| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-07-24 | |||
| 14:17:04 | openstack | Launchpad bug 1705680 in OpenStack Compute (nova) "nova compute does nothing on receiving sighup signal" [Undecided,New] | |
| 14:17:15 | bauzas | if we were inheriting from oslo.service manager, then OK | |
| 14:17:19 | bauzas | but we're not | |
| 14:17:44 | dansmith | we do process sighup for rpc version pins | |
| 14:18:12 | dansmith | bauzas: https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L539 | |
| 14:18:35 | bauzas | dansmith: I saw the hook | |
| 14:18:50 | bauzas | dansmith: but I don't see how we define that hook to be called on a sighup signal | |
| 14:19:03 | dansmith | I think we do get it from oslo.service | |
| 14:19:07 | dansmith | but it's buried pretty deep | |
| 14:19:23 | bauzas | dansmith: that was my assumption | |
| 14:19:24 | bauzas | but | |
| 14:19:29 | bauzas | we don't inherit from it | |
| 14:19:34 | dansmith | we do | |
| 14:19:46 | dansmith | bauzas: https://github.com/openstack/nova/blob/master/nova/service.py#L98-L98 | |
| 14:19:52 | dansmith | bauzas: service is oslo.service there | |
| 14:20:08 | dansmith | L27 | |
| 14:20:11 | bauzas | https://github.com/openstack/nova/blob/master/nova/compute/manager.py#L482 | |
| 14:20:22 | bauzas | manager is nova.manager, right? | |
| 14:20:37 | dansmith | yes, but the service hooks Service.reset() to Manager.reset() | |
| 14:20:47 | dansmith | https://github.com/openstack/nova/blob/master/nova/service.py#L289-L289 | |
| 14:21:00 | bauzas | but https://github.com/openstack/nova/blob/master/nova/manager.py#L91 | |
| 14:21:13 | mriedem | dansmith: the mutable config stuff is in oslo.config docs | |
| 14:21:14 | bauzas | dansmith: ooooh, it's fucking cryptic | |
| 14:21:25 | mriedem | https://docs.openstack.org/oslo.config/latest/reference/mutable.html | |
| 14:21:34 | dansmith | bauzas: not really, managers and services have always had this relationship | |
| 14:21:55 | bauzas | I see | |
| 14:21:59 | dansmith | mriedem: oh right and I think we're writing into our default nova config that log level is mutable or something | |
| 14:22:04 | dansmith | because of that | |
| 14:22:56 | mriedem | i thought it was only 'debug' | |
| 14:23:19 | mriedem | https://review.openstack.org/#/c/280851 | |
| 14:23:39 | mriedem | https://review.openstack.org/#/c/254821/ | |
| 14:23:52 | bauzas | mriedem: I was just looking at the mutable-config series | |
| 14:23:53 | dansmith | mriedem: okay I never saw this, but it completely wasn't working | |
| 14:23:55 | mriedem | so default_log_levels config wouldn't use that | |
| 14:24:17 | dansmith | I think it was debug= they were toggling, but maybe not | |
| 14:24:51 | mriedem | https://docs.openstack.org/nova/latest/sample_config.html | |
| 14:24:51 | mriedem | we only have 4 | |
| 14:24:58 | mriedem | if you search for "Note: This option can be changed without restarting." | |
| 14:25:28 | dansmith | wonder if we're supposed to be hooking our sighup handler to oslo.log somehow? | |
| 14:25:37 | bauzas | mriedem: correct, the mutable-config series was mostly still in progress when lxsli left | |
| 14:25:38 | dansmith | unless it registers its own signal handler quietly | |
| 14:26:41 | bauzas | the point with https://bugs.launchpad.net/nova/+bug/1705680 is that I suspect the sighup to be caught but just the fact that we only reload a very few flags made the reporter thinking it wasn't working | |
| 14:26:42 | openstack | Launchpad bug 1705680 in OpenStack Compute (nova) "nova compute does nothing on receiving sighup signal" [Undecided,New] | |
| 14:26:59 | bauzas | either way, I can call out for details | |
| 14:31:11 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/nova master: Add Neutron port capabilities to devspec in request https://review.openstack.org/451777 | |
| 14:55:54 | openstackgerrit | Eric Fried proposed openstack/nova master: Adopt new pypowervm power_off APIs https://review.openstack.org/476945 | |
| 14:56:05 | openstackgerrit | Sean Dague proposed openstack/nova master: Increase cpu time for image conversion https://review.openstack.org/486642 | |
| 14:59:16 | s-dean | can somebody please confirm that the cells table in nova db is meant to be empty, I have been trying to setup the cells database and every time i run su -s /bin/sh -c "nova-manage db sync" nova , I get the following output ERROR: could not access cell mapping database - has api db been created?, I have been at this for 5 days and same error everytime i reinstall | |
| 14:59:28 | jangutter | sean-k-mooney: are you inline? | |
| 14:59:35 | jangutter | s/inline/online/? | |
| 15:02:48 | sean-k-mooney | yes though i have to drop for meeting in an hour | |
| 15:03:14 | sean-k-mooney | jangutter: ^ | |
| 15:03:52 | jangutter | sean-k-mooney: I split off https://review.openstack.org/#/c/486426/ but I'm not sure I wrote the test right. | |
| 15:05:49 | sean-k-mooney | jangutter: well that is partly a technically question and partly a political one. you added _set_config_VIFHostDevice | |
| 15:06:53 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/os-vif master: Migration from 'ip' commands to pyroute2 https://review.openstack.org/484386 | |
