push it through.10:46
sdagueper your email, them not being listed is mostly, no one yet cared10:46
*** Marga__ has joined #openstack-nova10:47
*** otter768 has joined #openstack-nova10:47
*** Marga_ has quit IRC10:47
*** Marga__ has quit IRC10:48
*** VW has quit IRC10:48
*** Marga_ has joined #openstack-nova10:48
*** VW has joined #openstack-nova10:49
*** VW has quit IRC10:49
*** VW has joined #openstack-nova10:49
*** otter768 has quit IRC10:51
*** Marga_ has quit IRC10:53
openstackgerritGorka Eguileor proposed openstack/nova: Preserve usage and reservations on quota deletion  https://review.openstack.org/16342310:55
*** dulek_ has joined #openstack-nova10:57
*** mpaolino has quit IRC10:59
*** Marga_ has joined #openstack-nova11:05
openstackgerritMatthew Booth proposed openstack/nova: Remove unused variables from images api  https://review.openstack.org/18347511:07
*** tobe4333 has quit IRC11:07
danpbwho owns XenProject CI   ?  it no longer tells you what the syntax is for a xen-only recheck when it fails, and both the wiki links it provides are wrong11:09
BobBallThe links are wrong?  Sorry - I didn't check them... Will correct that.11:10
danpbBobBall: in https://wiki.openstack.org/wiki/ThirdPartySystems/ there is a bogus trailing slash that needs removing11:10
BobBalld'oh!11:10
danpbBobBall: and https://wiki.openstack.org/wiki/GerritJenkinsGit#Test_Failures has been deleted :-(11:10
* BobBall was just using the standard template from os-ext-testing :/11:11
*** ociuhandu has joined #openstack-nova11:11
BobBallWill update11:11
danpbBobBall:   does  "xenserver: recheck" still work as the right syntax ?11:11
BobBallxenserver: recheck is the XenServer ci, which is (joy of joys) different to the xenproject CI (which checks libvirt+xen)11:12
*** Longgeek has quit IRC11:12
BobBallSyntax for XenProject ci recheck is currently "recheck xen", again just following the conventions from the default job defintions but I should have updated that to xen: recheck and will do that at the same time.11:13
danpbah, ok, i should have noticed they were different :-)11:13
danpbcould you make the XenProject CI  message included  "recheck xen" in its text11:13
BobBallI'm a bit surprised actually - I was certain I'd added it to the third party CI systems but it's not on the wiki page!11:14
BobBallYup11:14
*** ociuhandu_ has joined #openstack-nova11:19
*** ociuhandu has quit IRC11:19
*** ociuhandu_ is now known as ociuhandu11:19
*** zhenguo has quit IRC11:21
*** VW has quit IRC11:24
*** VW has joined #openstack-nova11:25
kashyapdanpb: Thanks for that detailed response on that min libvirt versions thread.  Specifically, I appreciate the ratioanle behind why one need to be conservative w.r.t hypervisor platform while it's okay to move relatively faster w/ the higher layer (OpenStack).11:27
BobBalldanpb: I hope that's fixed now... :)11:35
*** bfic has joined #openstack-nova11:36
gilllliardNova API meeing in 22 minutes?11:38
*** alexpilotti has quit IRC11:42
danpbBobBall: cool, thanks for taking care of it so quickly11:43
*** Longgeek has joined #openstack-nova11:45
*** VW has quit IRC11:45
*** baoli has joined #openstack-nova11:45
*** VW has joined #openstack-nova11:45
tdurakovjohnthetubaguy: do you think we need spec for new rpc method that checks instance  could be resized?11:48
*** baoli_ has joined #openstack-nova11:51
johnthetubaguytdurakov: its borderline. its copying the pattern used for live-migrate (see can_live_migrate calls), as a bug fix, I think that could be OK11:52
tdurakovok, then I'll code patch first11:53
tdurakovjohnthetubaguy: thank you11:53
*** baoli has quit IRC11:54
openstackgerritDaniel Berrange proposed openstack/nova: libvirt: check min required qemu/libvirt versions on s390/s390x  https://review.openstack.org/18348611:55
*** Marga_ has quit IRC11:56
gilllliardNova API meeting in #openstack-meeting in 2 minutes. alex_xu oomichi eliqiao1 johnthetubaguy sdague11:58
*** VW has quit IRC11:58
sdaguegilllliard: thanks for the reminder, I'm dealing with packing / laundry for the trip right now, so won't be able to be there11:59
*** VW has joined #openstack-nova11:59
gilllliardit'll be a quiet on for sure :)11:59
*** dprince has joined #openstack-nova12:02
johnthetubaguydansmith: I attempted to capture your thoughts on mentoring and things in this wiki, added it to the agenda for the next nova meeting: https://wiki.openstack.org/wiki/Nova/Mentoring12:05
johnthetubaguysdague: I think some of these points were your ideas as well, possibly bent out of shape: https://wiki.openstack.org/wiki/Nova/Mentoring12:05
*** Marga_ has joined #openstack-nova12:07
gilllliard^^ I like this12:08
*** sudipto has quit IRC12:09
*** gilllliard is now known as gilliard_en_rout12:09
*** ajo has joined #openstack-nova12:09
johnthetubaguygilliard_en_rout: cool, feel free to send ideas and enhancements (via wiki edits) etc12:10
*** artom has joined #openstack-nova12:12
*** links has joined #openstack-nova12:13
alaskijohnthetubaguy: thanks for creating the bp.  I have a bad habit of not creating them until the spec is approved (and then forgetting)12:13
*** ajo has quit IRC12:13
johnthetubaguyalaski_eto: no worries12:14
*** lucasagomes is now known as lucas-hungry12:17
openstackgerritChangBo Guo(gcb) proposed openstack/nova: libvirt: use instance UUID with exception InstanceNotFound  https://review.openstack.org/18349212:17
sdaguejohnthetubaguy: ok, I'll try to take a look today. Warning though, I'm sort of in odds and ends wrap up mode today as I leave tomorrow afternoon, and a bunch of house tasks have to be accomplished before I get on a plane12:18
*** coolsvap is now known as coolsvap|afk12:23
*** afazekas has quit IRC12:23
*** rfolco has joined #openstack-nova12:24
*** IanGovett has quit IRC12:26
*** VW_ has joined #openstack-nova12:28
*** Marga_ has quit IRC12:30
*** VW has quit IRC12:30
*** mwagner_lap has joined #openstack-nova12:32
*** IanGovett has joined #openstack-nova12:32
*** VW_ has quit IRC12:33
*** gabriel-bezerra has quit IRC12:38
*** gabriel-bezerra has joined #openstack-nova12:40
*** okrieg has joined #openstack-nova12:41
*** aysyanne has joined #openstack-nova12:42
*** dboik has joined #openstack-nova12:42
*** mpaolino has joined #openstack-nova12:43
*** tojuvone has joined #openstack-nova12:44
*** dboik has quit IRC12:47
openstackgerritChangBo Guo(gcb) proposed openstack/nova: libvirt: use instance UUID with exception InstanceNotFound  https://review.openstack.org/18349212:47
*** otter768 has joined #openstack-nova12:48
*** gabriel-bezerra has quit IRC12:48
*** annegentle has joined #openstack-nova12:48
*** derekh_ has joined #openstack-nova12:49
baoli_garyk: ping, RE: https://review.openstack.org/#/c/183061/12:50
*** claudiub has joined #openstack-nova12:51
*** otter768 has quit IRC12:52
*** dulek_ is now known as dulek12:52
*** derekh has quit IRC12:52
johnthetubaguysdague: totally, similar here really12:56
*** links has quit IRC12:57
*** sdake has joined #openstack-nova13:00
*** gabriel-bezerra has joined #openstack-nova13:01
*** sdake_ has quit IRC13:02
*** sdake_ has joined #openstack-nova13:03
*** cbader has joined #openstack-nova13:05
*** sdake_ has quit IRC13:07
*** bfic has quit IRC13:07
*** sdake has quit IRC13:07
openstackgerritjichenjc proposed openstack/nova: Create Nova Scheduler Num Instances Weighter  https://review.openstack.org/13052613:09
*** sdake has joined #openstack-nova13:11
*** cl__ has quit IRC13:15
*** dboik has joined #openstack-nova13:16
*** dboik_ has joined #openstack-nova13:17
*** jecarey has joined #openstack-nova13:18
*** dboik has quit IRC13:21
*** erlon has joined #openstack-nova13:22
*** kylek3h has quit IRC13:22
openstackgerritDaniel Berrange proposed openstack/nova: libvirt: mark NUMA huge page mappings as shared access  https://review.openstack.org/17636013:23
openstackgerritjichenjc proposed openstack/nova: Remove policy check security_group_default_rule  https://review.openstack.org/18146713:23
*** sdake has quit IRC13:23
*** _gryf is now known as _gryf|afk13:24
*** busterswt has joined #openstack-nova13:28
*** lucas-hungry is now known as lucasagomes13:31
openstackgerritjichenjc proposed openstack/nova: Don't update RT status when set instance to ERROR  https://review.openstack.org/18142613:33
*** claudiub has quit IRC13:34
*** andrewbogott_afk is now known as andrewbogott13:34
*** sdake has joined #openstack-nova13:35
*** garyk has quit IRC13:36
openstackgerritPetrut Lucian proposed openstack/nova: Hyper-V: Nova serial console access support  https://review.openstack.org/14500413:39
openstackgerritPetrut Lucian proposed openstack/nova: Hyper-V: restart serial console workers after instance power change  https://review.openstack.org/16308013:39
*** takedakn has quit IRC13:42
*** sdake has quit IRC13:42
*** aix has quit IRC13:44
*** gholler has joined #openstack-nova13:44
*** burt has joined #openstack-nova13:45
*** gszasz has quit IRC13:50
*** apuimedo has quit IRC13:53
*** jlanoux has quit IRC13:54
openstackgerritjichenjc proposed openstack/nova: Add missing @require_context  https://review.openstack.org/17645413:55
