| *** andreaf has quit IRC | 00:00 | |
| *** cdub has quit IRC | 00:00 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: This changeset provides: - improved framework for API versioning, with separated controllers for each API version - Taken from nova:'Resource' class in WSGI framework. This class is a container for serializers, deserializers and controller - Better deseri https://review.openstack.org/1283 | 00:01 |
|---|---|---|
| nati2 | Hi I have a question about gerrit | 00:01 |
| *** bsza has quit IRC | 00:04 | |
| openstackgerrit | Vish Ishaya proposed a change to openstack/nova: Add local storage of context for logging https://review.openstack.org/1022 | 00:09 |
| *** jdg has quit IRC | 00:11 | |
| openstackgerrit | Salvatore Orlando proposed a change to openstack/quantum: This changeset provides: - improved framework for API versioning, with separated controllers for each API version - Taken from nova:'Resource' class in WSGI framework. This class is a container for serializers, deserializers and controller - Better deseri https://review.openstack.org/1283 | 00:16 |
| *** jog0 has quit IRC | 00:19 | |
| *** nati2 has quit IRC | 00:21 | |
| *** vladimir3p has joined #openstack-dev | 00:22 | |
| *** dtroyer has joined #openstack-dev | 00:24 | |
| *** cdub has joined #openstack-dev | 00:27 | |
| *** adjohn has quit IRC | 00:29 | |
| *** jakedahn has quit IRC | 00:29 | |
| *** jakedahn has joined #openstack-dev | 00:29 | |
| *** adjohn has joined #openstack-dev | 00:29 | |
| *** novas0x2a|laptop has quit IRC | 00:31 | |
| *** nati2 has joined #openstack-dev | 00:36 | |
| *** uncleofthestick has quit IRC | 00:41 | |
| *** dragondm has quit IRC | 00:48 | |
| *** cdub has quit IRC | 00:49 | |
| *** jakedahn has quit IRC | 00:56 | |
| *** adjohn has quit IRC | 01:01 | |
| *** danwent has quit IRC | 01:17 | |
| *** cdub has joined #openstack-dev | 01:21 | |
| *** jog0 has joined #openstack-dev | 01:24 | |
| *** ron-slc has quit IRC | 01:28 | |
| *** jog0 has quit IRC | 01:31 | |
| *** jakedahn has joined #openstack-dev | 01:39 | |
| *** cdub has quit IRC | 01:47 | |
| *** jog0 has joined #openstack-dev | 01:48 | |
| *** jog0 has quit IRC | 02:07 | |
| *** jdurgin has quit IRC | 02:17 | |
| *** pixelbeat has quit IRC | 02:22 | |
| *** danwent has joined #openstack-dev | 02:26 | |
| openstackjenkins | Project nova-docs build #1,550: SUCCESS in 4 min 22 sec: https://jenkins.openstack.org/job/nova-docs/1550/ | 02:35 |
| openstackjenkins | Vishvananda Ishaya: Add local storage of context for logging | 02:35 |
| *** mgius has quit IRC | 02:37 | |
| *** zns has joined #openstack-dev | 02:40 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Fixing immediate delete after boot on Libvirt https://review.openstack.org/1236 | 02:51 |
| *** hugokuo has joined #openstack-dev | 02:54 | |
| openstackgerrit | Jay Pipes proposed a change to openstack/glance: Fixes LP Bug#878411 - No docs for cache management https://review.openstack.org/1284 | 02:55 |
| openstackgerrit | Jay Pipes proposed a change to openstack/glance: Fixes LP Bug#885529 - NameError on _ function https://review.openstack.org/1285 | 02:55 |
| *** helloworld has joined #openstack-dev | 02:59 | |
| *** helloworld has left #openstack-dev | 02:59 | |
| *** yeming has joined #openstack-dev | 03:02 | |
| openstackjenkins | Project nova-docs build #1,551: SUCCESS in 4 min 33 sec: https://jenkins.openstack.org/job/nova-docs/1551/ | 03:05 |
| openstackjenkins | renuka.apte: Xen Storage Manager Volume Driver | 03:05 |
| *** rohitk has joined #openstack-dev | 03:13 | |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Fixes LP Bug#885434 - Documentation showing multiple tenants misleading https://review.openstack.org/1286 | 03:14 |
| *** joonwon has quit IRC | 03:14 | |
| *** mmetheny has joined #openstack-dev | 03:15 | |
| openstackjenkins | Project nova-docs build #1,552: SUCCESS in 4 min 28 sec: https://jenkins.openstack.org/job/nova-docs/1552/ | 03:15 |
| openstackjenkins | ewan.mellor: Bug 884863: nova logs everything to syslog twice | 03:15 |
| uvirtbot | Launchpad bug 884863 in nova "nova logs everything to syslog twice" [Undecided,Fix committed] https://launchpad.net/bugs/884863 | 03:15 |
| *** novas0x2a|laptop has joined #openstack-dev | 03:18 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: X.509 client authentication with Keystone. Implements blueprint 2-way-ssl https://review.openstack.org/1038 | 03:21 |
| *** sleepsontheflo-1 has joined #openstack-dev | 03:41 | |
| *** rods has quit IRC | 03:44 | |
| *** adjohn has joined #openstack-dev | 03:46 | |
| *** rohitk has quit IRC | 03:47 | |
| *** rohitk has joined #openstack-dev | 03:48 | |
| openstackjenkins | Project nova-docs build #1,553: SUCCESS in 4 min 30 sec: https://jenkins.openstack.org/job/nova-docs/1553/ | 03:50 |
| openstackjenkins | Johannes Erdfelt: Speed up tests yet another 45 seconds | 03:50 |
| *** rohitk has quit IRC | 03:52 | |
| *** negronjl has joined #openstack-dev | 03:59 | |
| *** nati2_ has joined #openstack-dev | 04:00 | |
| *** nati2 has quit IRC | 04:02 | |
| *** vladimir3p has quit IRC | 04:08 | |
| *** nati2_ has quit IRC | 04:11 | |
| *** nati2 has joined #openstack-dev | 04:12 | |
| openstackjenkins | Project nova-docs build #1,554: SUCCESS in 4 min 22 sec: https://jenkins.openstack.org/job/nova-docs/1554/ | 04:15 |
| openstackjenkins | Vishvananda Ishaya: Moves a-zone scheduling into simple scheduler | 04:15 |
| *** nati2 has quit IRC | 04:16 | |
| *** jog0 has joined #openstack-dev | 04:19 | |
| openstackgerrit | joe-gordon0 proposed a change to openstack/nova: fix bug 816630 https://review.openstack.org/1237 | 04:23 |
| uvirtbot | Launchpad bug 816630 in nova "broadcast ip is being assigned out as IP address" [Medium,Confirmed] https://launchpad.net/bugs/816630 | 04:23 |
| *** jog0 has quit IRC | 04:29 | |
| *** negronjl has quit IRC | 04:31 | |
| *** rohitk has joined #openstack-dev | 04:32 | |
| *** rohitk has quit IRC | 04:39 | |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: X.509 client authentication with Keystone. Implements blueprint 2-way-ssl https://review.openstack.org/1038 | 04:40 |
| *** gyee has quit IRC | 04:44 | |
| *** rohitk has joined #openstack-dev | 04:45 | |
| *** thickskin has left #openstack-dev | 04:56 | |
| *** chomping has joined #openstack-dev | 04:58 | |
| *** hugokuo has quit IRC | 05:01 | |
| *** vladimir3p has joined #openstack-dev | 05:15 | |
| *** zaitcev has quit IRC | 05:30 | |
| *** vladimir3p has quit IRC | 05:31 | |
| *** rohitk has quit IRC | 05:31 | |
| *** donaldngo_hp has quit IRC | 05:32 | |
| *** adjohn has quit IRC | 05:40 | |
| *** clayg has quit IRC | 05:41 | |
| *** 31NAADY1U is now known as clayg | 05:41 | |
| *** mandela123 has joined #openstack-dev | 06:03 | |
| *** cdub has joined #openstack-dev | 06:10 | |
| *** novas0x2a|laptop has quit IRC | 06:11 | |
| *** jog0 has joined #openstack-dev | 06:13 | |
| *** jog0 has quit IRC | 06:15 | |
| *** hugokuo has joined #openstack-dev | 06:24 | |
| *** andreaf has joined #openstack-dev | 06:59 | |
| *** mnour has quit IRC | 07:19 | |
| *** andreaf has quit IRC | 07:38 | |
| *** zns has quit IRC | 07:56 | |
| *** jdag has quit IRC | 08:10 | |
| *** jeffjapan has quit IRC | 08:10 | |
| *** jdag has joined #openstack-dev | 08:11 | |
| *** reidrac has joined #openstack-dev | 08:14 | |
| *** rohitk has joined #openstack-dev | 08:16 | |
| *** jdag has left #openstack-dev | 08:17 | |
| *** jdag has quit IRC | 08:17 | |
| *** rohitk has quit IRC | 08:25 | |
| *** rohitk has joined #openstack-dev | 08:29 | |
| openstackgerrit | Nikolay Sokolov proposed a change to openstack/nova: Sync URLs with trunk https://review.openstack.org/1287 | 08:32 |
| openstackgerrit | Nikolay Sokolov proposed a change to openstack/nova: Ported NetworkAdmin extention from openstackx. Migrated from bazaar. https://review.openstack.org/801 | 08:32 |
| *** mandela123 has quit IRC | 08:53 | |
| *** mnour has joined #openstack-dev | 08:56 | |
| *** yeming has quit IRC | 09:03 | |
| openstackgerrit | garyk proposed a change to openstack/nova: Blueprint - lasterror https://review.openstack.org/1288 | 09:07 |
| *** pixelbeat has joined #openstack-dev | 09:23 | |
| *** darraghb has joined #openstack-dev | 09:31 | |
