| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-06-15 | |||
| 12:38:21 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Remove max_size parameter from fake_libvirt_utils.fetch_*image methods https://review.openstack.org/574911 | |
| 12:38:22 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix nits from trusted certs notification change https://review.openstack.org/575521 | |
| 12:38:23 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add trusted certs to feature support matrix docs https://review.openstack.org/574890 | |
| 12:43:28 | sq4ind | Hey | |
| 12:43:29 | sq4ind | Quick question: service_user for send_service_user_token should be the same across all the services ? | |
| 12:43:43 | sq4ind | So in nova and cinder the same user should be used ? | |
| 12:51:53 | bhagyashris | finucannot: So I have try to install using the $sudo python setup.py install and try to run the test $ tox -e all-plugin intel_nfv_ci_tests but got the error as ERROR: unknown environment 'all-plugin' | |
| 13:00:29 | finucannot | bhagyashris: have you cloned it? | |
| 13:00:35 | fried_rice | tonyb: got a sec for an easy cherry-pick? https://review.openstack.org/#/c/574255/ | |
| 13:00:43 | fried_rice | or other stable core ^ | |
| 13:02:05 | bhagyashris | finucannot: yes | |
| 13:02:47 | finucannot | bhagyashris: And how have you deployed it? With DevStack | |
| 13:02:49 | finucannot | ? | |
| 13:03:10 | bhagyashris | finucannot: yes | |
| 13:03:40 | finucannot | bhagyashris: I usually 'cd /opt/tempest' and run tox from there | |
| 13:04:04 | finucannot | It's a tempest plugin so it will appear in the list of tests for Tempest | |
| 13:04:22 | fried_rice | sq4ind: I'm not an expert here, but since there's only one [service_user] section in the conf, there can only be one registered in your nova.conf. So from the perspective of nova talking to cinder or neutron or whatever, that token had better be accepted by those services. | |
| 13:04:51 | fried_rice | sq4ind: What I don't know is whether the other services themselves have a similar [service_user] concept and/or whether that has to sync up with what's in your nova.conf. | |
| 13:05:03 | sq4ind | fried_rice, just trying | |
| 13:05:32 | sq4ind | fried_rice, answering your question: yes cinder has the same section. | |
| 13:05:43 | fried_rice | sq4ind: It's possible this would be a question better answered in #openstack-keystone | |
| 13:06:11 | sq4ind | fried_rice, will try it first in my deployment and if not I will ask there, thanks a lot ! | |
| 13:06:24 | fried_rice | sq4ind: I suspect the answer is something like: "in order for service A to talk to service B, the [service_user] in A must correspond to a <registered user thingy?> in B" | |
| 13:07:12 | fried_rice | "...but that doesn't mean that the [service_user] in B has to be that same <registered user thingy> unless B needs to talk to itself via that same token." | |
| 13:07:13 | bhagyashris | finucannot: ok let me check | |
| 13:07:25 | fried_rice | sq4ind: only guessing, though. | |
| 13:08:35 | fried_rice | PapaOurs: You appear to be a stable core. Would you mind pushing https://review.openstack.org/#/c/574255/ ? Clean 6-line cherry pick only in unit test code. | |
| 13:10:29 | bhagyashris | finucannot: yeah it's running but all are failing... | |
| 13:12:42 | PapaOurs | fried_rice: roger. | |
| 13:13:18 | bhagyashris | finucannot: total 8 test cases ran and all are failing ... so is there any configuration setup or environment needed to run that test successfully Because i have single node setup whith all the basic services running with devstack | |
| 13:15:20 | giblet | Kevin_Zheng: are you still around? | |
| 13:15:40 | giblet | Kevin_Zheng: I have some suggestion for https://review.openstack.org/#/c/564092/11/releasenotes/notes/add-full-traceback-to-exceptionpayload-06cf8d55d2918eab.yaml | |
| 13:25:26 | fried_rice | Merci PapaOurs | |
| 13:25:43 | PapaOurs | fried_rice: tbh, I need to look more at the stable changes | |
| 13:25:49 | PapaOurs | thanks for the reminder | |
| 13:25:52 | fried_rice | :) | |
| 13:30:12 | finucannot | bhagyashris: You'll probably need to look into why they're failing | |
| 13:30:26 | finucannot | They do make some assumptions about the hardware you're running | |
| 13:32:25 | bhagyashris | finucannot: in your environment all test are passing? | |
| 13:32:45 | finucannot | I don't know. I don't have it installed currently | |
| 13:32:55 | finucannot | They were, last time I ran them | |
| 13:33:13 | bhagyashris | ok | |
| 13:35:05 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Make nova list ignore down cells https://review.openstack.org/575734 | |
| 13:36:38 | openstackgerrit | Georg Hoesch proposed openstack/nova master: handle logfiles for tcp-based consoles https://review.openstack.org/575735 | |
| 13:38:48 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Microversion 2.63 - Add trusted_image_certificates https://review.openstack.org/500396 | |
| 13:38:55 | fried_rice | dansmith: http://logs.openstack.org/45/575545/1/check/build-openstack-sphinx-docs/77f4d6a/html/specs/rocky/approved/nova-cyborg-flow.html | |
| 13:45:17 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Microversion 2.63 - Add trusted_image_certificates https://review.openstack.org/500396 | |
| 13:45:49 | mriedem | trusted certs novaclient change should be good to go https://review.openstack.org/#/c/500396/ | |