*** dansmith is now known as superdan13:56
*** kylek3h has joined #openstack-nova13:57
*** hightall has joined #openstack-nova13:58
*** annegentle has quit IRC13:59
*** takedakn has joined #openstack-nova13:59
*** BadCub-Hiding is now known as BadCub13:59
mriedemdanpb: so....i'm getting the feeling you don't want to bump the min version of libvirt? :P13:59
danpblol14:00
danpbwell we can bump it a little bit given some of the eol ubuntu versions14:01
danpbor rather add a deprecation warning to notify users of our intent to bump it in Mxx14:01
*** hightall has quit IRC14:02
mriedemyeah, i was thinking of the same last night, i.e. in init_host check if the version < 0.10.2 and warn if so, then bump to that in M14:02
danpbshame we didn't think of this a few weeks back when we could have warned in kilo14:02
mriedemand g-r has libvirt-python >= 1.2.5 so we could add that to test-requirements.txt to remove that workaround in host.py14:02
* danpb doesn't really understand how g-r maps / relates to project local requirements/text-requirements14:02
andreykurilin1jogo: hi! can you re-review my change( https://review.openstack.org/#/c/169292/ ) based on process of porting nova to python 3?14:03
mriedemi know i removed libvirt-python from test-requirements.txt but that was because it was being used in tests rather than fakelibvirt, i don't think that's an issue anymore since the tests explicitly import fakelibvirt now14:03
mriedemdanpb: since nova is registered with g-r, when we add libvirt-python to nova it has to be the same version as in g-r14:03
danpbi thought we removed it from test-requirements.txt so that people testing vmware didn't need to have libvirt-pytho ninstalled14:03
mriedemand if we update the version in g-r that will sync to nova14:03
mriedemdanpb: no, it was b/c it was corrupting the tests if it was installed14:04
danpbthe fact that tests use fakelibvirt was just the enabler to made it possible to remove it14:04
mriedemb/c of a conditional import in the tests14:04
mriedemyeah, but test-requirements is now also kind of a place to add optional dependencies based on which driver you're using, e.g. we moved oslo.vmware to test-requirements14:04
mriedemb/c you only need that if you're using the vmware driver14:04
danpbok14:04
mriedemso, i'll add the warning for < 0.10.2 and the move to that in M,14:05
mriedemand add libvirt-python to test-requirements to remove that workaround in host.py14:05
danpbok14:06
danpbso this would drop support for debian wheezy - not a problem as it has wheezy-backports available14:07
danpband suse 12.2, which i think is ok as there's several newer suse versions avail14:07
*** mpaolino has quit IRC14:08
*** sdake has joined #openstack-nova14:08
*** salv-orlando has joined #openstack-nova14:08
openstackgerritjichenjc proposed openstack/nova: Add common function for v2.1 API flavor_get.  https://review.openstack.org/13656814:09
openstackgerritjichenjc proposed openstack/nova: Add common function for v2 API flavor_get.  https://review.openstack.org/13656714:09
*** archers has joined #openstack-nova14:09
mriedemkashyap: btw, we already have libvirt 1.2.2 in the gate in the trusty nodes14:10
mriedemthat's why i went with that version as the min14:10
kashyapmriedem: Oh, yeah now that you mention, I remember seeing this somewhere :-)14:10
*** signed8bit has joined #openstack-nova14:11
*** nihilifer has quit IRC14:11
*** DziQs has joined #openstack-nova14:12
*** DziQs has left #openstack-nova14:12
*** Murali_ has joined #openstack-nova14:14
mriedemalex_xu: if i change my fixed-ips microversion change to depend on https://review.openstack.org/#/c/176177/ does that also mean i'm signing up to merge the v2 and v3 fixed ips api samples tests in my change?14:16
*** dims is now known as dimsum__14:16
*** sdake has quit IRC14:16
*** Murali has quit IRC14:16
*** Murali_ is now known as Murali14:16
*** amotoki has joined #openstack-nova14:16
alex_xumriedem: emm....yes, the merge sample tests work didn't process the microversion case. https://review.openstack.org/#/c/176177/ add related code.14:18
*** eharney has joined #openstack-nova14:18
mriedemalex_xu: i guess it's unclear to me why i need to depend on that change?14:19
mriedemthe tests are running in my change as far as i can tell14:19
alex_xumriedem: yes, you are right, but _api_version = 'v3' looks wrong, we should use 'v2.1'14:20
*** jcoufal has quit IRC14:21
alex_xumriedem: but let me check again, I'm ok without dependence14:21
*** cfriesen has joined #openstack-nova14:21
alex_xumriedem: If your patch merge first, let's change it again if https://review.openstack.org/#/c/176177/ get merged14:22
*** garyk has joined #openstack-nova14:22
*** aix has joined #openstack-nova14:22
garykbaoli_: hi, i'll look at it now14:25
garykbaoli_: ok. i understand. i think that you need to rebase the unit tests are broken.14:26
*** mpaolino has joined #openstack-nova14:27
*** sdake has joined #openstack-nova14:28
*** zsais has joined #openstack-nova14:28
mriedemalex_xu: can you drop your -1 then?14:30
*** vladikr has joined #openstack-nova14:30
*** annegentle has joined #openstack-nova14:31
alex_xumriedem: yea14:31
*** ZZelle has joined #openstack-nova14:31
alex_xumriedem: it's done, sorry for that14:33
*** dave-mccowan has quit IRC14:34
openstackgerritBaodong (Robert) Li proposed openstack/nova: Split instance event/tag correctly  https://review.openstack.org/18306114:35
*** achanda has joined #openstack-nova14:36
*** mpaolino has quit IRC14:37
*** archers has quit IRC14:38
*** cknight has joined #openstack-nova14:40
*** nelsnelson has joined #openstack-nova14:40
*** achanda has quit IRC14:43
*** mragupat has joined #openstack-nova14:45
*** salv-orlando has quit IRC14:46
*** mragupat has quit IRC14:46
*** dave-mccowan has joined #openstack-nova14:48
*** otter768 has joined #openstack-nova14:49
*** armax has joined #openstack-nova14:49
*** alexpilotti has joined #openstack-nova14:49
*** takedakn has quit IRC14:50
*** otter768 has quit IRC14:53
*** lpetrut has quit IRC14:53
garykbaoli_: thanks!14:55
baoli_garyk, thanks for your review14:56
*** mpaolino has joined #openstack-nova14:56
*** nihilifer has joined #openstack-nova14:57
*** ZZelle has quit IRC14:57
*** mikal has joined #openstack-nova14:57
*** ZZelle has joined #openstack-nova14:57
garykmriedem: any idea how i recheck xen?14:58
mriedemalex_xu: thanks14:59
danpbgaryk: (12:13:04) BobBall: Syntax for XenProject ci recheck is currently "recheck xen", again just following the conventions from the default job defintions but I should have updated that to xen: recheck and will do that at the same time.14:59
mriedemgaryk: i think ti's busted anyway14:59
danpbassuming you mean the XenProject rather than XenServer14:59
mriedemi mentioned that late last night to BobBall15:00
garykdanpb: yes, correct the xen ci15:00
mriedemi think i'm going to just disable it15:00
alex_xumriedem: np :), another choice is we didn't pass 'reserved' into subs parameter, just hard-code resevered is false in sample template file.15:01
alex_xumriedem: I want to verify that, but I need rebuild tox env, it's pretty slow....15:01
mriedemgaryk: this is what i've been seeing when xenproject ci fails15:01
mriedem2015-05-14 22:03:33.163 14933 TRACE nova.compute.manager [instance: 095ab848-b7d0-4785-b78d-9e6976c80fbb]15:01
BobBallgaryk: Which change?  Just want to check the time it ran to confirm that it was before I fixed it today :)15:01
mriedemoops15:02
mriedemlibvirtError: internal error: libxenlight failed to restore domain 'instance-0000001b'15:02
mriedemBobBall: http://logs.openstack.xenproject.org/20/183220/1/check/dsvm-tempest-xen/dfbf491/15:02
*** claudiub has joined #openstack-nova15:02
garykBobBall: it was one i rebase yesterday morning - patches after that are ok. i did a recheck. will let you know if it still happens15:02
danpbmriedem: XenProject CI is mostly working for me - i rechecked on earlier and it succeeded15:02
BobBallmriedem: That libvirt error is an interesting one and we're trying to track it down; it needs https://review.openstack.org/#/c/183461/ so we can collect the Xen logs15:02
BobBalldanpb: I had to manually trigger yours because I'd screwed up the syntax :/15:03
danpbthanks :-)15:03
BobBallmriedem: Thought to be an issue with nics (libxl: error: libxl_create.c:1226:domcreate_attach_vtpms: unable to add nic devices) - but we haven't figured it out yet15:03
*** sdake_ has joined #openstack-nova15:03
*** nelsnels_ has joined #openstack-nova15:04
BobBallgaryk: http://zuul.openstack.xenproject.org/ your recheck did indeed get queued! :)15:04
*** nelsnelson has quit IRC15:04
garykBobBall: tx15:05
*** armax has quit IRC15:05
openstackgerritAndrey Kurilin proposed openstack/python-novaclient: Remove redundant check for version of `requests`  https://review.openstack.org/18355915:07