| *** adjohn has joined #openstack-dev | 09:40 | |
| *** chomping has quit IRC | 09:56 | |
| *** chomping has joined #openstack-dev | 10:00 | |
| *** Oneiroi has joined #openstack-dev | 10:11 | |
| *** viraptor has joined #openstack-dev | 10:21 | |
| *** dolphm has joined #openstack-dev | 10:47 | |
| *** dolphm has quit IRC | 10:51 | |
| *** markmc has joined #openstack-dev | 10:51 | |
| *** dolphm has joined #openstack-dev | 11:17 | |
| *** naehring has joined #openstack-dev | 11:18 | |
| *** bsza has joined #openstack-dev | 11:20 | |
| *** markvoelker has joined #openstack-dev | 11:34 | |
| *** dolphm has quit IRC | 11:47 | |
| zykes- | danwent: around ? | 12:08 |
| zykes- | probably sleeping still over there but ;) | 12:08 |
| *** rkukura has joined #openstack-dev | 12:16 | |
| *** rkukura has quit IRC | 12:21 | |
| openstackgerrit | tom-hancock proposed a change to openstack/glance: Fix two issues with checked_iter https://review.openstack.org/1289 | 12:22 |
| *** dolphm has joined #openstack-dev | 12:22 | |
| *** lorin1 has joined #openstack-dev | 12:28 | |
| *** cdub_ has joined #openstack-dev | 12:29 | |
| *** cdub has quit IRC | 12:29 | |
| *** rkukura has joined #openstack-dev | 12:34 | |
| viraptor | there's a nice speedup patch with 2 +1s - maybe someone wants to have a look / approve? ;) https://review.openstack.org/1225 | 12:36 |
| *** dolphm has quit IRC | 12:45 | |
| *** dprince has joined #openstack-dev | 12:47 | |
| *** rbasak has joined #openstack-dev | 12:56 | |
| *** rbasak has joined #openstack-dev | 12:56 | |
| *** lorin1 has quit IRC | 12:58 | |
| *** dolphm has joined #openstack-dev | 13:00 | |
| *** lorin1 has joined #openstack-dev | 13:04 | |
| *** lorin1 has quit IRC | 13:04 | |
| *** rkukura has left #openstack-dev | 13:04 | |
| *** Oneiroi has quit IRC | 13:11 | |
| *** rohitk has quit IRC | 13:12 | |
| *** bcwaldon has joined #openstack-dev | 13:16 | |
| *** beekhof_muc has joined #openstack-dev | 13:17 | |
| *** rkukura has joined #openstack-dev | 13:19 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: exception.KeypairNotFound usage correction https://review.openstack.org/1290 | 13:23 |
| *** rods has joined #openstack-dev | 13:27 | |
| *** mmetheny has quit IRC | 13:28 | |
| *** mmetheny_ has joined #openstack-dev | 13:28 | |
| *** dolphm has quit IRC | 13:32 | |
| openstackgerrit | Sandy Walsh proposed a change to openstack/nova: Added a new PickledScheduler which takes its configuration from a remote web server. https://review.openstack.org/1291 | 13:33 |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Fixing immediate delete after boot on Libvirt https://review.openstack.org/1236 | 13:37 |
| openstackgerrit | Sandy Walsh proposed a change to openstack/nova: Added a new PickledScheduler which takes its configuration from a remote web server. https://review.openstack.org/1291 | 13:38 |
| *** dolphm has joined #openstack-dev | 13:39 | |
| openstackgerrit | Sandy Walsh proposed a change to openstack/nova: This branch adds the ability to define a local JSON file that will hold configuration information that the scheduler can use. https://review.openstack.org/1291 | 13:39 |
| zykes- | annegentle: you around ? | 13:40 |
| *** rohitk has joined #openstack-dev | 13:46 | |
| *** dolphm has quit IRC | 13:46 | |
| *** dolphm has joined #openstack-dev | 13:48 | |
| *** rbasak has quit IRC | 13:56 | |
| openstackgerrit | tom-hancock proposed a change to openstack/glance: Fix two issues with checked_iter https://review.openstack.org/1289 | 14:00 |
| *** rbasak has joined #openstack-dev | 14:01 | |
| *** pringles_ has quit IRC | 14:01 | |
| *** zns has joined #openstack-dev | 14:06 | |
| openstackgerrit | Sandy Walsh proposed a change to openstack/nova: Add auto-reloading JSON config file support to scheduler. https://review.openstack.org/1291 | 14:08 |
| *** mnour has quit IRC | 14:12 | |
| *** tryggvil_ has joined #openstack-dev | 14:14 | |
| *** joesavak has joined #openstack-dev | 14:16 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: X.509 client authentication with Keystone. Implements blueprint 2-way-ssl https://review.openstack.org/1038 | 14:18 |
| *** lorin1 has joined #openstack-dev | 14:20 | |
| *** jsavak has joined #openstack-dev | 14:22 | |
| *** joesavak has quit IRC | 14:26 | |
| *** mattray has joined #openstack-dev | 14:28 | |
| *** zns has quit IRC | 14:32 | |
| zykes- | would something like http://etherpad.openstack.org/HkEvt4crw9 < be interesting? | 14:32 |
| soren | Yes. | 14:36 |
| zykes- | i think it's cool that you could get a "service" record for a XaaS service instance pointing to the XaaS vm instance automagically when going to the XaaS service for example... | 14:37 |
| *** dubsquared has joined #openstack-dev | 14:38 | |
| *** danwent has quit IRC | 14:38 | |
| *** zns has joined #openstack-dev | 14:40 | |
| zykes- | soren: is there any concept of a "service" or utility vm in openstack ? | 14:40 |
| *** rnirmal has joined #openstack-dev | 14:41 | |
| *** cp16net has joined #openstack-dev | 14:42 | |
| soren | zykes-: There's the cloudpipe one. | 14:43 |
| zykes- | soren: how's that managed in the sence of nova? | 14:43 |
| soren | zykes-: Don't really know. I've never used it. | 14:43 |
| soren | So "magic and pixie dust". | 14:44 |
| zykes- | ;) | 14:44 |
| zykes- | i guess i can look at reddwarf and atlas | 14:44 |
| rnirmal | zykes-: reddwarf is on top of nova... so basically it's a forked version of nova, with very minor changes and the new additions specific to reddwarf | 14:46 |
| *** code_franco has joined #openstack-dev | 14:47 | |
| zykes- | rnirmal: doesn't reddwarf run inside nova? | 14:47 |
| *** clayg_ has joined #openstack-dev | 14:47 | |
| zykes- | i mean, like a vm controlled by nova | 14:47 |
| rnirmal | yes it is | 14:47 |
| zykes- | same thing this should be | 14:47 |
| rnirmal | but it has it's own api and a guest agent | 14:47 |
| rnirmal | zykes-: I guess I just jumped in not knowing what you wanted | 14:48 |
| rnirmal | I thought you were asking for a skeleton project | 14:48 |
| zykes- | rnirmal: guest agent meaning ? | 14:48 |
| rnirmal | it's a another service that runs inside the vm | 14:48 |
| zykes- | rnirmal: i'm using the atlas / rd as a skeleton for the service yeah | 14:48 |
| rnirmal | to manage the database and such | 14:48 |
| zykes- | as in you have a api that's available to other users and you have dns servers running in instances | 14:49 |
| rnirmal | essentially its a managed vm | 14:49 |
| bcwaldon | mtaylor: marco | 14:49 |
| rnirmal | zykes-: yes there's a api and database servers in the instances | 14:49 |
| rnirmal | not dns | 14:50 |
| rnirmal | database server per instance | 14:50 |
| openstackjenkins | Project nova-docs build #1,555: SUCCESS in 4 min 39 sec: https://jenkins.openstack.org/job/nova-docs/1555/ | 14:50 |
| openstackjenkins | stanislaw.pitucha: Enable admin access to EC2 API server | 14:50 |
| zykes- | rnirmal: you have dbaas vm instance, in that vm instance you have a db instance pr db ? | 14:51 |
| rnirmal | zykes-: a single mysql instance per dbaas vm | 14:52 |
| zykes- | rnirmal: what's openvz for then ? | 14:52 |
| zykes- | i thought that was to be able to run multiple databases pr vm ? | 14:52 |
| rnirmal | openvz is the hypervisor in this case... a container based one | 14:52 |
| rnirmal | we are not using, kvm or xenserver | 14:53 |
| zykes- | yeah, but within nova, you have 1 vm instance on kvm or whatever hypervisor you want, within that vm you have multiple openvz instances where the mysql processes runs ? | 14:53 |
| openstackgerrit | Dan Prince proposed a change to openstack/openstack-chef: Add boolean flag for allow_ec2_admin_api. https://review.openstack.org/1292 | 14:54 |
| rnirmal | zykes-: no... openvz is the hypervisor we are using with nova. so nova and the virt layer is openvz... the driver for that is being proposed into nova | 14:54 |
| rnirmal | so openvz is a replacement for kvm in this case | 14:55 |
| zykes- | rnirmal: so you run this on a physical box or ? | 14:55 |
| *** crobinso has joined #openstack-dev | 14:55 | |
| rnirmal | zykes-: openvz runs on the physical compute node where the compute service is | 14:56 |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Correcting libvirt tests that were failing https://review.openstack.org/1293 | 14:56 |