| 13:45:54 | openstackgerrit | Dan Smith proposed openstack/nova master: Update the disk_cachemodes to mention an rbd detail https://review.openstack.org/575740 | |
| 13:46:03 | mriedem | glorious docs and tests throughout | |
| 13:46:48 | giblet | fried_rice: I left two comments in https://review.openstack.org/#/c/575545/1 . Nice diagram! | |
| 13:47:06 | fried_rice | giblet: Cool, thank you. | |
| 13:48:03 | superdan | fried_rice: I think the configure step would happen before the plug, but yeah, otherwise that's what I think | |
| 13:48:45 | fried_rice | superdan: "configure"? Is that a new thing or is it the "bind" step we talked about yesterday? | |
| 13:49:01 | superdan | fried_rice: you have it in your diagram homey :) | |
| 13:49:24 | superdan | and I would think cyborg could be doing that after the bind, in anticipation of the plug | |
| 13:49:36 | fried_rice | oh, gotcha. Yeah, that makes sense, will fix. Thanks. | |
| 13:50:11 | superdan | mriedem: my patch above to the disk_cachemodes text patches stuff you wrote about a year ago, if you want to look | |
| 13:50:34 | fried_rice | hm, how tf does cyborg get access to the device to configure it? That's gotta be platform-specific code. | |
| 13:51:36 | superdan | they will need an agent per compute, like neutron | |
| 13:51:48 | superdan | unless they embed all the programming (and image acquisition) code in os-acc, but I wouldn't expect they would | |
| 13:52:13 | bhagyashris | finucannot: thank you :) | |
| 13:52:16 | superdan | the latter is the cinder approach.. brick knows how to configure iscsi and fc and .. | |
| 13:54:40 | mriedem | superdan: +2ed it long long ago | |
| 13:54:56 | mriedem | superdan: the docs i copied into there came from some ibm doc on that stuff, and then i think kashyap added some edits | |
| 13:54:58 | superdan | heh okay | |
| 13:55:03 | superdan | mriedem: which are gone now btw | |
| 13:55:17 | kashyap | Just noticed it; it looks good. | |
| 13:55:38 | giblet | mriedem: in the trusted cert nova client patch. Is there any particular reason why the create() has a formal trusted_image_certificates param while rebuild() expect it via kwargs? | |
| 13:55:42 | giblet | mriedem: https://review.openstack.org/#/c/500396/49/novaclient/v2/servers.py | |
| 13:55:57 | mriedem | giblet: because with rebuild you can unset the trusted certs, | |
| 13:56:09 | mriedem | so if we have a named kwarg defaulting to None, it could be confused as the user intentionally trying to unset the cert ids | |
| 13:56:17 | mriedem | giblet: see my comments in the previous patch set on that same thing | |
| 13:56:33 | giblet | mriedem: thanks | |
| 14:01:19 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Add full traceback to ExceptionPayload in versioned notifications https://review.openstack.org/564092 | |
| 14:02:55 | giblet | mriedem: btw I think ^^ is good to go now | |
| 14:03:16 | giblet | mriedem: Kevin_Zheng made a good job I've just enhance the reno a bit | |
| 14:04:11 | mriedem | ack | |
| 14:08:40 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start populating NUMACell.network_info field https://review.openstack.org/564441 | |
| 14:11:56 | mriedem | finucannot: are you aware of a sphinx extension for doing a nice .. since:: <version> in CLI docs? | |
| 14:16:00 | finucannot | mriedem: It's in core http://www.sphinx-doc.org/en/master/usage/restructuredtext/directives.html#directive-versionadded | |
| 14:17:49 | mriedem | ooo yeah father's day came early! https://photos.app.goo.gl/Ha2JZFZmTo5EETC8A | |
| 14:18:24 | melwitt | noyce | |
| 14:25:40 | giblet | mriedem: what would it mean if the client send [] as trusted_image_certificates in rebuild? | |
| 14:25:56 | mriedem | finucannot: thanks | |
| 14:26:10 | mriedem | giblet: via the python API binding? | |
| 14:26:10 | giblet | mriedem: from shell you cannot do that but from python binding or from curl you can | |
| 14:26:54 | mriedem | giblet: you can't pass [] to the api | |
| 14:27:01 | mriedem | it must be minlength 1 if it's a list, or None | |
| 14:28:01 | mriedem | giblet: https://github.com/openstack/nova/blob/master/nova/api/validation/parameter_types.py#L482 | |
| 14:28:04 | giblet | mriedem: cool | |
| 14:28:18 | giblet | mriedem: I was affraid that it would mean trust nothing | |
| 14:28:48 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Remove doc/build when building docs https://review.openstack.org/575753 | |
| 14:28:48 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Add CLI docs reference for flavor-update https://review.openstack.org/575754 | |
| 14:29:22 | giblet | mriedem: I left some comments in https://review.openstack.org/#/c/500396 nothing serious | |
| 14:30:36 | mriedem | giblet: ok i can address those quick | |
| 14:37:20 | openstackgerrit | Eric Fried proposed openstack/nova master: Enforce placement minimum in nova.cmd.status https://review.openstack.org/568965 | |
| 14:40:28 | openstackgerrit | Matt Riedemann proposed openstack/python-novaclient master: Microversion 2.63 - Add trusted_image_certificates https://review.openstack.org/500396 | |
| 14:40:29 | mriedem | giblet: done ^ | |
| 14:41:09 | mriedem | giblet: if you're ok with that now, can you also go back to the support matrix docs change which is now dependent on the novaclient change and i'm using the nova boot CLI instead of OSC CLI https://review.openstack.org/#/c/574890/ | |