*** sdake has quit IRC15:08
*** gus has quit IRC15:08
*** zzzeek has joined #openstack-nova15:09
*** gus has joined #openstack-nova15:09
*** kmartin has joined #openstack-nova15:12
*** jistr is now known as jistr|mtgh15:15
*** jistr|mtgh is now known as jistr|mtg15:15
*** nelsnels_ has quit IRC15:16
*** nelsnelson has joined #openstack-nova15:17
alex_xumriedem: "reserved": %(reserved)s works for me. api sample test support non-quoted replacement https://github.com/openstack/nova/blob/master/nova/tests/functional/api_samples_test_base.py#L4715:17
mriedemalex_xu: ok, i'll try that in a bit15:18
alex_xumriedem: ok, thanks15:18
*** claudiub has quit IRC15:20
alex_xumriedem: then I think I can +1, after rechecking no more question from me15:20
*** yaguang has joined #openstack-nova15:22
*** artom has quit IRC15:25
*** tsekiyama has joined #openstack-nova15:25
*** Nikolay_St has quit IRC15:25
*** irenab has joined #openstack-nova15:26
*** jistr|mtg is now known as jistr15:28
*** mtanino has joined #openstack-nova15:29
*** yaguang has quit IRC15:29
*** lpetrut has joined #openstack-nova15:30
*** yamahata has joined #openstack-nova15:30
*** Murali has quit IRC15:33
mikal.15:33
*** destiny has quit IRC15:34
*** VW has joined #openstack-nova15:36
*** tonytan4ever has joined #openstack-nova15:37
*** krtaylor has quit IRC15:39
*** dulek has quit IRC15:39
*** VW has quit IRC15:40
*** hemnafk is now known as hemna15:47
*** vilobhmm1 has joined #openstack-nova15:48
*** apuimedo has joined #openstack-nova15:49
*** browne has quit IRC15:49
*** vilobhmm11 has joined #openstack-nova15:49
*** krtaylor has joined #openstack-nova15:51
openstackgerritChris St. Pierre proposed openstack/nova: Allow blacklisting os-api extensions  https://review.openstack.org/16098615:51
*** vilobhmm1 has quit IRC15:52
*** dsanders has joined #openstack-nova15:53
openstackgerritMatt Riedemann proposed openstack/nova: libvirt: deprecate libvirt version usage < 0.10.2  https://review.openstack.org/18322015:53
mriedemdanpb: ^15:54
*** BobBall is now known as BobBall_AWOL15:56
*** tonytan4ever has quit IRC15:57
*** kmartin has quit IRC15:58
*** shakamunyi has quit IRC16:00
*** barra204 has quit IRC16:00
*** sdake_ has quit IRC16:00
*** Marga_ has joined #openstack-nova16:00
*** harlowja has joined #openstack-nova16:01
*** lucasagomes is now known as lucas-packing16:02
*** harlowja has quit IRC16:03
*** annegentle has quit IRC16:04
johnthetubaguymriedem: I do wonder if we can deprecate more on top of that as well…16:05
cfriesenso wrt the summit...are there any speakers who are so good that it doesn't matter what they're talking about?16:06
*** VW has joined #openstack-nova16:06
*** kmartin has joined #openstack-nova16:08
*** xyang1 has quit IRC16:08
*** xyang1 has joined #openstack-nova16:09
*** kmartin has quit IRC16:10
*** BadCub is now known as BadCub-Traveling16:12
mriedemcfriesen: that would be me16:12
mriedemjohnthetubaguy: such as?16:12
johnthetubaguymriedem: drafting an ML post16:12
cfriesenmriedem: :)16:12
*** arnaud____ has joined #openstack-nova16:12
*** armax has joined #openstack-nova16:13
* mriedem goes to lunch16:13
johnthetubaguymriedem: thinking about scheduling the higher version for deprecating after N maybe?16:13
mriedemjohnthetubaguy: as in, 0.10.2 is M and 1.2.2 is N?16:14
johnthetubaguymriedem: yeah, something like that16:14
mriedemidk, i'd ask danpb16:15
johnthetubaguymriedem: really I want us to warn on every version thats non tested somehow16:15
mriedemjohnthetubaguy: we can do that separately today16:15
johnthetubaguymriedem: yeah, the 1.2.2 might need to be 1.1.x or something given Fedora16:15
mriedemi.e. we could put a warning in there that if you're on libvirt < 1.2.2 it's not tested in the gate16:15
*** nihilifer has quit IRC16:15
johnthetubaguymriedem: yeah, and the other way too really, so its != 1.2.2 although thats a little too harsh maybe16:16
mriedemtoo strict16:16
mriedemimo16:16
*** unicell1 has quit IRC16:16
mriedemwe do have a fc21 job on the experimental queue that runs 1.2.916:16
mriedemso we do have a way of validating some things there16:16
mriedemand actually, it sounds like ubuntu trusty LTS cloud archive has a much newer libvirt we could be testing against too16:17
johnthetubaguymriedem: yeah, I kinda want to encourage an actually list, with third party testing, etc16:17
mriedem"The Ubuntu Cloud Archive for Kilo includes updates for Nova, Glance, Keystone, Neutron, Cinder, Horizon, Swift, Ceilometer and Heat; Ceph (0.94.1), RabbitMQ (3.4.2), QEMU (2.2), libvirt (1.2.12) and Open vSwitch (2.3.1) back-ports from 15.04 have also been provided."16:17
mriedemanyway, lunch for real this time16:18
*** mriedem is now known as mriedem_lunch16:18
*** claudiub has joined #openstack-nova16:18
danpbit think it is a bit over the top to warn on every single version that doesn't match int he gate16:20
danpbthat means essentially every user who isn't running the matching version of ubuntu is going to have their logs spammed with messages16:20
danpbwith the result that there will be an endless stream of bug reports about the warnings, and users will learn to ignore them16:20
danpbso when we need to warn about something important they then won't notice16:20
danpbi think the release notes are the appropriate place to call out the distinction between what is enabled in the code, vs what is formally CI tested16:22
johnthetubaguydanpb: or people start to get their system tested against master, so they are included in the "tested" list16:22
johnthetubaguydanpb: yeah, having a good way to do that in the release notes / docs is also important, we just use log messages because its in the code I guess16:23
danpbif people get their nova from a vendor then its reasonable that they can assume the vendor has run the tempest CI16:23
danpbif people get their nova from upstream directly, then the release notes can tell them to run tempest against their deployment platform to validate it16:24
*** claudiub has quit IRC16:24
danpbeither way those people aren't going to want their logs polluted with warnings every time nova starts16:24
* danpb wonders if vmware/xenserver do any version checking 16:25
*** derekh_ has quit IRC16:25
*** dguitarbite has quit IRC16:25
johnthetubaguyyeah, a better alternative to logs probably makes sense, it just they are "easy"16:25
*** vilobhmm11 has quit IRC16:25
superdanthere is a vmware version check, IIRC16:26
johnthetubaguydanpb: xenserver probably doesn't I should add that16:26
danpbgaryk: does vmware driver check or enforce any vcenter version number ?16:26
* danpb also wonders if xenserver/vmware nova drivers have ever dropped support for older versions of the hypervisor16:27
*** patrickeast has joined #openstack-nova16:28
*** tonytan4ever has joined #openstack-nova16:28
*** claudiub has joined #openstack-nova16:29
superdanvmware did16:29
superdandropped ESX4 and then I think ESX entirely, IIRC16:29
johnthetubaguysuperdan: +1 thats what I was remembering as well16:31
*** cknight has quit IRC16:32
*** tonytan4ever has quit IRC16:34
superdanjohnthetubaguy: did you see the discussion yesterday about exposing live migrations and evacuations to the user?16:35
baoli_superdan: https://review.openstack.org/#/c/183061/2/nova/compute/manager.py, about your comments, is the change at line 1705 in the dependent patch https://review.openstack.org/#/c/183131/3/nova/tests/unit/compute/test_compute_mgr.py, good enough?16:35
johnthetubaguysuperdan: I did not16:35
*** Marga_ has quit IRC16:35
superdanbaoli_: you should push that back into the previous patch and then I'll be okay yeah16:35
*** haypo has quit IRC16:36
*** ZZelle has quit IRC16:36
superdanjohnthetubaguy: so with my robustify-evacuate set, we will get migration records for live migrations and evacuations16:36
*** redbeard has quit IRC16:36
superdanjohnthetubaguy: the question is whether to show them to the user, because right now they don't see those things16:36
baoli_superdan, sure, I can do that16:36
*** nihilifer has joined #openstack-nova16:36
superdanjohnthetubaguy: there were two views, one op wanting users to always see them, and one op wanting them to be still hidden16:36
*** browne has joined #openstack-nova16:37
superdanjohnthetubaguy: so the plan is to introduce a hidden boolean on the migration record, and let admins provide hidden=True to live migrations and evacuations if they want, and the API will just filter them out if you don't have can-see-admin-migrations=True in your policy16:37
superdanjohnthetubaguy: so I need to amend my spec to say that I'm adding the field, and changing the API thusly, but that the actual is-hidden flags being added to the migration apis is a separate deal16:38
johnthetubaguysuperdan: oh I had patches for live-migrate migrate records I was going to push, but they are a complete hack, thats totally worth it16:39
superdanjohnthetubaguy: oh, well, mine are up16:39
superdanhttps://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/robustify-evacuate,n,z16:39
johnthetubaguysuperdan: sweet, thats all good, I should review that16:39
superdanjohnthetubaguy: so we can hold on the api one (which is last) until I update the spec16:39
johnthetubaguysuperdan: gotcha, I think that makes sense16:40
johnthetubaguysuperdan: so hidden = True, I guess its a bit like the "admin" lock API16:40