| zykes- | ah | 14:56 |
| rnirmal | the openvz containers are created on the physical node | 14:56 |
| zykes- | i thought you did openvz within a vm node | 14:56 |
| rnirmal | nope | 14:57 |
| *** rbasak has quit IRC | 14:57 | |
| annegentle | Jenkins folks (mtaylor, jeblair): does anyone know why Jenkins can't see that the stable/diablo branch of openstack-manuals changed based on the configuration in https://jenkins.openstack.org/view/Openstack-manuals/job/openstack-admin-manuals-diablo/configure ? Do I need a different build trigger when it's on a different branch than master? | 14:57 |
| zykes- | would it be a bad idea though rnirmal ? | 14:57 |
| *** zns has quit IRC | 14:59 | |
| rnirmal | well not in the managed environment, the user won't have access to the actual instance... just the mysql service within it | 14:59 |
| *** adjohn has joined #openstack-dev | 15:00 | |
| zykes- | yeah, but won't it be the same if you have the db server instance spawn up like cloudpipe? | 15:00 |
| rnirmal | zykes-: I haven't looked at cloudpipe, so I'm not sure | 15:01 |
| rnirmal | ok looks like it's adding some extra payload and starting an instance, so yeah in a way sort of, but there's a whole lot of other things we do | 15:02 |
| zykes- | rnirmal: i mean, about spawning a say x gb big vm inside of nova and have it with openvz so you can "plug" it into multiple tenant networks and or others? | 15:03 |
| zykes- | like citrix does with the "vpx" concept | 15:03 |
| *** zns has joined #openstack-dev | 15:05 | |
| *** dragondm has joined #openstack-dev | 15:05 | |
| *** clayg has quit IRC | 15:06 | |
| *** Gordonz has joined #openstack-dev | 15:06 | |
| rnirmal | zykes-: so like a virtual appliance? | 15:06 |
| *** clayg_ is now known as clayg | 15:06 | |
| *** clayg_ has joined #openstack-dev | 15:06 | |
| *** jdg has joined #openstack-dev | 15:07 | |
| jdg | rnirmal: Morning! | 15:09 |
| *** f-yang has joined #openstack-dev | 15:09 | |
| vishy | kpepple: what if you have instances in history that are using that flavor? The usage calculations will get lost if you delete them. I don't like that this is the one and only place that nova actually permanently deletes stuff from the db | 15:09 |
| *** fred_ has joined #openstack-dev | 15:09 | |
| *** fred_ is now known as Guest89527 | 15:09 | |
| *** f-yang has quit IRC | 15:13 | |
| *** Guest89527 has quit IRC | 15:13 | |
| *** f-yang has joined #openstack-dev | 15:14 | |
| kpepple | vishy: it was a usability compromise for easy installation and private clouds. for easy installation, the migration for this change creates the amazon ec2 defnitions as the defaults -- cutting a possibly confusing step out of the install process (which was difficult enough at the times). for private clouds, many of them want to start fresh with just their numbers from 1. | 15:15 |
| kpepple | vishy: having said that, i am not against purging the --purge option | 15:15 |
| kpepple | vishy: but people will howl :) | 15:15 |
| vishy | kpepple: i am not that unhappy with using purge as a first step | 15:15 |
| vishy | as in when you are installing the cloud | 15:15 |
| vishy | but using it later seems like a bad idea | 15:15 |
| vishy | this is all coming up because someone added code in purge to delete instance_type_extra specs | 15:16 |
| vishy | and accidentally broke it | 15:16 |
| vishy | and it made me realize how non-obvious the code and the ideas are | 15:16 |
| vishy | because no other part of the code does it. | 15:16 |
| kpepple | vishy: agreed on the non-obviousness part | 15:17 |
| kpepple | vishy: we could also just not set default flavors | 15:18 |
| vishy | I don't understand fully the problem with flavor_id and instance_name being non-unique? | 15:18 |
| vishy | can't you just change your code to check for deleted=0 as well? | 15:18 |
| kpepple | we key our billing and rating off the flavor id -- it's our offer code | 15:18 |
| vishy | flavor_id should be unique across deleted=0 | 15:18 |
| *** jdg_ has joined #openstack-dev | 15:18 | |
| *** jdg has quit IRC | 15:19 | |
| kpepple | oh, sorry, i get it now. yeah, names and/or flavor id wouldn't have to be unique there and we could leave it to operator choice | 15:20 |
| vishy | i still think that they should be unique for non-deleted rows | 15:20 |
| kpepple | yes | 15:20 |
| vishy | the unique constraint breaks things when we try to have deletd=1 | 15:20 |
| vishy | ok cool | 15:21 |
| *** rbasak has joined #openstack-dev | 15:23 | |
| rnirmal | jdg_: you back? | 15:23 |
| jdg_ | rnimral: :) Yep | 15:23 |
| f-yang | sandywalsh: ??? | 15:24 |
| sandywalsh | f-yang, hey! | 15:24 |
| *** Gordonz has quit IRC | 15:25 | |
| *** Gordonz has joined #openstack-dev | 15:25 | |
| f-yang | sandywalsh: I saw flatten scheduler got merged, but it still missing ZoneManager object passes to filter_hosts | 15:25 |
| *** adjohn has quit IRC | 15:27 | |
| *** dolphm has quit IRC | 15:27 | |
| sandywalsh | f-yang, there is a new object call HostInfo that gets passed in. It has all the services as ReadOnly dicts in it. You can add your "virtual" service to HostInfo just like the others | 15:28 |
| f-yang | sandywalsh: ok, let me check it out, thX | 15:28 |
| *** lorin1 has quit IRC | 15:29 | |
| *** lorin1 has joined #openstack-dev | 15:29 | |
| *** dolphm has joined #openstack-dev | 15:30 | |
| *** dtroyer has quit IRC | 15:33 | |
| *** cdub_ has quit IRC | 15:33 | |
| *** mnour has joined #openstack-dev | 15:34 | |
| openstackjenkins | Project nova-docs build #1,556: SUCCESS in 4 min 30 sec: https://jenkins.openstack.org/job/nova-docs/1556/ | 15:35 |
| openstackjenkins | ewan.mellor: Bug #884534: nova-ajax-console-proxy crashes on shutdown | 15:35 |
| uvirtbot | Launchpad bug 884534 in nova "nova-ajax-console-proxy crashes on shutdown" [Undecided,Fix committed] https://launchpad.net/bugs/884534 | 15:35 |
| *** beekhof_muc has quit IRC | 15:38 | |
| zykes- | rnirmal yeah | 15:39 |
| zykes- | what you think vishy ? | 15:39 |
| *** rbasak has quit IRC | 15:40 | |
| vishy | about? | 15:40 |
| zykes- | would something like http://etherpad.openstack.org/HkEvt4crw9 < be interesting? | 15:41 |
| *** mnour has quit IRC | 15:42 | |
| *** lorin1 has quit IRC | 15:45 | |
| *** lorin1 has joined #openstack-dev | 15:46 | |
| *** yogirackspace has joined #openstack-dev | 15:47 | |
| *** zns has quit IRC | 15:51 | |
| *** dtroyer has joined #openstack-dev | 15:57 | |
| openstackgerrit | Nikolay Sokolov proposed a change to openstack/nova: Ported NetworkAdmin extention from openstackx. Migrated from bazaar. https://review.openstack.org/801 | 15:57 |
| openstackgerrit | Nikolay Sokolov proposed a change to openstack/nova: Ported NetworkAdmin extention from openstackx. Migrated from bazaar. https://review.openstack.org/801 | 15:59 |
| *** zaitcev has joined #openstack-dev | 16:00 | |
| f-yang | sandywalsh: ??? | 16:01 |
| f-yang | sandywalsh: we still need zoneMgr besides HostInfo - We cache some data within ZoneMgr and servicees, which will be needed by filter_hosts(0 to recheck for consitency, since the filter_hosts(0 can only see data which is ~1 minutes old | 16:03 |
| f-yang | that is why we would need go back to services in ZoneMgr again for checking | 16:04 |
| *** jdg_ has quit IRC | 16:04 | |
| sandywalsh | f-yang, where does your data live in zonemanager? | 16:05 |
| openstackgerrit | Joe Heck proposed a change to openstack/keystone: adding docs to test classes, updating run_tests.sh to match reality adding debug middleware factory adding docs on enabling debug middleware https://review.openstack.org/1278 | 16:06 |
| rnirmal | zykes-: did you look at this blueprint https://blueprints.launchpad.net/nova/+spec/public-and-private-dns | 16:06 |
| f-yang | Sandywalsh: pls look at https://review.openstack.org/#patch,unified,675,2,nova/scheduler/filters/json_filter_integrity.py line#96, though the code is old | 16:06 |
| sandywalsh | f-yang, I thought it was in zm.service_states['virtual'] = {xxx} | 16:06 |
| f-yang | sandywalsh: the above code relatively show the usage more than service_state{} | 16:07 |
| *** reed has joined #openstack-dev | 16:09 | |