| 15:08:04 | mriedem | s-dean: yes the cells table is for cells v1 only | |
| 15:08:06 | jangutter | sean-k-mooney: yep, and used "unplugin" rather than "plugin" -> other VIF tests that don't go out via an os-vif plugin seem to use that, rather than "plugin" | |
| 15:08:07 | sean-k-mooney | jangutter: technically in test_vif.py you should only assert that the designer was called. and then in https://github.com/openstack/nova/blob/master/nova/tests/unit/virt/libvirt/test_designer.py you should assert that correct xml is generated when you pass in the os_vif_hostdevice | |
| 15:08:25 | mriedem | s-dean: you have a nova_api db yes? | |
| 15:08:33 | mriedem | s-dean: did you run nova-manage api_db sync ? | |
| 15:09:08 | mriedem | s-dean: also https://docs.openstack.org/nova/latest/cells.html and https://docs.openstack.org/ocata/install-guide-ubuntu/nova.html for docs | |
| 15:09:22 | sean-k-mooney | jangutter: but politcally nova has not done it this way else where in the file so you should follow the files convention of have test_vif.py also test desighern fuctionality though really this should be change across the board in that file in a seperate patch | |
| 15:09:37 | jangutter | sean-k-mooney: ah, that makes it a lot finer-grained. | |
| 15:10:57 | sean-k-mooney | jangutter: yes unfutrnetlly that is not how the other unitest have been written so it better to follow the convention in the file though we should proably file a bug to make it more granular | |
| 15:11:35 | sean-k-mooney | jangutter: i think the current patch is fine given the convention that is inplace | |
| 15:11:43 | mriedem | s-dean: it could also mean that the nova.conf that you're running nova-manag db sync against doesn't have the [api_database]connection option set? | |
| 15:12:20 | s-dean | yes I have nova_api database, I have run the commands in order as specified in the documentation, However ERROR: could not access cell mapping database - has api db been created? | |
| 15:12:29 | s-dean | keeps appearing | |
| 15:12:45 | mriedem | s-dean: the command is looking for entries in the cell_mappings table in the nova_api db | |
| 15:12:59 | s-dean | they exist | |
| 15:13:07 | jangutter | sean-k-mooney: thanks, I and that clears up the other questions I also had. I had wondered why there seemed to be two sets of tests in test_vif.py | |
| 15:13:11 | mriedem | or whatever you called it, but it would be defined by the [api_database]connection url in nova.conf | |
| 15:13:13 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Enable cold migration with target host(2/2) https://review.openstack.org/408964 | |
| 15:13:26 | mriedem | s-dean: how do the cell_mappings entries exist if you don't have a nova_api db? | |
| 15:13:33 | mriedem | oh you said you have it | |
| 15:13:34 | stvnoyes | mriedem good morning Matt, when you get some time, please take a look at the updated cinder v3 migrate review. thanks. - https://review.openstack.org/#/c/463987/ | |
| 15:13:37 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: api-ref: Add parameters in cold migrate action https://review.openstack.org/410042 | |
| 15:13:49 | mriedem | s-dean: is [api_database]connection set in nova.conf when running nova-manage db sync? | |
| 15:13:54 | s-dean | I have created it, Is this a potential bug | |
| 15:14:13 | s-dean | yes my connection string is correct | |
| 15:14:40 | mriedem | s-dean: just to be clear, so you have both [database]/connection and [api_database]/connection set in nova.conf? | |
| 15:14:44 | mriedem | and they are different values, yes? | |
| 15:14:45 | s-dean | yes | |
| 15:14:48 | s-dean | yes | |
| 15:15:08 | s-dean | one for nova_api | |
| 15:15:10 | mriedem | ok, and you ran nova-manage api_db sync ? | |
| 15:15:15 | s-dean | and on for nova db | |
| 15:15:19 | s-dean | yes | |
| 15:15:28 | sean-k-mooney | jangutter: i think its a legacy thing. but generally when writhing unit test you should mock any fuction call to other modules in the current project and test only how the output of the fuction is used and test the internals of that fuction in a seperate test in that modules test file | |
| 15:15:43 | mriedem | did you run nova-manage cell_v2 map_cell0 | |
| 15:16:26 | openstackgerrit | Gábor Antal proposed openstack/nova master: Transform HostAPI-set_enabled notifications https://review.openstack.org/482622 | |
| 15:16:27 | sean-k-mooney | jangutter: but in all cases the convention of the file you are extending take presedence over teh convetions of the project as a whole though one would hope they more or less agree | |
| 15:16:36 | mriedem | stephenfin: fyi, docs theme is jacked here https://docs.openstack.org/nova/latest/cells.html | |
| 15:17:06 | s-dean | ERROR: could not access cell mapping database - has api db been created? | |
| 15:17:06 | s-dean | I have run Step 3,4,5,6 at step 6 following error | |
| 15:17:19 | jangutter | sean-k-mooney: yep. No joke "writhing unit test" is pretty accurate. | |
| 15:17:53 | mriedem | s-dean: steps in which doc? the install guide? | |
| 15:17:59 | s-dean | ubuntu 16.04 | |
| 15:18:12 | s-dean | ocata install guide | |
| 15:18:40 | mriedem | s-dean: so you have a nova_cell0 database as well? | |