| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-20 | |||
| 19:33:46 | openstackgerrit | Roman Dobosz proposed openstack/nova-specs master: FPGA resource use cases https://review.openstack.org/318047 | |
| 19:33:59 | mriedem | this isn't the nova ubuntu packaging channel | |
| 19:34:39 | mriedem | coreycb: do you do ubuntu packaging for nova stuff? | |
| 19:35:06 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Remove deprecated baremetal CLIs and APIs https://review.openstack.org/447657 | |
| 19:35:08 | mriedem | jroll: just for you ^ | |
| 19:35:13 | wally | yesh i use apt install nova-placement-api | |
| 19:35:47 | openstackgerrit | Claudiu Belu proposed openstack/nova master: POC: tests: autospecs all the mock.patch usages https://review.openstack.org/447505 | |
| 19:36:01 | coreycb | mriedem, hi, yes | |
| 19:36:09 | mriedem | EmilienM: package mirror issue here? http://logs.openstack.org/22/447522/1/check-tripleo/gate-tripleo-ci-centos-7-ovb-nonha/b8990d9/console.html | |
| 19:36:23 | mriedem | coreycb: are you getting reports of packaging issues with the placement API in ocata? | |
| 19:36:35 | coreycb | mriedem, no, not yet | |
| 19:36:41 | openstack | bugzilla.redhat.com bug 1430540 in openstack-nova "[Ocata] openstack-nova-placement-api-15.0.0-1.el7.noarch missing <Directory> parameter" [Unspecified,New] - Assigned to nlevinki | |
| 19:36:41 | mriedem | like https://bugzilla.redhat.com/show_bug.cgi?id=1430540#c0 | |
| 19:37:29 | mriedem | wally: what does your httpd conf look like for the placement service? | |
| 19:37:39 | wally | one set | |
| 19:38:32 | wally | please see here | |
| 19:38:33 | wally | http://paste.openstack.org/show/603485/ | |
| 19:38:47 | wally | root@controller:/etc/apache2/sites-enabled# vim nova-placement-api.conf | |
| 19:39:10 | wally | this is my apache nova-placement-api.conf located here | |
| 19:39:20 | mriedem | coreycb: is that what you'd expect to see? | |
| 19:40:19 | wally | one question, i cannot do curl http://10.10.10.11:8778 on compute node, not working, but it work for other endpoint | |
| 19:40:20 | openstackgerrit | Anusha Unnam proposed openstack/nova master: Add description to policies to availability_zone https://review.openstack.org/440836 | |
| 19:40:27 | wally | curl http://10.10.10.11:8778 curl: (7) Failed to connect to 10.10.10.11 port 8778: Connection refused | |
| 19:40:29 | coreycb | wally, mriedem: that looks a little different from the site config we ship with the package | |
| 19:41:10 | coreycb | wally, mriedem: here's what the package defines, https://git.launchpad.net/~ubuntu-server-dev/ubuntu/+source/nova/tree/debian/nova-placement-api.conf?h=stable/ocata | |
| 19:41:37 | wally | so i should use this one?> | |
| 19:41:51 | wally | the config, i modified a little bit myself, | |
| 19:42:35 | coreycb | wally, well it's the default shipped by the package so you can use that, or most deployment projects will likely modify that (ie. ansible, puppet, juju etc) | |
| 19:42:59 | wally | ok, let me try to use this default one and try again | |
| 19:43:05 | coreycb | wally, ok | |
| 19:43:20 | cdent | coreycb: as a somewhat related aside: I think the wsgi script should not be in /usr/bin and the Directory statement should not grant on /usr/bin, but whatever the wsgi script dir is. It is pbr that is in the habit of installing the wsgi script in /usr/bin or /usr/local/bin and that's probably bad. | |
| 19:43:36 | cdent | wally: make sure you restart your apache2, of course | |
| 19:43:57 | wally | this config is locate root@controller:/etc/apache2/sites-enabled# ls | |
| 19:43:58 | wally | right | |
| 19:44:27 | cdent | brb | |
| 19:44:45 | openstackgerrit | Anusha Unnam proposed openstack/nova master: Add description to volumes_attachments.py https://review.openstack.org/445252 | |
| 19:45:05 | openstackgerrit | Anusha Unnam proposed openstack/nova master: Add description to policies in volumes.py https://review.openstack.org/445637 | |
| 19:45:39 | EmilienM | mriedem: not mirror but building the rpm http://logs.openstack.org/22/447522/1/check-tripleo/gate-tripleo-ci-centos-7-ovb-nonha/b8990d9/console.html#_2017-03-20_19_03_26_926347 | |
| 19:45:45 | EmilienM | mriedem: I'm not sure about that one :/ | |
| 19:46:03 | EmilienM | mriedem: i'm looking into logs now | |
| 19:47:24 | wally | hi cdent , i have update the confi file could you please see here, | |
| 19:47:25 | wally | http://paste.openstack.org/show/603486/ | |
| 19:47:53 | wally | now i can get some error message when i run root@compute1:~# curl http://controller:8778 from compute node | |
| 19:48:15 | EmilienM | mriedem: http://paste.openstack.org/show/603487/ - I don't know why | |
| 19:48:17 | wally | maybe still the apache config file problem | |
| 19:48:30 | mriedem | wally: check the placement service logs | |
| 19:48:30 | EmilienM | mriedem: i'm poking packagers | |
| 19:50:18 | wally | i check the nova_placement_error.log | |
| 19:50:31 | wally | please see herer | |
| 19:50:32 | wally | http://paste.openstack.org/show/603488/ | |
| 19:52:17 | wally | thsi is my config file http://paste.openstack.org/show/603489/ | |