| f-yang | sandywalsh: line#147 of https://review.openstack.org/#patch,unified,675,2,nova/scheduler/manager_integrity.py will be the zoneMgr supported functions | 16:10 |
| sandywalsh | f-yang, ok, so line 85 can be handled if you have "trust_state" added to HostInfo (just like compute/volume/network) ... but I'm not sure if we should be attempted to update the zonemanager in host_filter | 16:11 |
| sandywalsh | (uh, filter_hosts) | 16:11 |
| openstackgerrit | garyk proposed a change to openstack/nova: Blueprint - lasterror https://review.openstack.org/1288 | 16:11 |
| sandywalsh | f-yang, perhaps you need a hook at the end of the DistributedScheduler._schedule() method to do some other post-processing on the selected hosts? | 16:12 |
| sandywalsh | (you could derive from DistributedScheduler, override _schedule(), call the base-class first and do your work afterwards? | 16:13 |
| *** jog0 has joined #openstack-dev | 16:13 | |
| f-yang | sandywalsh: need to digest .... | 16:13 |
| *** negronjl_mobile has joined #openstack-dev | 16:13 | |
| *** reidrac has quit IRC | 16:14 | |
| *** zns has joined #openstack-dev | 16:15 | |
| f-yang | sandywalsh: immediate response is that is to create yet another scheduler, rather than to add on a new filter | 16:15 |
| zykes- | uhm, rnirmal but isn't dhcp i wonder handled by netstack and not nova? vishy ? | 16:22 |
| rnirmal | zykes-: not sure... that one is a old blueprint. | 16:22 |
| rnirmal | zykes-: haven't kept up with all the netstack changes | 16:23 |
| *** pringles_ has joined #openstack-dev | 16:26 | |
| *** negronjl_mobile has quit IRC | 16:29 | |
| vishy | handled by nova | 16:31 |
| vishy | atm | 16:31 |
| *** masom has joined #openstack-dev | 16:39 | |
| *** jdurgin has joined #openstack-dev | 16:39 | |
| *** jog0 has quit IRC | 16:43 | |
| *** jog0 has joined #openstack-dev | 16:43 | |
| * pixelbeat is away: Away | 16:57 | |
| jog0 | How do I change a commit message on gerrit when there is no code change ? | 17:02 |
| *** reed has quit IRC | 17:05 | |
| *** mnour has joined #openstack-dev | 17:06 | |
| *** rohitk has quit IRC | 17:12 | |
| *** rods has quit IRC | 17:12 | |
| viraptor | jog0: isn't the commit accepted again after you do commit --amend and rewrite the message? | 17:13 |
| *** dolphm has quit IRC | 17:16 | |
| *** naehring has quit IRC | 17:17 | |
| openstackgerrit | joe-gordon0 proposed a change to openstack/nova: Partial fix for bug 817107 https://review.openstack.org/1245 | 17:18 |
| uvirtbot | Launchpad bug 817107 in nova "Exceptions should be united and exception.wrap_exception should be updated or removed" [Low,Triaged] https://launchpad.net/bugs/817107 | 17:18 |
| *** gyee has joined #openstack-dev | 17:19 | |
| *** jdurgin has quit IRC | 17:19 | |
| jog0 | viraptor: it was a conflict issue, thanks | 17:19 |
| *** stuntmachine has joined #openstack-dev | 17:20 | |
| *** rods has joined #openstack-dev | 17:20 | |
| viraptor | does anyone know where can I find a list of ssh commands for gerrit? | 17:20 |
| *** Gordonz has quit IRC | 17:20 | |
| openstackjenkins | Project nova-docs build #1,557: SUCCESS in 4 min 23 sec: https://jenkins.openstack.org/job/nova-docs/1557/ | 17:20 |
| openstackjenkins | brian.waldon: exception.KeypairNotFound usage correction | 17:20 |
| *** rohitk has joined #openstack-dev | 17:21 | |
| *** gyee has quit IRC | 17:21 | |
| *** lorin1 has quit IRC | 17:26 | |
| openstackgerrit | Stuart McLaren proposed a change to openstack/glance: Adds option to encrypt 'location' metadata. https://review.openstack.org/1189 | 17:27 |
| *** mnour has quit IRC | 17:28 | |
| *** novas0x2a|laptop has joined #openstack-dev | 17:29 | |
| *** tryggvil_ has quit IRC | 17:30 | |
| *** tryggvil_ has joined #openstack-dev | 17:30 | |
| *** lorin1 has joined #openstack-dev | 17:32 | |
| *** joesavak has joined #openstack-dev | 17:32 | |
| bcwaldon | mtaylor: | 17:34 |
| *** Gordonz has joined #openstack-dev | 17:34 | |
| openstackjenkins | Project nova-docs build #1,558: SUCCESS in 4 min 23 sec: https://jenkins.openstack.org/job/nova-docs/1558/ | 17:35 |
| openstackjenkins | Johannes Erdfelt: Log original dropped exception when a new exception occurs | 17:35 |
| *** jsavak has quit IRC | 17:36 | |
| *** jsavak has joined #openstack-dev | 17:37 | |
| *** bengrue has quit IRC | 17:37 | |
| *** joesavak has quit IRC | 17:38 | |
| *** novas0x2a|lapto1 has joined #openstack-dev | 17:38 | |
| *** jdurgin has joined #openstack-dev | 17:38 | |
| *** bengrue has joined #openstack-dev | 17:38 | |
| *** novas0x2a|laptop has quit IRC | 17:39 | |
| *** novas0x2a|laptop has joined #openstack-dev | 17:43 | |
| *** novas0x2a|lapto1 has quit IRC | 17:45 | |
| *** bengrue has quit IRC | 17:47 | |
| *** gyee has joined #openstack-dev | 17:47 | |
| *** novas0x2a|laptop has quit IRC | 17:47 | |
| openstackgerrit | joe-gordon0 proposed a change to openstack/nova: fix bug 816630 https://review.openstack.org/1237 | 17:52 |
| uvirtbot | openstackgerrit: Error: Could not parse data returned by Launchpad: The read operation timed out | 17:52 |
| *** dprince has quit IRC | 17:55 | |
| *** reed has joined #openstack-dev | 17:55 | |
| *** jk0 has quit IRC | 17:56 | |
| *** pyhole has quit IRC | 17:56 | |
| *** jk0 has joined #openstack-dev | 18:00 | |
| *** jk0 has joined #openstack-dev | 18:00 | |
| *** ChanServ sets mode: +v jk0 | 18:00 | |
| *** adjohn has joined #openstack-dev | 18:01 | |
| *** dolphm has joined #openstack-dev | 18:03 | |
| *** dolphm_ has joined #openstack-dev | 18:05 | |
| *** lorin1 has quit IRC | 18:06 | |
| *** dolphm has quit IRC | 18:07 | |
| *** gyee has quit IRC | 18:09 | |
| *** gyee has joined #openstack-dev | 18:09 | |
| *** seats has joined #openstack-dev | 18:10 | |
| *** gyee has joined #openstack-dev | 18:12 | |
| *** masom has quit IRC | 18:13 | |
| *** jk0 has quit IRC | 18:17 | |
| *** lorin1 has joined #openstack-dev | 18:20 | |
| openstackgerrit | Brian Lamar proposed a change to openstack/openstack-integration-tests: Tests were not passing for test_servers in Kong tests. Kong was expecting too much information back from the POST. https://review.openstack.org/1296 | 18:23 |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Gracefully handle Xen resize failure https://review.openstack.org/1297 | 18:24 |
| *** dolphm_ has quit IRC | 18:30 | |
| *** dolphm has joined #openstack-dev | 18:31 | |
| *** cdub has joined #openstack-dev | 18:40 | |
| *** darraghb has quit IRC | 18:42 | |
| zykes- | vishy: how would you think would be the best way to do it ? | 18:42 |
| *** pyhole has joined #openstack-dev | 18:42 | |
| *** jk0 has joined #openstack-dev | 18:42 | |
| *** jk0 has joined #openstack-dev | 18:42 | |
| *** ChanServ sets mode: +v jk0 | 18:42 | |
| *** bengrue has joined #openstack-dev | 18:44 | |
| vishy | zykes-: no idea. :) | 18:45 |
| zykes- | guess i'll give a shot at Dan | 18:47 |
| *** bcwaldon has quit IRC | 18:47 | |
| vishy | automatic dns would be awesome, I haven't thought about how to implement it thtough | 18:48 |
| *** novas0x2a|laptop has joined #openstack-dev | 18:48 | |
| *** jk0 has quit IRC | 18:48 | |
| *** pyhole has quit IRC | 18:48 | |
| openstackgerrit | Gabe Westmaas proposed a change to openstack/nova: Don't update database before to resize https://review.openstack.org/1303 | 18:52 |
| *** bcwaldon has joined #openstack-dev | 18:52 | |
| *** zns has quit IRC | 18:54 | |
| *** Gordonz_ has joined #openstack-dev | 18:54 | |
| *** Gordonz_ has quit IRC | 18:55 | |
| *** Gordonz_ has joined #openstack-dev | 18:55 | |
| *** pyhole has joined #openstack-dev | 18:55 | |
| *** jk0 has joined #openstack-dev | 18:56 | |
| *** jk0 has joined #openstack-dev | 18:56 | |
| *** ChanServ sets mode: +v jk0 | 18:56 | |
| *** novas0x2a|laptop has quit IRC | 18:57 | |
| *** Gordonz has quit IRC | 18:58 | |
| *** zns has joined #openstack-dev | 18:59 | |
| openstackgerrit | Gabe Westmaas proposed a change to openstack/nova: Don't update database before to resize https://review.openstack.org/1303 | 19:01 |
| *** dolphm has quit IRC | 19:02 | |
| *** reed has quit IRC | 19:02 | |
| *** stuntmachine has quit IRC | 19:03 | |
| *** adjohn has quit IRC | 19:05 | |
| *** adjohn has joined #openstack-dev | 19:05 | |
| *** pyhole has quit IRC | 19:08 | |
| f-yang | sandywalsh: question again?! | 19:09 |