*** tjones1 has joined #openstack-nova16:41
superdanjohnthetubaguy: howso?16:41
johnthetubaguysuperdan: where admin means "has access but is not owner" or something16:41
johnthetubaguysuperdan: oh, so admins would see all migrations, users would see a sub set16:41
superdanjohnthetubaguy: I was just going to add a policy element "can see admin migrations" and if you have that, you see migrations with hidden=True16:41
*** unicell has joined #openstack-nova16:41
garykdanpb: no we do not enforce. we support multiple versions with the driver.16:42
superdanit's not really admin, it's hidden, and "can you see hidden migrations" is the thing16:42
*** ddutta has joined #openstack-nova16:42
johnthetubaguysuperdan: right, thats cool16:42
garykin some cases we need to check attributes that are exposed by certain versions16:42
johnthetubaguygaryk: but you have a min_version I thought?16:42
*** arnaud____ has quit IRC16:42
garykjohnthetubaguy: the min version that we support is 5.116:42
johnthetubaguygaryk: do you check that?16:42
*** unicell has quit IRC16:42
garykno, we do not. an example on where we check it dynamically in the code is - https://review.openstack.org/16575016:43
*** unicell has joined #openstack-nova16:43
garykjohnthetubaguy: we also would like to start to suggest that people use 5.516:43
superdangaryk: so what happens if someone points you at <5.1?16:43
garyksuperdan: some features may not work, for example VSAN16:43
*** apuimedo has quit IRC16:43
superdanso that's not really a minimum version then, right?16:44
garykbut anyone who wants to use these fetaures will use the correct VC version16:44
garyksuperdan: vsan is supported from 5.516:44
garyki am not sure i understand your question/comment16:44
superdangaryk: what happens if someone points openstack at version 4.016:44
garyksuperdan: vms deployments will work16:45
garykthey will not be able to use some features, for example, storage policies. these are documented that they require 5.5 and above16:45
superdanyou test this or you just assume so?16:45
danpbgaryk: presumably there is some version that is so old that openstack will just be broken  ?16:45
danpbie inability to launch vms or something similarly major16:45
*** andreykurilin1 has quit IRC16:45
garykour current CI testing is done with 5.5 and I am using 5.5 and 6.016:45
garykin the wiki, docs etc. we have 5.1 as the min suggested supported version16:46
superdanwe're not asking about docs or suggestions16:46
superdanwe're asking if it actually works at 4.016:47
superdanor 3.016:47
*** yamahata has quit IRC16:47
superdanor whatever :)16:47
johnthetubaguysuperdan: anyways, is there a reason we want to hide live-migrates and migrations, you would see the state change in the API, maybe you we always want to expose that…not sure…16:47
*** subscope has joined #openstack-nova16:47
garyksuperdan: no, no one has tried those versions. it should officially be supported from 5.116:47
danpbanyway, weekend time - enjoy the summit folks - sorry i won't be there16:47
superdanjohnthetubaguy: there was the argument that right now people don't see that record, if they didn't notice it at the time it was happening16:48
*** ildikov_afk has quit IRC16:48
garykdanpb: sorry to hear that. would have been nice to see you16:48
garyksuperdan: i am not really sure i understand what you want, do we need a warning in the driver? if so I can add that16:48
superdanjohnthetubaguy: so it's just "ops choice"16:48
*** danpb has quit IRC16:48
*** nelsnelson has quit IRC16:48
superdangaryk: we're just asking because in libvirt we go to great pains to try to be very upfront about what is supported by which version, and what our minimum is (enforced, not suggested)16:49
johnthetubaguysuperdan: yeah, I just wonder if we should not add that feature and see if people dislike that, and add it later if needed?16:49
superdanjohnthetubaguy: well, it was alaski that wanted the ability to do that16:49
*** otter768 has joined #openstack-nova16:49
superdanjohnthetubaguy: but the deal is,16:49
superdanjohnthetubaguy: I'm adding the flag, defaulting it to False16:50
garyksuperdan: ok, thanks for the clarification.16:50
superdanjohnthetubaguy: if we never add the ability to set it, then it's zero cost, and if someone needs it suddenly, it'll be there in the schema and we can just do it16:50
superdanjohnthetubaguy: or they can flip it manually in the db16:50
johnthetubaguysuperdan: true, its not a big deal16:50
johnthetubaguysuperdan: I just feel like we should tell people about all their moves16:51
superdanjohnthetubaguy: that was the other side of the argument16:51
superdanjohnthetubaguy: the thing is, os-migrations is admin-only right now anyway :)16:51
superdanjohnthetubaguy: but just trying to think ahead16:51
*** mpaolino has quit IRC16:51
*** alanf-mc has joined #openstack-nova16:51
johnthetubaguysuperdan: yeah, it does seem like a good compromise16:52
johnthetubaguysuperdan: its the host names that we would not want to expose, I think16:52
superdanjohnthetubaguy: regardless, the spec said no api changes, so even just exposing live migrations now is kindof an api change16:52
*** armax has quit IRC16:52
johnthetubaguysuperdan: true16:52
johnthetubaguysuperdan: thanks for the heads up on that though, good to know!16:53
*** nelsnelson has joined #openstack-nova16:53
*** VW has quit IRC16:53
superdannp16:53
superdanthe schema part has already landed, so the hidden flag is there, whether you like it or not :P16:53
johnthetubaguysuperdan: :)16:54
superdanI dunno what the appropriate britishism is16:54
superdanin america, we would say "neener neener"16:54
*** otter768 has quit IRC16:55
johnthetubaguysuperdan: we just turn away and then smile internally I guess?16:56
superdanhehe16:56
*** armax has joined #openstack-nova16:58
*** Sukhdev has joined #openstack-nova16:58
*** thedodd has joined #openstack-nova17:00
openstackgerritBaodong (Robert) Li proposed openstack/nova: Split instance event/tag correctly  https://review.openstack.org/18306117:00
*** ildikov_afk has joined #openstack-nova17:01
superdansomeone please approve this trivial fix: https://review.openstack.org/#/c/183061/317:01
*** jogo is now known as flashgordon17:02
*** artom has joined #openstack-nova17:02
*** markmcclain has quit IRC17:03
flashgordonalaski: you beat me to the punch17:03
superdansomeone other than alaski17:03
superdanalaski really doesn't know how to have a day off17:03
alaski:)17:04
* superdan plans to talk to caroline about this17:04
alaskiuh oh, she totally agrees with you17:04
*** cknight has joined #openstack-nova17:04
*** dimsum__ has quit IRC17:04
*** mpaolino has joined #openstack-nova17:04
superdanwell, we'll work on an improvement plan17:04
*** Marga_ has joined #openstack-nova17:04
openstackgerritBaodong (Robert) Li proposed openstack/nova: Split instance event/tag correctly  https://review.openstack.org/18306117:05
*** annegentle has joined #openstack-nova17:05
*** dimsum__ has joined #openstack-nova17:05
openstackgerritBaodong (Robert) Li proposed openstack/nova: Use EnumField for instance external event name  https://review.openstack.org/18313117:05
openstackgerritBaodong (Robert) Li proposed openstack/nova: Handle port delete event  https://review.openstack.org/17939017:05
*** thedodd has quit IRC17:07
*** annegentle has quit IRC17:10
*** yamahata has joined #openstack-nova17:10
*** markmcclain has joined #openstack-nova17:11
*** mpaolino has quit IRC17:11
flashgordonsuperdan: want to do the honors on this one https://review.openstack.org/#/c/183220/217:12
*** jistr has quit IRC17:12
superdannot sure it's an honor :)17:12
*** jecarey has quit IRC17:13
cburgessOnly 0.10.2? Thats... ancient.17:14
*** aix has quit IRC17:15
superdanyeah, I wish we would be a little more aggressive, TBH17:15
cburgessNewer libvirts and qemu versions would solve a number of bugs that we handle poorly or with complex code today.17:16
cburgessits sad really17:16
cburgessSame argument we always have.17:16
flashgordonyeah, baby steps are still steps17:16
cburgessBetter then nothing yes.17:17
cburgessNot saying don't do it.17:17
flashgordonalso min version != recommended version or version we test with17:17
*** nihilifer has quit IRC17:21
*** afaranha has joined #openstack-nova17:22
flashgordondimsum__: thanks for doing https://review.openstack.org/#/c/181554/1017:25
*** mestery has joined #openstack-nova17:26
*** achanda has joined #openstack-nova17:27
*** kmartin has joined #openstack-nova17:29
*** amotoki has quit IRC17:33
*** nihilifer has joined #openstack-nova17:35
openstackgerritDan Smith proposed openstack/nova: Make InstanceExternalEvent use an Enum for status  https://review.openstack.org/18362617:35
*** annegentle has joined #openstack-nova17:35
*** markmcclain has quit IRC17:37
*** annegentle has quit IRC17:40
openstackgerritMatt Riedemann proposed openstack/nova: Show 'reserved' status in os-fixed-ips  https://review.openstack.org/16896617:41
mriedem_lunchalex_xu: thanks, you're my hero :) ^17:41