| 19:52:23 | mriedem | from signing_dir from nova.conf | |
| 19:52:54 | mriedem | you have signing_dir=/var/cache/nova but signing_dir is deprecated | |
| 19:53:03 | mriedem | and you don't have access to /var/cache/nova anyway | |
| 19:53:11 | mriedem | s/you/the user running the placement service/ | |
| 19:53:27 | wally | is this igning_dir=/var/cache/nova in the nova.conf | |
| 19:54:14 | openstackgerrit | Sarafraj Singh proposed openstack/nova-specs master: Spec to add per instance timeout for LM https://review.openstack.org/442809 | |
| 19:56:50 | coreycb | cdent, that's a good point, thanks. we'll take a look at changing that. | |
| 19:59:15 | rfolco | cdent, can you please see if all good for another +2 on https://review.openstack.org/#/c/416669/ please ? | |
| 19:59:38 | openstackgerrit | Anusha Unnam proposed openstack/nova master: Add descripiton to policies in virtual_interfaces.py https://review.openstack.org/445652 | |
| 20:01:50 | EmilienM | mriedem: nova Ron broke in RDO because a new rep related to tooz, we are working on it | |
| 20:01:55 | wally | now i got this error root@compute1:~# curl http://controller:8778 {"error": {"message": "The request you have made requires authentication.", "code": 401, "title": "Unauthorized"}}root@compute1:~# | |
| 20:02:18 | cfriesen | Hi cores...https://review.openstack.org/#/c/265282/ (spec for flavor information in the instance details) has one +2...anyone up for giving it another? I'm happy to fix johnthetubaguy's nit if that pushes it over. :) | |
| 20:02:19 | EmilienM | Sorry, nova I'm RDO ( on my phone) | |
| 20:02:25 | wally | 2017-03-20 13:01:37.065 25869 WARNING keystonemiddleware.auth_token [-] AuthToken middleware is set with keystone_authtoken.service_token_roles_required set to False. This is backwards compatible but deprecated behaviour. Please set this to True. | |
| 20:02:32 | wally | where to set this | |
| 20:04:31 | efried | claudiub If you're still in a fixing-holes-in-mock mood: https://github.com/testing-cabal/mock/issues/398 | |
| 20:04:52 | mriedem | rfolco: it's on my list | |
| 20:05:00 | mriedem | cfriesen: on my list too | |
| 20:05:39 | rfolco | mriedem, thanks | |
| 20:09:32 | wally | hi mriedem i getting this erro now | |
| 20:09:41 | wally | root@compute1:~# curl http://controller:8778 {"error": {"message": "The request you have made requires authentication.", "code": 401, "title": "Unauthorized"}} | |
| 20:09:44 | wally | any idea? | |
| 20:16:14 | mriedem | wally: pass a token | |
| 20:16:45 | wally | how | |
| 20:16:49 | openstackgerrit | Sarafraj Singh proposed openstack/nova-specs master: Spec to add per instance timeout for LM https://review.openstack.org/442809 | |
| 20:16:52 | wally | from compute node? | |
| 20:16:59 | mriedem | wally: https://developer.openstack.org/api-guide/quick-start/api-quick-start.html#openstack-api-quick-guide | |
| 20:17:26 | mriedem | well, the nova-compute service will be using keystone, and use the credentials for the placement service user setup in the [placement] section of nova.conf | |
| 20:17:32 | mriedem | you can test manually though with curl | |
| 20:17:50 | mriedem | OS_TOKEN=$(openstack token issue -f value -c id) | |
| 20:18:12 | mriedem | curl -s -H "X-Auth-Token: $OS_TOKEN" http://controller:8778/ | |
| 20:20:22 | wally | what can i get the $OS_TOKEN? | |
| 20:20:32 | wally | how can i get the $OS_TOKEN | |
| 20:21:33 | claudiub | efried: hm, that's rather surprising. the mock.patch stuff with autospec is kinda' broken, which is probably why it doesn't have the assert_called_with / assert_not_called methods. it creates NonCallableMagicMock objects, which are not callable, so it would make sense it doesn't have those assert methods. | |
| 20:21:42 | wally | [placement] auth_uri = http://controller:5000 auth_url = http://controller:35357 memcached_servers = controller:11211 auth_type = password project_domain_name = default user_domain_name = default project_name = service username = placement password = conalab os_region_name = RegionOne | |
| 20:21:59 | wally | this is my [placement] in nova.conf | |
| 20:22:05 | wally | is this correct | |
| 20:22:26 | claudiub | efried: i've addressed the mock.patch issue here: https://github.com/testing-cabal/mock/pull/397 . it might also solve the issue you linked. | |
| 20:22:57 | efried | claudiub Looking.. | |
| 20:24:14 | mriedem | wally: project_domain_name = default user_domain_name = default | |
| 20:24:19 | mriedem | i think default is supposed to be Default | |
| 20:24:22 | mriedem | https://review.openstack.org/#/c/438328/13/doc/install-guide/source/nova-compute-install.rst | |
| 20:24:37 | mriedem | auth_url = http://controller:35357/v3 | |
| 20:24:49 | wally | ok let me try | |
| 20:24:49 | mriedem | you don't have the v3 suffix | |
| 20:25:00 | mriedem | you will have to restart the nova-compute process after making those changes | |
| 20:25:07 | wally | yeah | |
| 20:25:13 | efried | claudiub That one might entail a serious behavior change | |
| 20:25:50 | claudiub | true, but only if autospec=True, which never really worked anyways. | |