| *** novas0x2a|laptop has joined #openstack-dev | 19:09 | |
| f-yang | sandywalsh: HostInfo() only picks up compute+volume+network caps, am I correct? | 19:10 |
| openstackjenkins | Project nova-docs build #1,559: SUCCESS in 4 min 25 sec: https://jenkins.openstack.org/job/nova-docs/1559/ | 19:10 |
| openstackjenkins | renuka.apte: Adds documentation for Xen Storage Manager | 19:10 |
| *** mdomsch has joined #openstack-dev | 19:10 | |
| *** lorin1 has quit IRC | 19:10 | |
| *** bcwaldon_ has joined #openstack-dev | 19:11 | |
| *** bcwaldon has quit IRC | 19:11 | |
| *** danwent has joined #openstack-dev | 19:12 | |
| *** reed has joined #openstack-dev | 19:14 | |
| *** stuntmachine has joined #openstack-dev | 19:14 | |
| *** reed has quit IRC | 19:17 | |
| *** bcwaldon_ has quit IRC | 19:17 | |
| *** bcwaldon has joined #openstack-dev | 19:17 | |
| *** lorin1 has joined #openstack-dev | 19:17 | |
| *** lorin1 has quit IRC | 19:18 | |
| *** lorin1 has joined #openstack-dev | 19:18 | |
| *** lorin1 has joined #openstack-dev | 19:19 | |
| *** jk0 has quit IRC | 19:23 | |
| sandywalsh | f-yang, correct, you'll need to add your 'trusted' block | 19:27 |
| *** viraptor has quit IRC | 19:28 | |
| f-yang | sandywalsh: how am I to add it? overwirt HostInfo() ? | 19:29 |
| sandywalsh | f-yang, is your branch going into nova core or is it an optional component? | 19:31 |
| f-yang | sandywalsh: Nova core | 19:32 |
| sandywalsh | f-yang, ok, then just add it to the HostInfo object directly | 19:33 |
| sandywalsh | (modify HostInfo) | 19:33 |
| f-yang | sandywalsh: ok, thX! but question though, what will be the way if it is optional component ? | 19:34 |
| f-yang | sandywalsh: there seems no easy or clean way to do it for other "block" | 19:35 |
| sandywalsh | f-yang, then we need a way for you to hook in your own ZoneManager | 19:35 |
| sandywalsh | f-yang, we may need to do that anyway for something else going on | 19:35 |
| f-yang | sandywalsh: should I submit a patch on HostInfo() 1st? | 19:36 |
| f-yang | sandy: I didn't catch the purpose of only hardcode those 3 components?! | 19:37 |
| sandywalsh | f-yang, I wanted to lessen the dependency on dict keys | 19:38 |
| *** tryggvil_ has quit IRC | 19:38 | |
| sandywalsh | (I should say having to know what keys are important) | 19:38 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Bug #884930 Support/Remove additional calls for for Tenant. https://review.openstack.org/1281 | 19:39 |
| uvirtbot | Launchpad bug 884930 in keystone "Support/Remove additional calls for for Tenant." [Medium,Confirmed] https://launchpad.net/bugs/884930 | 19:39 |
| f-yang | sandy, (???) not quite nderstand | 19:39 |
| *** yogirackspace has left #openstack-dev | 19:40 | |
| *** reed has joined #openstack-dev | 19:43 | |
| *** pyhole has joined #openstack-dev | 19:44 | |
| *** jk0 has joined #openstack-dev | 19:44 | |
| *** jk0 has quit IRC | 19:44 | |
| *** jk0 has joined #openstack-dev | 19:44 | |
| *** ChanServ sets mode: +v jk0 | 19:44 | |
| *** adjohn has quit IRC | 19:51 | |
| *** cdub has quit IRC | 19:53 | |
| f-yang | sandywalsh: I will 1st submit patch for both HostInfo() and json_filter._parse_string()?? though I still need to create my own json_filter() and derived _scheduler(). Is this ok to you? | 19:55 |
| *** dolphm has joined #openstack-dev | 19:55 | |
| sandywalsh | f-yang, sure, we'll give it a look to make sure we're saying the same thing :) | 19:56 |
| openstackgerrit | A change to openstack/nova has been rejected: Fix deletion of instances without a domain. https://review.openstack.org/968 | 19:56 |
| f-yang | ok! | 19:56 |
| openstackjenkins | Project nova-docs build #1,560: SUCCESS in 4 min 26 sec: https://jenkins.openstack.org/job/nova-docs/1560/ | 20:00 |
| openstackjenkins | brian.waldon: Fixing immediate delete after boot on Libvirt | 20:00 |
| *** zns has quit IRC | 20:00 | |
| openstackgerrit | Gaurav Gupta proposed a change to openstack/nova: Added support for creating nova volume snapshots using OS API. https://review.openstack.org/1202 | 20:01 |
| *** jakedahn has quit IRC | 20:03 | |
| *** jakedahn has joined #openstack-dev | 20:03 | |
| *** jk0 has quit IRC | 20:06 | |
| openstackgerrit | A change to openstack/nova has been rejected: Update the EC2 API versions. https://review.openstack.org/1093 | 20:07 |
| *** zns has joined #openstack-dev | 20:12 | |
| *** gyee has quit IRC | 20:13 | |
| *** jk0 has joined #openstack-dev | 20:16 | |
| *** jk0 has joined #openstack-dev | 20:16 | |
| *** ChanServ sets mode: +v jk0 | 20:16 | |
| *** jk0 has quit IRC | 20:17 | |
| *** dolphm has quit IRC | 20:17 | |
| *** jk0 has joined #openstack-dev | 20:17 | |
| *** jk0 has joined #openstack-dev | 20:17 | |
| *** ChanServ sets mode: +v jk0 | 20:17 | |
| *** zns has quit IRC | 20:19 | |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Bug #884930 Support/Remove additional calls for for Tenant. - Supported call to get users for a tenant for a specific role. - Removed calls to get specific role for a user and to get all the roles for a specific tenant as they are not useful. - Fixed LDAP https://review.openstack.org/1281 | 20:21 |
| uvirtbot | Launchpad bug 884930 in keystone "Support/Remove additional calls for for Tenant." [Medium,Confirmed] https://launchpad.net/bugs/884930 | 20:21 |
| *** zns has joined #openstack-dev | 20:22 | |
| *** pixelbeat has quit IRC | 20:22 | |
| *** bcwaldon has quit IRC | 20:23 | |
| *** dtroyer has quit IRC | 20:24 | |
| *** bengrue has quit IRC | 20:29 | |
| openstackgerrit | yunmao proposed a change to openstack/keystone: fixed the output message error on granting user a role https://review.openstack.org/1313 | 20:29 |
| openstackjenkins | Project nova-docs build #1,561: SUCCESS in 4 min 32 sec: https://jenkins.openstack.org/job/nova-docs/1561/ | 20:30 |
| openstackjenkins | brian.waldon: Gracefully handle Xen resize failure | 20:30 |
| *** jakedahn_ has joined #openstack-dev | 20:31 | |
| *** jdg_ has joined #openstack-dev | 20:32 | |
| jdg_ | Does anybody know why when running devstack on a VM (Vbox) eth0 is lost after a reboot of the VM? | 20:33 |
| *** jakedahn has quit IRC | 20:34 | |
| *** jakedahn_ is now known as jakedahn | 20:34 | |
| *** rkukura has quit IRC | 20:34 | |
| zykes- | rnirmal: here still ? | 20:40 |
| zykes- | danwent: around ? | 20:40 |
| openstackjenkins | Project nova-docs build #1,562: SUCCESS in 4 min 41 sec: https://jenkins.openstack.org/job/nova-docs/1562/ | 20:40 |
| openstackjenkins | jogo: fix bug 816630 | 20:40 |
| uvirtbot | openstackjenkins: Error: Could not parse data returned by Launchpad: timed out | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| uvirtbot | openstackjenkins: Error: "did" is not a valid command. | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| uvirtbot | openstackjenkins: Error: "did" is not a valid command. | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| uvirtbot | openstackjenkins: Error: "did" is not a valid command. | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| uvirtbot | openstackjenkins: Error: "did" is not a valid command. | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| uvirtbot | openstackjenkins: Error: "did" is not a valid command. | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'Error:' | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| uvirtbot | openstackjenkins: You've given me 5 invalid commands within the last minute; I'm now ignoring you for 10 minutes. | 20:41 |
| openstackjenkins | uvirtbot did you mean me? Unknown command 'You've' | 20:41 |
| openstackjenkins | Use '!jenkins help' to get help! | 20:41 |
| zykes- | what the hellp :/ | 20:41 |
| *** bengrue has joined #openstack-dev | 20:43 | |
| jdg_ | That was oddd | 20:43 |
| *** Gordonz_ has quit IRC | 20:43 | |
| f-yang | sandywalsh: ??? Not a right way to overwrite _schedule() since it is to duplicate whole logic of _Schedule() only for one hook. The best way is to create a "common hook" right after selected_filter.filter_hosts() in DS. that is, call out to FLAG.PostFilter_hosts if not None. Though Kludge, hope this won't overloading it | 20:43 |
| *** mdomsch has quit IRC | 20:49 | |
| *** jdg_ has quit IRC | 20:52 | |
| *** jdg has joined #openstack-dev | 20:53 | |