*** mriedem_lunch is now known as mriedem17:41
*** angdraug has joined #openstack-nova17:41
peceif I return NotificationResult.REQUEUE from endpoint handler, notification listener is presented once again with requeued message - the loop is created. How to handle this?17:42
*** subscope has quit IRC17:44
garykmriedem: do you have the link on hand with the libvirt version?17:47
Glauprebaoli_ irenab sadasu moshele : one of you are here?17:47
mriedemgaryk: https://wiki.openstack.org/wiki/LibvirtDistroSupportMatrix17:47
garyktx17:48
*** ddutta has quit IRC17:48
*** subscope has joined #openstack-nova17:49
baoli_Glaupre: hi17:50
*** tonytan4ever has joined #openstack-nova17:51
Glauprebaoli_: hello, I have a question regarding the binding of one VF for SR-IOV17:54
Glauprebaoli_: you redirected to moshele last time.17:55
openstackgerritJoe Gordon proposed openstack/nova: Update docs layout  https://review.openstack.org/18322117:55
openstackgerritJoe Gordon proposed openstack/nova: Prune development.environment doc  https://review.openstack.org/18288417:55
openstackgerritJoe Gordon proposed openstack/nova: Remove jenkins, launchpad and gerrit docs  https://review.openstack.org/18334217:55
openstackgerritJoe Gordon proposed openstack/nova: Make blueprints doc a reference for nova blueprints  https://review.openstack.org/18334317:55
openstackgerritJoe Gordon proposed openstack/nova: Add note to doc explaining scope  https://review.openstack.org/18288117:55
flashgordonjohnthetubaguy: if your around it would be great to get some very preliminary feedback on my docs changes ^17:55
*** VW has joined #openstack-nova17:56
baoli_Glaupre: ok17:58
Glauprebaoli_: I am not sure to understand where nova compute bind a VF to an instance.17:58
Glauprefor SR-IOV17:58
baoli_Glaupre: it's the same as the regular port.17:59
Glauprewhere that happens? I didn't find that17:59
*** Nic has joined #openstack-nova17:59
*** browne has quit IRC17:59
baoli_Glaupre, it happens in the nova compute service, checkout the libvirt driver and the vif driver in nova18:00
baoli_Glaupre: assuming libvirt/kvm is used18:00
*** browne has joined #openstack-nova18:01
Glaupreall right, tks18:01
*** busterswt has quit IRC18:01
*** tjones2 has joined #openstack-nova18:05
*** tjones1 has quit IRC18:07
*** Nic has quit IRC18:08
*** vilobhmm1 has joined #openstack-nova18:09
*** vilobhmm11 has joined #openstack-nova18:11
*** annegentle has joined #openstack-nova18:11
*** mriedem has quit IRC18:13
*** vilobhmm1 has quit IRC18:14
artomhttps://github.com/openstack/python-novaclient/blob/master/novaclient/v2/contrib/host_evacuate_live.py#L6118:14
artomWould that ever return more than 1 hypervisor?18:14
*** Nic has joined #openstack-nova18:15
*** Murali has joined #openstack-nova18:15
artomTo me it would only make sense if there was, for example kvm and xen on the same host... Does Nova support that?18:15
*** Nikolay_St has joined #openstack-nova18:15
superdanartom: it would for vmware I think18:16
*** ArxCruz has joined #openstack-nova18:16
superdanand maybe ironic18:16
artomI don't think live migrations make sense in ironic :P18:16
superdanyou asked18:17
artom(Is Friday casual Friday, only instead of Hawaiian shirts and jeans it's funky IRC handles?)18:18
superdanyes18:18
artomsuperdan, but live migrations *do* make sense in vmware...18:18
superdanyes, but regardless, that client method doesn't know what driver is in use, so even with ironic, the line you quoted would return multiple things for ironic, AFAIK18:19
openstackgerritVladik Romanovsky proposed openstack/nova-specs: Allow admins to query user keypairs and create new on their behalf  https://review.openstack.org/17557918:19
artomYeah... I mean, if it's looping over hypervisors, it must mean there are cases where there's more than 118:19
artomI was trying to understand what those cases are18:19
*** artom is now known as notartom18:20
superdanit's confusing because I think that the terminology is different in the api and elsewhere in nova.18:20
superdanbut I think that's what it's doing there, yeah18:20
notartom... when's a hypervisor not a hypervisor?18:21
*** openstackgerrit has quit IRC18:22
superdanvladikr: on the GET request, I meant that the result of a GET request doesn't currently include user-id in the body, does it?18:22
superdannotartom: there are hosts and nodes18:23
*** openstackgerrit has joined #openstack-nova18:23
superdannotartom: a distinction we introduced for baremetal and have regretted ever since18:23
*** eglynn has quit IRC18:23
notartomsuperdan, what was the distinction for baremetal?18:24
*** mriedem has joined #openstack-nova18:24
superdannotartom: in baremetal and ironic, one compute host is responsible for more than one "nodes", where those nodes are actual ironic boxes that look like hypervisors that can only take one vm each18:24
notartomsuperdan, so strictly speaking, 'node' should only exist in baremetal/ironic world.18:26
*** Marga_ has quit IRC18:26
superdannotartom: no, strictly speaking, node should not exist anywhere :)18:26
superdanbut the ironic and vmware drivers use the construct18:27
notartomsuperdan, hah :) So what would you call the 'node' in ironic?18:27
notartomA hypervisor?18:27
superdanit's not the naming that is the problem,18:27
superdanit's the 1:N relationship18:27
notartomSo in Ironic you would have prefered each err, 'node' to run nova-compute?18:28
notartomExcept that's impossible...18:29
superdanno18:29
superdanand it's not impossible :)18:29
superdananyway, it's a long discussion, and not directly related to your question18:30
superdanwe have a session at summit to revisit it a little18:30
*** ijw has joined #openstack-nova18:31
mriedemwe have ~13 minutes at the summit to discuss it18:32
openstackgerritMichael Bayer proposed openstack/nova: Provide working SQLA_VERSION attribute  https://review.openstack.org/16599318:32
mriedem:)18:32
superdanmriedem: which is plenty of time since it's not going to matter anyway :P18:32
*** cknight has quit IRC18:32
notartommriedem, heh, I won't be at the summit - but I'm really just curious, not trying to question past decision.18:32
*** cknight has joined #openstack-nova18:32
*** shakamunyi has joined #openstack-nova18:36
*** barra204 has joined #openstack-nova18:36
openstackgerritVladik Romanovsky proposed openstack/nova: Adding user_id handling to keypair index api call  https://review.openstack.org/7048518:38
openstackgerritVladik Romanovsky proposed openstack/nova: Adding user_id handling to keypair show api call  https://review.openstack.org/18364318:38
openstackgerritVladik Romanovsky proposed openstack/nova: Adding user_id handling to keypair create api call  https://review.openstack.org/18364418:38
openstackgerritVladik Romanovsky proposed openstack/nova: tests: functional tests for keypairs create in API microversion 2.4  https://review.openstack.org/18364518:38
*** Murali has quit IRC18:40
*** lucas-packing has quit IRC18:43
openstackgerritDan Smith proposed openstack/nova-specs: Amend robustify-evacuate to cover additional details  https://review.openstack.org/18364918:45
superdanshould be an easy amendment to land ^18:45
*** tonytan4ever has quit IRC18:49
*** otter768 has joined #openstack-nova18:50
*** nihilifer has quit IRC18:53
*** markvoelker has joined #openstack-nova18:53
*** otter768 has quit IRC18:55
*** tonytan4ever has joined #openstack-nova18:55
*** tojuvone has quit IRC18:55
*** lucas-packing has joined #openstack-nova18:55
*** Marga_ has joined #openstack-nova18:57
*** Sukhdev has quit IRC18:58
flashgordonsuperdan: you see this before http://logs.openstack.org/38/145738/14/gate/gate-tempest-dsvm-nova-v21-full/0ada47f/logs/screen-n-api.txt.gz?level=TRACE#_2015-05-15_18_12_14_69319:00
flashgordonmessage:"ObjectActionError: Object action destroy failed because: host changed"19:00
flashgordon8 hits in 2 days19:00
flashgordonfor cells and v21 jobs19:00
superdaninteresting19:00
*** Marga__ has joined #openstack-nova19:01
flashgordonjust caused https://review.openstack.org/145738 to bounce out of the gate19:01
superdanit's actually from the db api19:01
superdanConstraintNotMet19:01
flashgordonohh19:01
*** Marga_ has quit IRC19:01
superdanbut might not be host-related,19:01
superdandunno why we translate any constraint violation on delete to that19:02
flashgordonyeah at the very least its misleading19:02
*** angdraug has quit IRC19:04
*** nihilifer has joined #openstack-nova19:05
*** Marga__ has quit IRC19:05
flashgordonmriedem: question about https://review.openstack.org/#/c/182884/2..5/doc/source/development.environment.rst,cm19:09
openstackgerritEric Brown proposed openstack/nova: VMware: Missing docstring on parameter  https://review.openstack.org/18313319:09
flashgordondo you mean we don't need libvirt-dev to run the unit tests anymore?19:09
flashgordonso we should remove it?19:10
flashgordonalso removing the whole 12.04 section19:10
mriedemflashgordon: you don't need libvirt-dev to run the unit tests, no19:10
mriedemyou just need libvirt/libvirt-bin to run devstack19:11
flashgordonanything else we can remove?19:11
mriedemyou'd also need libvirt-python which we need to add back to test-requirements.txt for nova, i'm planning on doing that19:11
flashgordongraphiz?19:11