| *** jsavak has quit IRC | 20:56 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: Adding an entry for Nexus Extensions https://review.openstack.org/1314 | 20:56 |
| openstackgerrit | Aaron Lee proposed a change to openstack/nova: Move failed instances to error state https://review.openstack.org/1315 | 20:58 |
| *** cp16net has quit IRC | 20:59 | |
| *** zns has quit IRC | 21:00 | |
| *** stuntmachine has quit IRC | 21:01 | |
| *** stuntmachine has joined #openstack-dev | 21:01 | |
| *** gyee has joined #openstack-dev | 21:02 | |
| *** lorin1 has quit IRC | 21:03 | |
| *** lorin1 has joined #openstack-dev | 21:03 | |
| *** adjohn has joined #openstack-dev | 21:04 | |
| *** lorin1 has left #openstack-dev | 21:06 | |
| *** zns has joined #openstack-dev | 21:06 | |
| openstackgerrit | Edgar Magana Perdomo proposed a change to openstack/quantum: Adding an entry for Nexus Extensions https://review.openstack.org/1314 | 21:08 |
| *** stuntmachine has quit IRC | 21:13 | |
| *** zns has quit IRC | 21:17 | |
| *** zns has joined #openstack-dev | 21:21 | |
| *** adjohn has quit IRC | 21:31 | |
| vishy | comstud: ping | 21:35 |
| *** danwent has quit IRC | 21:39 | |
| openstackgerrit | Aaron Lee proposed a change to openstack/nova: Move failed instances to error state https://review.openstack.org/1315 | 21:43 |
| *** sdake has quit IRC | 21:45 | |
| *** bsza-I has joined #openstack-dev | 21:50 | |
| *** bsza has quit IRC | 21:51 | |
| *** bsza-I has quit IRC | 21:55 | |
| *** reed has quit IRC | 22:01 | |
| *** crobinso has quit IRC | 22:03 | |
| comstud | vishy: pong | 22:05 |
| vishy | we have some concurrency issues with kombu | 22:05 |
| vishy | i have a couple questions for ya | 22:05 |
| comstud | hm ok | 22:05 |
| comstud | fire away | 22:05 |
| *** mattray has quit IRC | 22:05 | |
| vishy | 1st: is there any reason why multicall doesn't use the with … as conn: like everything else does? | 22:06 |
| comstud | hm... it hangs onto the connection | 22:06 |
| comstud | until the caller iterates through the results | 22:06 |
| comstud | so... | 22:07 |
| comstud | I think it would add it back into the pool too early? | 22:07 |
| *** markvoelker has quit IRC | 22:08 | |
| vishy | hmm | 22:09 |
| comstud | ie, on exit of the 'with' | 22:09 |
| comstud | which you wouldn't want | 22:09 |
| comstud | because the caller is still using the connection | 22:09 |
| vishy | maybe i don't understand how the with works there | 22:09 |
| vishy | how does it get released? | 22:09 |
| comstud | multicall itself will release it | 22:09 |
| vishy | does the waiter manually release it at the end? | 22:09 |
| comstud | when it gets the last result | 22:09 |
| comstud | and I think I also added a __del__ | 22:09 |
| comstud | just in case | 22:09 |
| comstud | on the multicall object or the ConnectionContext | 22:09 |
| comstud | i'd have to look | 22:10 |
| vishy | ah gotcha | 22:10 |
| vishy | _connection.close | 22:10 |
| comstud | yes | 22:10 |
| vishy | ok | 22:10 |
| comstud | it's a bit tricky | 22:10 |
| vishy | so I've managed to break it by adding some tests that do a whole bunch of calls | 22:10 |
| comstud | i was actively trying to make sure that the connection get back into the pool no matter what | 22:10 |
| comstud | so I think I hae a __del__ on ConnectionContext | 22:10 |
| vishy | and then running three copies of the tests | 22:10 |
| *** dubsquared has quit IRC | 22:10 | |
| comstud | interesting | 22:10 |
| comstud | got a pointer to those? | 22:11 |
| vishy | but it randomly breaks with different error messages | 22:11 |
| vishy | yes, unfortunately i've just totally hosed my vm | 22:11 |
| comstud | haha | 22:11 |
| comstud | i'm going to be out of town tomorrow until Tues | 22:11 |
| comstud | so I prob won't have much time to look until then | 22:12 |
| comstud | i can't think offhand why there would be concurrency issues with what I did | 22:13 |
| comstud | I thought i took extra care to make sure connections didn't get back into the pool until they were done being used | 22:13 |
| vishy | comstud: https://gist.github.com/1337956 | 22:14 |
| vishy | these could be kombu issues | 22:14 |
| vishy | I've seen three different errors | 22:15 |
| comstud | could be.. not sure what would cause it there also | 22:15 |
| comstud | but if this test reproduces it, i'm confident i'll figure it out | 22:15 |
| comstud | applying that now | 22:16 |
| vishy | so i run the test once with -N to create the db | 22:17 |
| vishy | then (./run_tests.sh -N -n test_rpc_kombu &);(./run_tests.sh -N -n test_rpc_kombu &);./run_tests.sh -N -n test_rpc_kombu | 22:17 |
| vishy | to run three at once without creating the db | 22:17 |
| vishy | and it fails almost every time (with different errors) | 22:17 |
| vishy | the last time i seem to have destroyed my ability to connect to rabbit | 22:17 |
| vishy | i can't even run the test anymore | 22:17 |
| vishy | i think i've maxed out my tcp connections to rabbit | 22:17 |
| vishy | looks like there are a whole bunch open | 22:17 |
| comstud | btw, sould just be able to self.flags(fake_rabbit=False) in those tests | 22:17 |
| comstud | should | 22:17 |
| comstud | and +1 on the use of '42' | 22:18 |
| vishy | comstud yeah but i'm setting up stuff in setup | 22:19 |
| vishy | i could have put it there | 22:19 |
| gyee | newbie question: I added my ssh public key to review.openstack.org/#settings,ssh-keys, but keeping getting "Permission denied (publickey)" when running "ssh -T git@github.com" | 22:19 |
| vishy | but whatev | 22:19 |
| *** reed has joined #openstack-dev | 22:19 | |
| vishy | (you probably don't want rabbit password on yours) | 22:19 |
| comstud | yea | 22:19 |
| vishy | comstud: i had xrange set to 10 during my testing | 22:21 |
| comstud | ok | 22:21 |
| vishy | comstud: when i up to 100 it doesn't even complete the test | 22:21 |
| comstud | i see that | 22:21 |
| comstud | i think that's because the pool is too small | 22:21 |
| comstud | although | 22:22 |
| comstud | hm | 22:22 |
| comstud | i guess it should work | 22:22 |
| vishy | comstud: yes | 22:23 |
| vishy | 20 works | 22:23 |
| vishy | 30 doesn't | 22:23 |
| vishy | so the pool doesn't seem to be working properly | 22:23 |
| comstud | alrighty | 22:23 |
| vishy | if it makes you feel any better it is broken in carrot too! | 22:23 |
| comstud | haha | 22:23 |
| comstud | i guess it does a little | 22:24 |
| comstud | hm, looks like self.flags doesn't work here for some reason | 22:24 |
| comstud | had to mod fakes | 22:24 |
| vishy | yeah | 22:25 |
| vishy | so we have two bugs | 22:25 |
| comstud | ok | 22:25 |
| comstud | so tests pass for me | 22:25 |
| comstud | with xrange 10 | 22:25 |
| comstud | upping it | 22:25 |
| comstud | 20 is ok for me | 22:25 |
| vishy | a) connection over pool size breaks | 22:26 |
| vishy | b) random errors from kombu | 22:26 |
| comstud | ok | 22:26 |
| comstud | i'm not getting b) at all | 22:27 |
| comstud | what version of kombu you running? | 22:27 |
| vishy | wow | 22:28 |
| vishy | i can't up pool size | 22:28 |
| vishy | apparently | 22:28 |
| vishy | example: AMQPConnectionException: (503, u"COMMAND_INVALID - second 'channel.open' seen", (20, 10), 'Channel.open') | 22:29 |
| comstud | shoot | 22:29 |
| comstud | what version of kombu? | 22:30 |
| vishy | comstud: are you doing: (./run_tests.sh -N -n test_rpc_kombu &);(./run_tests.sh -N -n test_rpc_kombu &);./run_tests.sh -N -n test_rpc_kombu | 22:30 |
| comstud | i assume 1.04 because that's in pip | 22:30 |
| comstud | oh no, i'm not | 22:30 |
| comstud | got it | 22:30 |
| vishy | no 1.4.3 | 22:30 |
| comstud | hm | 22:30 |
| comstud | ok | 22:30 |
| comstud | let me try the above | 22:30 |
| comstud | i'm at 1.0.4 i believe | 22:30 |
| *** danwent has joined #openstack-dev | 22:31 | |
| comstud | AMQPChannelException: (405, u"RESOURCE_LOCKED - cannot obtain exclusive access to locked queue 'a_direct' in vhost '/'", (50, 10), 'Channel.queue_declare') | 22:31 |
| comstud | you seeing that one too? | 22:31 |
| vishy | AMQPChannelException: (404, u"NOT_FOUND - no exchange 'a_direct' in vhost '/'", (50, 20), 'Channel.queue_bind') | 22:31 |
| vishy | yup got that one | 22:32 |
| comstud | ok | 22:32 |
| comstud | i'm curious why this only happens when running 2 copies of the tests | 22:32 |