mriedemflashgordon: graphiz is for tox -e docs19:11
flashgordonahh19:11
mriedemgraphviz19:11
flashgordonwhat uses libssl?19:11
mriedemprobably the pycrypto stuff19:12
sdagueyeh19:12
*** Marga_ has joined #openstack-nova19:12
sdaguethe cert management code needs it19:12
melwittflashgordon: that failure is supposed to be addressed by this patch https://review.openstack.org/#/c/182772/19:12
flashgordonahh I was thrown off because my env already had it installed19:14
openstackgerritJoe Gordon proposed openstack/nova: Prune development.environment doc  https://review.openstack.org/18288419:14
openstackgerritJoe Gordon proposed openstack/nova: Remove jenkins, launchpad and gerrit docs  https://review.openstack.org/18334219:14
openstackgerritJoe Gordon proposed openstack/nova: Make blueprints doc a reference for nova blueprints  https://review.openstack.org/18334319:14
flashgordonmelwitt: context?19:14
*** Marga__ has joined #openstack-nova19:14
*** Marga_ has quit IRC19:14
*** Nikolay_St has quit IRC19:14
*** ajo has joined #openstack-nova19:15
melwittflashgordon: the "ObjectActionError: Object action destroy failed because: host changed", for cells. sorry I missed that it's happened for non-cells too19:16
*** Nikolay_St has joined #openstack-nova19:16
flashgordonahhh19:17
*** angdraug has joined #openstack-nova19:19
*** angdraug has quit IRC19:20
*** ajo has quit IRC19:20
*** IanGovett has quit IRC19:31
*** IanGovett has joined #openstack-nova19:32
*** alanf-mc has quit IRC19:32
*** annegentle has quit IRC19:33
*** alanf-mc has joined #openstack-nova19:34
*** markmcclain has joined #openstack-nova19:38
*** Nic has quit IRC19:40
*** okrieg has quit IRC19:40
*** busterswt has joined #openstack-nova19:41
tonybsdague: Thanks.  I'll do it when I land in Vancouver :) Yay for 30 hours transit!19:44
*** ijw has quit IRC19:44
*** ijw has joined #openstack-nova19:45
superdanomg stop complaining19:46
superdanI have 9019:46
superdanwait, that's 90 MINUTES, nevermind :)19:46
*** apuimedo has joined #openstack-nova19:47
*** barra204 has quit IRC19:48
*** shakamunyi has quit IRC19:48
*** cknight has quit IRC19:50
*** subscope has quit IRC19:54
*** vilobhmm11 has quit IRC20:04
*** vilobhmm1 has joined #openstack-nova20:04
*** Sukhdev has joined #openstack-nova20:05
sdagueheh20:06
*** okrieg has joined #openstack-nova20:07
sdaguetonyb: how far is tokyo going to be for you? or can you just never win?20:07
mriedemwacka wacka http://logs.openstack.org/21/183321/1/gate/gate-tempest-dsvm-full/a263953/logs/screen-n-cond.txt.gz?level=TRACE#_2015-05-15_15_06_44_17620:08
mriedemwe should probably handle InstanceNotFound there20:08
mriedemrather than trace in n-cond20:08
*** Marga__ has quit IRC20:08
mriedemsuperdan: you agree? ^20:09
*** nelsnelson has quit IRC20:09
*** raildo has quit IRC20:10
*** Marga__ has joined #openstack-nova20:11
superdanmrda: yeah20:12
superdanoops20:12
superdanmriedem: I was going to say make it an expected exception, but in the case of that notification, yeah handle and ignore or something20:12
mriedemk, mtreinish and i were digging through logs20:13
mriedemfinding goodies20:13
*** alexpilotti has quit IRC20:17
*** achanda has quit IRC20:18
*** busterswt has quit IRC20:18
*** notartom has quit IRC20:21
*** vladikr has quit IRC20:23
*** ijw has quit IRC20:29
*** baoli_ has quit IRC20:31
*** baoli has joined #openstack-nova20:31
*** tonytan4ever has quit IRC20:32
openstackgerritChris Friesen proposed openstack/nova: fix "down" nova-compute service spuriously marked as "up"  https://review.openstack.org/16306020:33
*** ajo has joined #openstack-nova20:34
cfriesensuperdan: so what's the story on this one? ^  are we going to hold off and re-architect service reporting first, or submit something like this now and fix up later?20:35
superdancfriesen: is that relevant with the forced-down thing?20:36
superdanI guess it still kidna is20:36
superdanI'm not in favor of blocking things that fix issues in the db driver yet, no20:36
*** achanda has joined #openstack-nova20:37
cfriesensuperdan: not directly.  this one is related to the fact that the db backend uses the builtin "updated_at" field, which can be updated by doing things like enabling/disabling the service20:37
superdanright20:37
cfriesensuperdan: we'd still want a "forced_down" field or something similar20:38
*** afaranha has quit IRC20:38
superdancfriesen: surely we have a session where it'll be appropriate to discuss this next week, right?20:38
cfriesenseems likely20:38
superdandoubt it needs much20:38
cfriesenit was on hold till the 293 db migration went in20:39
cfriesenI just reworked it to build on top of the latest codebase20:39
superdanI meant doubt it needs much discussion20:39
cfriesenyeah20:39
superdandidn't I just win 293?20:39
cfriesenyep20:39
superdanokay20:39
*** Marga__ has quit IRC20:39
cfriesenor maybe it was on hold for 292, can't remember now. :)20:39
*** mwagner_lap has quit IRC20:40
*** Marga_ has joined #openstack-nova20:40
*** lucas-packing has quit IRC20:40
superdanI think 29220:40
superdanbut I won 292 and 29320:41
superdanI'm very lucky20:41
cfriesenhelps to have approver status. :)20:41
cfriesenon another note...where does the scheduler for Friday get put up?20:41
*** q3k has quit IRC20:42
superdaneh? you know we don't approve our own things right?20:42
superdancfriesen: no schedule on friday20:42
*** vilobhmm1 has quit IRC20:43
*** vilobhmm1 has joined #openstack-nova20:43
cfriesensuperdan: so how do people find out what's happening that day?  email, irc, word of mouth?20:44
*** annegentle has joined #openstack-nova20:44
superdancfriesen: there is nothing happening to find out about20:44
superdanit's discussion day20:44
superdanwe'll put things in an etherpad to discuss that come up during the week20:44
superdanand talk about them20:44
superdanthere's no schedule, no slots, you show up and participate20:45
superdanit's midcycle style20:45
*** okrieg has quit IRC20:46
openstackgerritJoe Gordon proposed openstack/nova: Remove unit_test doc  https://review.openstack.org/18369720:46
openstackgerritTomoki Sekiyama proposed openstack/nova: libvirt: Remove error messages from multipath command output before parsing  https://review.openstack.org/16556020:47
cfriesenthanks.  just making sure I know what's going on.20:48
*** DanyC has joined #openstack-nova20:50
superdancfriesen: you'll get it pretty quick once you're there20:51
superdancfriesen: find someone you know and ask if you're worried about missing things20:51
*** q3k has joined #openstack-nova20:51
*** otter768 has joined #openstack-nova20:51
*** aysyanne has quit IRC20:55
*** otter768 has quit IRC20:56
*** okrieg has joined #openstack-nova20:56
*** vilobhmm1 has quit IRC20:57
*** zzzeek has quit IRC20:58
*** davideagnello has quit IRC21:01
*** armax has quit IRC21:03
*** jwcroppe_ has joined #openstack-nova21:04
*** davideagnello has joined #openstack-nova21:04
*** jwcroppe has quit IRC21:06
DanyChi all (sorry to ask this type of question here but is the only hope i have that i can get a spot on answer). I'm running IceHouse on Ubuntu 14.04.2 and few days ago the kernel on a compute node crashed (that is the theory) and i just discovered due to being unable to access the instances on that compute node. I've paste the logs at http://paste.openstack.org/show/224197/ and it think s'thing crashed on libvirtd21:07
*** jwcroppe has joined #openstack-nova21:08
openstackgerritMatt Riedemann proposed openstack/nova: Add libvirt-python to test-requirements.txt to enforce min version  https://review.openstack.org/18370621:08
superdanDanyC: see topic please21:08
DanyCsorry, thx.21:09
*** DanyC has left #openstack-nova21:10
*** jwcroppe_ has quit IRC21:10
*** kylek3h has quit IRC21:11
*** htruta has quit IRC21:15
*** zzzeek has joined #openstack-nova21:16
*** cbader has quit IRC21:17
*** rainya has joined #openstack-nova21:19
openstackgerritgaryk proposed openstack/nova: VMware: enforce minimum support VC version  https://review.openstack.org/18371121:27
garyksuperdan: mriedem: johnthetubaguy: ^21:27
superdangaryk: we weren't really saying you needed to do that, but I do think it's a good idea21:28
superdanso.. thanks :)21:28
*** lpetrut has quit IRC21:28
*** jwcroppe_ has joined #openstack-nova21:28
*** gholler_ has joined #openstack-nova21:28
*** jwcroppe has quit IRC21:30
garyksuperdan: after the interaction this morning it seemd like the responsible thing to do21:31
mriedem-121:31
superdangaryk: I think it is21:31
mriedemgaryk: what you have there is not the same as https://review.openstack.org/#/c/183220/21:31
*** ociuhandu has quit IRC21:31
superdangaryk: what do you think it should do if it's too old?21:32
*** gholler has quit IRC21:32
mriedemthis prevents nova-compute from starting21:32
mriedemwhich is not ok w/o a deprecation cycle imo21:32
garykin the docs etc we say that it supports 5.1.0 onwards so i guess exiting is ok.21:32
garykif you guys think it is too extreme we can change to a warning and enforce in M21:33
*** jwcroppe has joined #openstack-nova21:33
garyki know that is is only tested from 5.1.0 onwards. mu understanding is that suds also does not work properly with 5.0.021:33