| comstud | if it only happens then | 22:33 |
| comstud | actually | 22:33 |
| comstud | does't this try to start 2 consumers on the same exclusive queue? | 22:33 |
| * comstud looks at the tests more closely | 22:34 | |
| comstud | no, i guess exclusive is False | 22:35 |
| comstud | but the 'cannot obtain exclusive access to locked queue' seems rather odd, if exclusive is false. | 22:37 |
| *** dtroyer has joined #openstack-dev | 22:37 | |
| comstud | i can't get any other error here right now, other than the 'locked' one | 22:38 |
| comstud | let me try 1.4.3 | 22:38 |
| comstud | hm, same thing | 22:39 |
| vishy | the errors are very intermittent | 22:40 |
| vishy | comstud: I'm concerned about: File "/usr/local/lib/python2.7/dist-packages/kombu/syn.py", line 49, in __eblocking__ | 22:42 |
| vishy | it goes into greenthread there | 22:42 |
| comstud | Ask mentioned something about that | 22:42 |
| vishy | so there is potential for some state to get messed up | 22:42 |
| comstud | in that bug report | 22:43 |
| comstud | i think he suggested something | 22:43 |
| comstud | might have been in private email | 22:43 |
| comstud | let me dig it out | 22:43 |
| comstud | okay | 22:44 |
| comstud | i know what the locked issue is | 22:45 |
| comstud | it's running the 'test_direct_send_receive' | 22:45 |
| comstud | direct is an exclusive queue | 22:45 |
| comstud | running it twice at same time will cause locked error | 22:45 |
| comstud | that's 'normal' | 22:45 |
| comstud | should never have 2 consumers on same queue name and the test hardcodes 'a_direct' | 22:46 |
| vishy | the exclusive access is for a_direct queue | 22:46 |
| comstud | yeah | 22:46 |
| comstud | this is a direct queue | 22:46 |
| comstud | test_direct_send_receive() | 22:46 |
| vishy | yeah | 22:46 |
| comstud | so | 22:46 |
| vishy | i just noticed taht at the same time | 22:46 |
| vishy | :0 | 22:46 |
| comstud | i'm going to ignore that one | 22:46 |
| comstud | ok :) | 22:46 |
| vishy | agreed | 22:46 |
| comstud | or i'll randomize the queue name | 22:46 |
| vishy | similarly: NOT_FOUND - no exchange 'a_direct' in vhost '/' | 22:47 |
| comstud | yeah | 22:47 |
| vishy | is probably 1 test deleting the queue | 22:47 |
| comstud | yep | 22:47 |
| vishy | before the other is done with it | 22:47 |
| vishy | also not a big deal | 22:47 |
| comstud | yeah, i just i randomized the queue name for that direct test now | 22:49 |
| comstud | i can't recreate any other errors | 22:49 |
| comstud | i'll continue to run | 22:49 |
| vishy | is this one: 503, u"COMMAND_INVALID - second 'channel.open' seen", (20, 10), 'Channel.open' the same? | 22:51 |
| vishy | ok i randomized the name | 22:51 |
| vishy | fixed those two issues | 22:51 |
| vishy | other still appears | 22:51 |
| vishy | comstud: http://paste.openstack.org/show/3068/ | 22:51 |
| comstud | hm | 22:51 |
| comstud | that feels like a concurrency issue | 22:52 |
| comstud | not sure what that's about | 22:52 |
| comstud | you should be able to self.connection.channel() to open new channels IIRC | 22:52 |
| comstud | either way.. pretty sure I close the old channel | 22:53 |
| vishy | i get that mb 1/3 times | 22:53 |
| vishy | maybe even less | 22:53 |
| comstud | i can't get an error at all | 22:54 |
| comstud | i can try running more | 22:54 |
| vishy | not sure if it gets us any closer to debugging the issue where instance create fails | 22:54 |
| vishy | i mean i guess it is two things both trying to create the same channel? | 22:55 |
| *** adjohn has joined #openstack-dev | 22:55 | |
| comstud | no clue | 22:55 |
| comstud | i didn't think there was really anything specific with channels that they could 'collide' | 22:56 |
| comstud | you can open multiple channels on a connection | 22:56 |
| vishy | comstud: https://answers.launchpad.net/nova/+question/177213 | 22:56 |
| vishy | it happens irl too | 22:56 |
| comstud | right now i'm confused about my environment | 22:58 |
| comstud | i'd added a debug msg to impl_kombu.py | 22:58 |
| comstud | and I've removed it now | 22:58 |
| comstud | yet it's still printing somehow | 22:58 |
| comstud | and I cleared the pyc and everything | 22:58 |
| comstud | *shakes head* | 22:59 |
| *** cdub has joined #openstack-dev | 23:00 | |
| vishy | hehe | 23:01 |
| comstud | i'm stumped | 23:01 |
| vishy | comstud: (nova.rpc): TRACE: AMQPChannelException: (404, u"NOT_FOUND - no exchange | 23:01 |
| vishy | '21937a9d485c483da78f2dbc9aeca78e' in vhost '/'", (60, 40), | 23:01 |
| vishy | 'Channel.basic_publish') | 23:01 |
| vishy | that is the really scary one | 23:01 |
| comstud | yeah. | 23:01 |
| comstud | what pyamqplib version do you have? | 23:01 |
| vishy | i'm stiil not sure how the channel can be deleted before the message is sent | 23:01 |
| vishy | 1.0.0 | 23:01 |
| comstud | hm | 23:02 |
| comstud | i think i was using that before | 23:02 |
| comstud | i have 1.0.2 right now | 23:02 |
| comstud | vishy: yeah, exactly | 23:02 |
| comstud | no matter what I do, I can't reproduce this | 23:03 |
| *** andreaf has joined #openstack-dev | 23:04 | |
| vishy | comstud: interesting let us see if updating helps | 23:04 |
| vishy | comstud: ineteesting the NOT_FOUND is happening on close | 23:04 |
| vishy | comstud: not on send | 23:04 |
| vishy | comstud: what happens if you send a message on a direct queue and it is consumed | 23:04 |
| comstud | hm | 23:04 |
| vishy | comstud: is it possible the consumer is closing the channel? | 23:04 |
| vishy | and then the publisher tries to close and fails? | 23:05 |
| comstud | a channel is tied to a connection | 23:05 |
| comstud | it's nothing shared | 23:05 |
| comstud | but.. | 23:05 |
| vishy | sorry | 23:05 |
| comstud | exchange? | 23:05 |
| vishy | the exchange | 23:05 |
| vishy | i mean | 23:05 |
| comstud | that's an interesting thought | 23:05 |
| comstud | but | 23:06 |
| comstud | the AMQP command seemed to be a 'basic_publish' | 23:06 |
| comstud | not a close | 23:06 |
| comstud | your NOT_FOUND above says while doing a Channel.basic_publish | 23:06 |
| *** andreaf has quit IRC | 23:07 | |
| vishy | yeah that one is different | 23:07 |
| comstud | ok | 23:07 |
| vishy | the NOT_FOUND in the bug report is on reset | 23:07 |
| vishy | btw, upgraded to 1.0.2 and still got the 503 | 23:07 |
| comstud | ok | 23:08 |
| *** andreaf has joined #openstack-dev | 23:08 | |
| comstud | i'll try this on another host | 23:08 |
| *** bhall has quit IRC | 23:09 | |
| *** sdake has joined #openstack-dev | 23:09 | |
| *** bhall has joined #openstack-dev | 23:11 | |
| vishy | i'm running in a vm, so it may be some odd bug that only happens when the kernel is locked or something | 23:11 |
| comstud | is your rabbit server on localhost or further away? | 23:11 |
| comstud | so am I | 23:12 |
| comstud | i was running it under nested VMs, actually | 23:12 |
| comstud | fusion -> xenserver -> squeeze | 23:12 |
| comstud | which is faster than any of my other machines | 23:12 |
| comstud | on this mac | 23:12 |
| comstud | i still can't reproduce it on this other host | 23:13 |
| vishy | comstud: 404, u"NOT_FOUND - no exchange that one is actually an easy fix | 23:13 |
| vishy | i still would like to know how the exchange is getting closed early | 23:13 |
| *** pixelbeat has joined #openstack-dev | 23:14 | |
| comstud | i could see there being a race condition if the consumer popped the msg off extremely quickly and closed the channel (so the exchange is removed)... before the sender side closed its tie to the exchange | 23:14 |
| comstud | although it seems kind of dumb that a channel close would care if the exchange still existed or not | 23:15 |
| comstud | i'm not sure why a channel.close() results in a Channel.basic_publish command to the server. | 23:16 |
| comstud | but I don't pretend to know AMQP very well | 23:16 |
| comstud | i'm starting to wonder what's different with our environments | 23:17 |
| comstud | because I can't reproduce this... and I don't think we've seen it at all in our group (even once) | 23:17 |
| *** cdub has quit IRC | 23:17 | |
| comstud | let me find that thing from Ask | 23:18 |
| comstud | for you to try | 23:18 |
| comstud | 'Maybe, you could try replacing kombu.syn.blocking with a dummy version that just | 23:18 |