superdanmriedem: so if we think it won't work on <5.1 we should start and fail to do things in a half-broken way?21:33
mriedemgaryk: can you link me to the docs?21:34
superdanseems like exiting is the right thing to do21:34
garyksec, i'll look21:34
superdanwe abort startup in libvirt for config issues21:35
*** jwcroppe_ has quit IRC21:35
superdanincompatible-with-success things21:35
*** eharney has quit IRC21:36
mriedemyes we abort libvirt if the version is less than 0.9.11 (and for parallels if you're using parallels but don't have new enough libvirt)21:36
superdanif it's in the docs, I don't think we need a docimpact, but putting it in the release notes that we actaully enforce it seems like a good thing21:36
garykmriedem: http://docs.openstack.org/kilo/config-reference/content/vmware.html21:36
garykfor vc 5.0.0 they need to do all kinds of things to get wsdl (it was planned and the code never ever pushed as it was not used)21:37
superdanhmm, does that actually say 4.1 is supported?21:37
mriedem"This section describes how to configure VMware-based virtual machine images for launch. vSphere versions 4.1 and later are supported."21:37
*** dprince has quit IRC21:37
mriedem"Copying VMDK files (vSphere 5.1 only). In vSphere 5.1, copying large image files (for example, 12 GB and greater) from Glance can take a long time. To improve performance, VMware recommends that you upgrade to VMware vCenter Server 5.1 Update 1 or later. For more information, see the Release Notes."21:37
mriedem"vSphere 5.0 and earlier additional set up"21:38
mriedemyeah so this isn't clear21:38
superdanso I think docimpact is right, and that should be updated to just say 5.1 across the board21:38
garyki think that the docs really need to be updated21:38
*** dave-mccowan has quit IRC21:39
superdanmriedem: I don't think we need a deprecation cycle for this really.. do you feel strongly on that?21:39
mriedemsuperdan: we do deprecation cycles for far less21:40
*** jwcroppe_ has joined #openstack-nova21:40
superdanmriedem: but we're not dropping functionality that we test and that works21:40
mriedemi'm not as familiar with vcenter as i am with libvirt versions, so i don't appreciate how old 4.1 may be21:40
superdanmriedem: 5.1 was released in 201221:40
superdan5.5 is what they recommend you run21:41
superdan4.1 was 201021:41
mriedemwhere do they recommend that?21:41
tjones24.1 is 5 years old21:41
tjones22010-07-1321:41
superdanthat's as old as openstack21:41
mriedem5.5 isn't in http://docs.openstack.org/kilo/config-reference/content/vmware.html21:41
superdanmriedem: they == garyk :)21:41
*** erlon has quit IRC21:41
mriedemexactly21:41
mriedemi can't tell from the nsx ci logs what version of vcenter is being run either21:42
superdan5.5 is still 18 months old21:42
superdanhe said they're on 5.1 right?21:42
garykoriginally our CI was 5.121:42
*** jwcroppe has quit IRC21:42
garykwe moved it to 5.5 a few weeks ago21:42
*** alanf-mc_ has joined #openstack-nova21:42
garykmaybe even a few months ago (everything is a blur with all the + and - 1's)21:43
mriedemso imo libvirt 0.9.11 is old but people came out of the woodwork on the mailing list over mentioning going to 1.2.2 (which is the version we test  against)21:43
mriedemso vcenter 4.1 is old balls, i get it21:43
garyki am not sure that anyone is using anything below 5.1.021:43
mriedemnot being sure and knowing are different things, so this is at least an openstack-operators mailing list thread imo21:43
superdan4.x is EOL http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=203956721:44
superdanhasn't been supported for a year now21:45
*** alanf-mc has quit IRC21:45
garykso what is the concensus - should i adda deprecation warning and then in M we enforce21:45
mriedemand if i were using 4.x i probably wasn't using openstack to begin with21:45
mriedemb/c i love pets21:45
mriedemgaryk: at least start in the ML21:45
garykok, will do.21:45
mriedemthe fact the kilo docs point out < 5 support is a problem imo  here21:45
superdanmriedem: if we knew it didn't work on 4.x would you feel differently?21:46
mriedemmayhap21:46
mriedemhonestly,21:46
mriedemit's nearly 5pm on a friday before leaving for the summit21:47
superdanFWIW, I also think we should be requiring a newer libvirt, so I'm clearly on the other end of this argument regardless :)21:47
mriedemi don't think this is worth rushing in21:47
mriedemi think ML first, see if anyone freaks out21:47
*** Marga_ has quit IRC21:47
superdanwell, I think we can rush the deprecation warning21:47
superdan"rush"21:47
mriedemif not, maybe < 5.0 is a failure in L, 5.1 is min in M? idk21:47
mriedemwe can certainly do the deprecation warning21:48
mriedemi'm all for that21:48
*** Marga_ has joined #openstack-nova21:48
superdanso let's do that now. sending the email could likely get ignored,21:48
superdaneven by ops folks, over the next week21:48
openstackgerritJay Pipes proposed openstack/nova-specs: Adds availability zone real data model spec  https://review.openstack.org/18371621:48
*** annegentle has quit IRC21:49
*** Marga_ has quit IRC21:50
garykmriedem: i sent a mail to the list. i think that we should enforce 5.1.0 and update the docs - which are outdated21:50
*** annegentle has joined #openstack-nova21:51
garykif someone actually uses a version less than 5.1.0 i'll buy you a beer ...21:51
superdanI'll watch you buy him a beer21:51
*** Marga_ has joined #openstack-nova21:52
garyk:)21:52
superdanbut,21:53
mriedemis there any conditional code that is for < 5.1 in here?21:53
cfriesenthink I found a doc bug.  http://docs.openstack.org/admin-guide-cloud/content/section_configuring-compute-migrations.html has "By default, the Compute service does not use the libvirt live migration function."  But I think it does now, right?  I see VIR_MIGRATE_LIVE in the default flags.21:53
superdanif there are people running 4.1, that just means we deprecate for a cycle, IMHO21:53
superdancfriesen: no, that is not wrong21:53
superdancfriesen: migrate and live-migrate are separate calls21:53
cfriesensuperdan: this is in the "enabling true live migration" section.  it used to be that even when you called the "live migration" api libvirt paused the VM before shipping the memory over21:54
superdancfriesen: eh? used to be when?21:55
cfriesensuperdan: grizzly?21:55
*** Marga_ has quit IRC21:55
cfriesensuperdan: the docs suggest setting "live_migration_flag=xxxx" but the suggestion is our current default values.21:55
dimsum__garyk: check with -operators too?21:55
dimsum__garyk: min VC version21:55
superdancfriesen: I never recall it being that way21:55
*** zsais has quit IRC21:56
mriedemi do see a docs bug for upgrade_levels.compute21:56
mriedemoh snap, it was a code bug21:57
cfriesensuperdan: git commit 26504d71 set the VIR_MIGRATE_LIVE flag by default.  went into 2014.2.b321:58
*** davideagnello has quit IRC21:58
mriedemsuperdan: heh https://review.openstack.org/#/c/183718/ - the kilo docs talk about upgrading from havana to icehouse21:59
superdanmriedem: for reference, libvirt 0.9.2 was 201421:59
superdanmriedem: so that's much more in line with 5.121:59
superdanmriedem: heh, nice21:59
superdancfriesen: so when you say "it used to be" you meant "by default" ?22:00
cfriesensuperdan: sorry, yes.  prior to that commit live-migration would pause the VM unless you edited the config file22:01
cfriesensuperdan: the docs haven't been updated since that commit22:01
superdancfriesen: alright22:01
superdancfriesen: the text in the docs seem to apply to regular migrations not live migrations22:02
superdancfriesen: where it says "the vm will be suspended, minutes of downtime" is referring to regular migration I think22:02
superdanisn't the flag bit still relevant, potentially just with different defaults?22:02
superdanregular migration being completely different than live migration, regardless of whether the vm is paused in the latter22:03
cfriesensuperdan: table 4.6, "description of live migration configuration options"22:03
superdanright22:03
superdanit says "In most cases, you will not need to configure any options. "22:04
superdanoh I see, below that is the statement you're concerned about?22:04
cfriesensuperdan: below that: "By default, the Compute service does not use the libvirt live migration function. To enable this function, add the following line to the nova.conf file: "22:04
superdangotcha22:04
cfriesenyeah, and the one below "The Compute service does not use libvirt's live migration by default because there is a risk that the migration process will never end"22:04
cfriesenthose statements are no longer true22:04
superdanthere's a similar one at the top that is talking about regular migration22:04
*** Sukhdev has quit IRC22:05
superdanwhich I thought you were talking about22:05
superdanso yeah,  I guess so22:05
cfriesenwhat project is that, openstack-manuals?22:06
superdanI don't know exactly where that is22:06
openstackgerritJay Hennessey proposed openstack/nova-specs: blueprint qualified-uuids  https://review.openstack.org/13262322:07
*** burt has quit IRC22:08
openstackgerritgaryk proposed openstack/nova: VMware: enforce minimum support VC version  https://review.openstack.org/18371122:08
garykmriedem: ^ - it now has a warning and in M we will take the dogs for a rabies shot22:09
*** hanzhang has quit IRC22:09
*** davideagnello has joined #openstack-nova22:09
*** hanzhang has joined #openstack-nova22:10
melwittcfriesen: I think it is openstack-manuals, in the upper right corner there is a bug icon and it links to that launchpad22:10
superdangaryk: more stronger wors22:11
superdaner, words22:12
cfriesenmelwitt: excellent, thanks.   that icon shows up as junk on my firefox.22:12
superdangaryk: and I still think we should poll and drop earlier than M if signs point to mostly no use22:13
melwittoh, odd22:13
cfriesenmelwitt: could be a local setting.  shows up okay in konqueror.22:14
melwitthah, I see it too in firefox. I usually use chrome. must be unicode thing22:14
*** Marga_ has joined #openstack-nova22:14
*** IanGovett has quit IRC22:14
annegentlemelwitt: oo tell me the link22:14
melwittannegentle: http://docs.openstack.org/admin-guide-cloud/content/section_configuring-compute-migrations.html22:14
cfriesenannegentle http://docs.openstack.org/admin-guide-cloud/content/section_configuring-compute-migrations.html22:14
cfriesen:)22:14
mriedemgaryk: the rabies shot would prevent the dog from getting rabies, right?22:15
mriedemi think you mean, take the rabid dog out back and shoot it22:15
*** claudiub has quit IRC22:17
*** claudiub has joined #openstack-nova22:17
garykits a pet that is starting to foam at the mouth22:18
*** ociuhandu has joined #openstack-nova22:22
mriedemwe had to take a bat up to the university of minnesota to get it's brain sliced open to check for rabies once, otherwise i would have had to pay for the big shots in my gut b/c we found the thing under the bed that i was sleeping in22:23
mriedemi have pictures of it in a tupperware container22:24
superdanno thanks22:24
mriedemno turds in this one22:25
melwittmy word.22:26
superdanmelwitt: have you seen the turd picture?22:27
superdanmriedem would love to show you22:27
melwittno please22:27
*** okrieg has quit IRC22:29
mriedemplease hold22:30
*** penick has joined #openstack-nova22:31
*** penick has quit IRC22:31
*** penick has joined #openstack-nova22:31
penickI, however, would love to see said turd picture.22:32
penick...22:32
penickJust kidding.22:32
*** zzzeek has quit IRC22:34
*** apuimedo has quit IRC22:34
*** rfolco has quit IRC22:34
mriedemmelwitt: penick: enjoy https://plus.google.com/106668358919254945795/posts/dsysZwrBWy1?pid=6149230493461843458&oid=10666835891925494579522:35
mriedemgoogle really blew it up22:35
superdanmriedem: gotta explain it22:35
mriedemthe caption does22:36
penickI’m having second thoughts.22:36
mriedemour dog ate a tube sock22:36
mriedemthat came out a week later22:36
* penick stares wide eyed at the URL, afraid to click22:36
penickWell hell, now I gotta.22:36
mriedemi was overjoyed b/c otherwise it would have been possibly surgery to remove it22:36
penickCan’t un-see that.22:36
*** jdurgin has joined #openstack-nova22:36
melwittso not clicking it22:36
penickYeah. Don't.22:37
mriedemoh come on22:37
mriedemthere are infinitely worse things to see22:37
* mriedem goes to find some22:37
claudiub:))22:37
penickThere are. Also, now if anyone googles me after the summit they’ll find an IRC log of me demanding a turd pic.22:38
penickI’m not sure how I feel about that yet.22:38
claudiubindeed, there are worse things in the world than that simple picture :P22:38
mriedemthe one i showed comstud in utah was probably worse22:38
mriedemsince it was human22:39
melwittlol22:39
penickMy word.22:39
*** yamahata has quit IRC22:48
*** jamielennox|away is now known as jamielennox22:48
openstackgerritJay Hennessey proposed openstack/nova-specs: blueprint qualified-uuids  https://review.openstack.org/13262322:49
*** rainya has quit IRC22:50
*** annegentle has quit IRC22:51
*** mwagner_lap has joined #openstack-nova22:52
*** otter768 has joined #openstack-nova22:52
*** diegows has joined #openstack-nova22:55
*** otter768 has quit IRC22:56
*** rainya has joined #openstack-nova22:57
*** signed8bit is now known as signed8bit_ZZZzz22:58
*** achanda_ has joined #openstack-nova23:01
*** ZZelle has joined #openstack-nova23:03
superdancfriesen: just +Wd your numa fix.. you going to do the backport?23:04
openstackgerritMatt Riedemann proposed openstack/nova: Handle InstanceNotFound when sending instance state update notification  https://review.openstack.org/18372723:04
*** achanda has quit IRC23:05
*** dims_ has joined #openstack-nova23:05
*** ociuhandu has quit IRC23:06
*** achanda_ has quit IRC23:06
openstackgerritMatt Riedemann proposed openstack/nova: Handle InstanceNotFound when sending instance update notification  https://review.openstack.org/18372723:07
*** dimsum__ has quit IRC23:07
*** dboik_ has quit IRC23:09
*** dboik has joined #openstack-nova23:10
*** Marga_ has quit IRC23:11
*** claudiub has quit IRC23:12
*** ArxCruz has quit IRC23:14
*** ArxCruz has joined #openstack-nova23:15
openstackgerritJay Hennessey proposed openstack/nova-specs: blueprint qualified-uuids  https://review.openstack.org/13262323:16
*** Marga_ has joined #openstack-nova23:18
*** ajo has quit IRC23:19
*** dave-mccowan has joined #openstack-nova23:19
*** ArxCruz has quit IRC23:24
*** hemna is now known as hemnafk23:27
openstackgerritJay Hennessey proposed openstack/nova-specs: blueprint qualified-uuids  https://review.openstack.org/13262323:28
*** andrewbogott is now known as andrewbogott_afk23:31
*** Marga_ has quit IRC23:31
*** andrewbogott_afk is now known as andrewbogott23:31
*** andrewbogott is now known as andrewbogott_afk23:33
*** gholler_ has quit IRC23:35
flashgordonsuperdan: http://interactive.blockdiag.com/?compression=deflate&src=eJxVj82KwkAQhO_zFM2cV1_AH9BEUMSNsOtJPHRmOibY2DIzsrhL3n1nohE9Nf1VdVFdspiTbfAIfwpgKa75lTMMpjDbrkaRxJG2Nd18kDM90AdkxWa7-14krQg1OUjK0cn10gVlxWcO4yj2vn0lbMke1JuNsSSGCeh8DmgMea9Tjg83pogt-ppsR2q8JMLNvYMRFpcOS0Zz0inVYsASPcG-N-uK5cfU6MKwF_Xh5amu5LPYo3TkX9lykauU2apWqfYfHMpTKQ23:35
flashgordonmriedem: ^23:35
flashgordonhow is that for a nova diagram?23:35
flashgordonas a first pass23:35
superdanlol23:35
superdantechnically mostly correct I guess,23:36
superdanbut visually kindof confusing :)23:36
superdanthe loop around sched from api to compute is.. silly :D23:36
flashgordonyeah the visial is sorta hard23:36
*** dboik has quit IRC23:37
flashgordonthat is the arrow to 'other' which is glance etc23:37
superdanwell,23:37
superdanit's also the api->compute line23:37
flashgordonthere is no api compute line in this diagram should there be?23:37
flashgordon I thought it all went through cond?23:37
superdansure23:37
superdanno :)23:37
superdanso that is compute->other?23:37
flashgordondropping other for a moment23:38
flashgordonand focus on the nova side23:38
superdanmy point is,23:38
*** VW has quit IRC23:38
superdanI though it was a line from api to compute,23:38
superdanwhich should be there23:38
superdanheh, adding api->compute makes it super confusing23:40
*** Marga_ has joined #openstack-nova23:40
flashgordonhttp://interactive.blockdiag.com/?compression=deflate&src=eJxVkEFPwzAMhe_9FVbPgz9QhgTtpCEEmwQ7TTu4ibdGs-oqMUID7b8TlzKxnKL3nOfvpWVxRx_wAN8FwCZRhG3qcCCYQ4lOJZa7qsjWIcrHMA7ZkRioV9QgfR4cJGrEoNXkWswMlnnoK_s39_CwfjLvbEH5btIznZJKT9VVdr16beAu2_XqZb15X8B2L-zJ764RGFtiI2weAZ2jlErLSXpiA_eYOvKjMlXh8LvJCUu0hy2jO5aW6lGxxUT_eu9ZPl2HUW__TPuEC_oIeQGboLP-Vi8XTVGMRXPV8w_LlGR223:43
flashgordonnot much better23:43
superdancan we not use a real drawing program?23:43
flashgordonwe can23:43
superdanthat gives me an error23:43
flashgordontry http://interactive.blockdiag.com/?compression=deflate&src=eJxtj80OgjAQhO99ig13fQHURIFEDwqJciIclrb8hI0lLcYYw7vbohhMPM7M7uy3BSneigYreDKA1EgNmamxk7AGD3mvtJf7U7LYwDY5-MzqSqtbN-4E8SmElY2C-JiklwiyUpGQIreZnYZ_yU8BYSHJnQt3gJxLYzx30fQPchQCTS3F6Hy4qLlKp7kipd1iQchbz7UK7LFAI2dPlKTuvEbdL6fw_dGXzeLPkEdt_XOwj0LmOgc2MDa8AEd2VRE23:43
superdanstill pretty roundabout, but better I guess :)23:44
flashgordonI think some people are afraid to open up a non text based editor23:45
superdanand "user" looks like a voodoo dol23:45
superdandoll23:45
flashgordonthat sounds about right23:45
superdanactually kinda does23:46
*** tjones2 has quit IRC23:47
flashgordonit turns out placement is a really really hard problem23:47
flashgordonfor diagrams23:47
flashgordonso computers are bad at ait23:47
flashgordonat it23:47
*** Sukhdev has joined #openstack-nova23:49
superdanyes, so we can leave it to the humans :)23:51
*** annegentle has joined #openstack-nova23:51
openstackgerritgaryk proposed openstack/nova: VMware: enforce minimum support VC version  https://review.openstack.org/18371123:53
*** baoli has quit IRC23:53
*** rainya has quit IRC23:56
*** annegentle has quit IRC23:57
*** garyk has quit IRC23:57
*** tsekiyam_ has joined #openstack-nova23:57
*** Marga_ has quit IRC23:57
*** dsanders has quit IRC23:57
*** tjones1 has joined #openstack-nova23:59

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