| comstud | executes the function?' | 23:18 |
| vishy | basic_publish? it doesn't | 23:19 |
| comstud | says that could be relevant difference compared to kombu | 23:19 |
| vishy | comstud: look at the trace here: https://launchpadlibrarian.net/83304253/nova-network.log | 23:19 |
| vishy | it is in close that it fails | 23:19 |
| comstud | vishy: I see Channel.basic_publish in the 404 error | 23:19 |
| comstud | pyamqplib is putting it in the exception | 23:19 |
| *** adjohn has quit IRC | 23:19 | |
| comstud | but i dunno if that string came from rabbit or not | 23:19 |
| vishy | comstud: 125 'auto_delete': True, | 23:20 |
| vishy | in direct consumer | 23:20 |
| comstud | yep | 23:20 |
| vishy | why do we try to delete it on the other end? | 23:20 |
| comstud | Oh | 23:20 |
| comstud | is the publisher trying to delete it? | 23:20 |
| comstud | (also) | 23:20 |
| vishy | well the publisher calls close at the end | 23:21 |
| comstud | looks like it. | 23:21 |
| comstud | i close the channel so I can open a new one... to re-use the same connection | 23:21 |
| *** zns has quit IRC | 23:21 | |
| vishy | (nova.rpc): TRACE: File "/opt/stack/nova/nova/rpc/impl_kombu.py", line 382, in reset | 23:21 |
| vishy | (nova.rpc): TRACE: self.channel.close() | 23:21 |
| *** adjohn has joined #openstack-dev | 23:21 | |
| vishy | so should we try accept around that? | 23:22 |
| comstud | but I see auto_delete=True in DirectPublisher | 23:22 |
| comstud | for some reason, i think that might need to match the consumer | 23:22 |
| comstud | but we could try removing that also | 23:22 |
| comstud | vishy: maybe | 23:22 |
| comstud | although | 23:22 |
| vishy | they are both auto_delete in carrot | 23:22 |
| comstud | i'm concerned that would leave the connection in a broken state | 23:22 |
| comstud | such that if it goes back into the pool for someone else, it'll be broken | 23:23 |
| vishy | it just needs to not try to delete teh exchange twice | 23:23 |
| comstud | i'm not sure that the publisher actually tries to delete it | 23:23 |
| comstud | let me look at docs | 23:23 |
| comstud | yeah | 23:24 |
| comstud | i don't think it does | 23:24 |
| comstud | auto_delete tells the server to delete the exchange when the consumer disconnects | 23:25 |
| comstud | well | 23:25 |
| comstud | to be clear: | 23:25 |
| comstud | http://en.wikipedia.org/wiki/Advanced_Message_Queuing_Protocol#Exchanges | 23:26 |
| comstud | see auto_delete under Exchanges and Queues | 23:26 |
| comstud | it really only applies to consumers | 23:26 |
| comstud | but for whatever reason, a publisher has to specify the setting to match the consumer | 23:26 |
| comstud | IIRC | 23:26 |
| comstud | amqp is kinda screwy if you ask me :) | 23:26 |
| comstud | this is all server side stuff | 23:27 |
| comstud | and I know a publisher disconnecting doesn't delete the queue... because I ran into this in testing | 23:27 |
| vishy | wait but it tries to do something with the exchange | 23:28 |
| comstud | you can declare an exchange/queue as auto_delete=True... push a msg to it | 23:28 |
| comstud | close the publusher | 23:28 |
| comstud | and the queue/exchange will stick around | 23:28 |
| comstud | until a consumer subscribes and disconnects | 23:28 |
| vishy | sure | 23:28 |
| vishy | but the error is on the publisher side | 23:29 |
| comstud | yeah, makes no sense to me | 23:29 |
| comstud | i almost want to blame rabbit for that. | 23:29 |
| *** rnirmal has quit IRC | 23:29 | |
| comstud | because i don't think the client is trying to delete anything | 23:29 |
| vishy | File "/opt/stack/nova/nova/rpc/impl_kombu.py", line 382, in reset | 23:30 |
| vishy | (nova.rpc): TRACE: self.channel.close() | 23:30 |
| comstud | auto_delete happens on the server | 23:30 |
| comstud | that's just closing the amqp channel | 23:30 |
| comstud | essentially that's the same as closing the socket | 23:31 |
| comstud | althouhg it doesn't close the socket | 23:31 |
| vishy | comstud: check out 211 in 211 After sending this method any received met | 23:31 |
| comstud | 211 in where? | 23:31 |
| *** misheska has joined #openstack-dev | 23:32 | |
| vishy | usr/lib/python2.7/dist-packages/amqplib/client_0_8/channel.py | 23:33 |
| vishy | wow | 23:33 |
| vishy | it really looks like _do_close is called twice | 23:34 |
| comstud | immediately after i close the channel, i re-open a new one | 23:34 |
| comstud | essentially that 'resets' things for the next person to grab the connection out of the pool | 23:34 |
| comstud | there def should be no more commands on the channel after closing it | 23:35 |
| vishy | comstud: but it says any other method besides close_ok should be ignored | 23:35 |
| comstud | on the channel | 23:35 |
| comstud | not the connection | 23:35 |
| vishy | the method we get back is : (60, 40), 'Channel.basic_publish') | 23:35 |
| vishy | why isn't it being ignored | 23:36 |
| vishy | ? | 23:36 |
| comstud | hm, true | 23:36 |
| comstud | i see | 23:36 |
| comstud | it's unclear to me where that basic_publish is coming from | 23:37 |
| comstud | it feels like that the server is giving us a 404 when trying to do that command | 23:37 |
| vishy | unless that is a copy paste error form the close method above | 23:37 |
| comstud | ie... amqplib sent the basic_publish | 23:37 |
| *** jdg has quit IRC | 23:38 | |
| comstud | and this is where there might be a concurrency thing | 23:38 |
| vishy | the server sent basic publish instead of close_ok? | 23:38 |
| comstud | if 2 things somehow got ahold of the same connection | 23:38 |
| comstud | it's possible we have 1 greenthread doing a publish | 23:39 |
| comstud | at the same time another is closing | 23:39 |
| vishy | oh interesting | 23:43 |
| *** cdub has joined #openstack-dev | 23:44 | |
| *** novas0x2a|laptop has quit IRC | 23:44 | |
| comstud | i'm not sure i see a point to this kombu.syn.blocking() call for eventlet | 23:44 |
| comstud | it just does a spawn().wait() | 23:45 |
| comstud | if that created a _real_ thread, maybe i'd understand | 23:45 |
| vishy | comstud: https://github.com/pika/pika/issues/84 | 23:46 |
| vishy | comstud: https://github.com/wulczer/pika/commit/2e32388d97bbecdd83e90572e203783ca1b2d01c | 23:47 |
| vishy | i think a close-ok isn not being sent properly sometimes | 23:48 |
| vishy | because the stuff talked about in that issue is exactly what we are seeing | 23:49 |
| comstud | hm, that's pika tho | 23:49 |
| *** andreaf has quit IRC | 23:49 | |
| comstud | isn't that just an alternative to amqplib? | 23:49 |
| vishy | i know but i't is the same issue | 23:49 |
| vishy | getting back weird responses | 23:49 |
| vishy | when you send a close | 23:49 |
| comstud | ah | 23:49 |
| vishy | there is an explanation there of why it happens | 23:50 |
| *** jeffjapan has joined #openstack-dev | 23:50 | |
| comstud | reading further | 23:50 |
| *** andreaf has joined #openstack-dev | 23:50 | |
| comstud | oh | 23:51 |
| comstud | interesting. | 23:51 |
| *** cdub has quit IRC | 23:51 | |
| vishy | (from here on after, every operation the app will attempt using the closed channel will immediately raise an ChannelClosed error with a 404 code) | 23:51 |
| *** code_franco has quit IRC | 23:51 | |
| vishy | so that message may be referring to an earlier call? | 23:52 |
| vishy | i'm still at a loss to why this works with carrot | 23:52 |
| vishy | this might be a lot easier with amqp debugging turned on | 23:53 |
| *** dtroyer_ has joined #openstack-dev | 23:53 | |
| vishy | perhaps i need to reproduce his error with debugging on and request logging active | 23:53 |
| vishy | so i can actually see the different threads | 23:54 |
| vishy | and make it log message_id as well | 23:54 |
| comstud | i wish i could reproduce it | 23:54 |
| comstud | the one thing I can look at is the pool exhaustion thing | 23:54 |
| *** andreaf has quit IRC | 23:55 | |
| *** dtroyer has quit IRC | 23:56 | |
| *** dtroyer_ is now known as dtroyer | 23:56 | |
| *** andreaf has joined #openstack-dev | 23:56 | |
| comstud | ah | 23:56 |
| comstud | i guess i have an explanation for that. | 23:56 |
| comstud | well, for the nested case | 23:58 |
| comstud | hm | 23:58 |
| *** andreaf has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!