| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-11-17 | |||
| 14:56:20 | mriedem | are you sending a microversion header in the request when creating the server group? | |
| 14:56:31 | mriedem | what tool are you using to create the server group? nova CLI? curl, other? | |
| 14:59:00 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Check quota before creating volume snapshots https://review.openstack.org/520158 | |
| 15:01:49 | ArchiFleKs | mriedem: "status": "CURRENT", "updated": "2013-07-23T11:33:21Z", "links": [{"href": "https://compute:8774/v2.1/", "rel": "self"}], "min_version": "2.1", "version": "2.53", "id": "v2.1"}]}. I'd like to do it with openstack cli | |
| 15:02:35 | ArchiFleKs | can I pass microversion with --os-compute-api-version ? | |
| 15:03:14 | mriedem | ArchiFleKs: that's just the version doc | |
| 15:03:22 | mriedem | telling you what the compute endpoint supports | |
| 15:03:26 | mriedem | microversions are per-request | |
| 15:03:30 | mriedem | and have different meanings | |
| 15:03:59 | mriedem | nova CLI defaults to the latest available microversion in the compute endpoint and what the client side code understands, | |
| 15:04:05 | mriedem | openstack CLI does not do that | |
| 15:04:18 | mriedem | so with OSC, you'll need to specify the --os-compute-api-version option | |
| 15:04:33 | mriedem | openstack --os-compute-api-version 2.15 server group create ... | |
| 15:04:35 | mriedem | something like that | |
| 15:05:07 | mriedem | fyi https://developer.openstack.org/api-guide/compute/microversions.html https://docs.openstack.org/nova/latest/reference/api-microversion-history.html | |
| 15:06:54 | ArchiFleKs | mriedem: thanks | |
| 15:07:58 | openstackgerrit | Eric Fried proposed openstack/nova master: ProviderTree.all_by_uuid() https://review.openstack.org/520243 | |
| 15:07:59 | openstackgerrit | Eric Fried proposed openstack/nova master: SchedulerReportClient._get_providers_in_tree https://review.openstack.org/520663 | |
| 15:07:59 | openstackgerrit | Eric Fried proposed openstack/nova master: Scheduler set_inventory_for_provider does nested https://review.openstack.org/520643 | |
| 15:08:00 | openstackgerrit | Eric Fried proposed openstack/nova master: WIP: Use update_provider_tree from resource tracker https://review.openstack.org/520246 | |
| 15:08:00 | openstackgerrit | Eric Fried proposed openstack/nova master: ProviderTree.populate_from_iterable https://review.openstack.org/520756 | |
| 15:08:01 | openstackgerrit | Eric Fried proposed openstack/nova master: WIP: Scheduler[Report]Client.get_provider_tree https://review.openstack.org/521098 | |
| 15:08:01 | openstackgerrit | Eric Fried proposed openstack/nova master: SchedulerReportClient._get_providers_in_aggregates https://review.openstack.org/521097 | |
| 15:08:22 | ArchiFleKs | mriedem: I tried to do that with os-compute-api-version but it stil fails, I can see that the compute version used is 2.15 in the --debug log though | |
| 15:09:32 | ArchiFleKs | mriedem: but it works with nova cli | |
| 15:09:47 | mriedem | can you put the osc --debug output in a paste? | |
| 15:12:51 | fried_rice | leakypipes I have a test error in https://review.openstack.org/#/c/520643/ which I'm fixing, but would you please take a look and let me know if you're satisfied with my other responses, so I can make all the changes at once? | |
| 15:21:34 | zigo | bauzas: Hi there! Do you know if Nova 16.0.3 includes CVE-2017-16239_Validate_new_image_via_scheduler_during_rebuild.patch ? | |
| 15:21:41 | zigo | It looks like it to me... | |
| 15:21:49 | zigo | But I'd like to be 100% sure. | |
| 15:22:06 | bauzas | zigo: AFAIR, yes | |
| 15:22:09 | leakypipes | fried_rice: yup, doing now. | |
| 15:22:34 | zigo | bauzas: The git log shows it does. I'll push that to Debian unstable then. | |
| 15:24:49 | ArchiFleKs | mriedem: http://paste.openstack.org/show/626622/ | |
| 15:25:59 | zigo | bauzas: Do you know if that bug can be triggered by someone just having a tenant in the cloud, or is it less grave? | |
| 15:26:02 | openstackgerrit | Nguyen Van Trung proposed openstack/nova-specs master: Add validation to improve pep8 check for nova-specs project https://review.openstack.org/519561 | |
| 15:26:22 | zigo | bauzas: I'm just trying to figure out if I should upload to stretch-security or just to the next point release... | |
| 15:26:48 | bauzas | zigo: I need to disappear for a couple of minutes (children at school) | |
| 15:27:10 | zigo | Sure ! | |
| 15:28:09 | zigo | mriedem: Maybe you could reply to me about this? | |
| 15:28:26 | mriedem | ArchiFleKs: that looks like a bug in openstackclient | |
| 15:28:32 | mriedem | i wonder if it's whitelisting the policy options | |
| 15:28:52 | zigo | mriedem: The question was: for the last CVE on Nova (re-schedule issue), can it be triggered by a non-priviledged tenant, or is it just an admin issue? | |
| 15:29:07 | mriedem | zigo: tenant | |
| 15:29:26 | zigo | mriedem: Allright, so this grants a debian-security upload then. | |
| 15:29:28 | mriedem | create server with image A, rebuild server with image B | |
| 15:29:28 | zigo | Thanks. | |
| 15:34:01 | mriedem | leakypipes: fried_rice: fyi, cyborg is looking for some granuler provider filtering stuff in placement, https://etherpad.openstack.org/p/cyborg-feature-tag - looking over that, it sounds like they are looking for a solver scheduler though | |
| 15:34:22 | fried_rice | mriedem Cool, looking... | |
| 15:34:38 | fried_rice | mriedem What's a "solver scheduler"? | |
| 15:34:44 | mriedem | https://www.youtube.com/watch?v=Fsw-hjtQ8e4 | |
| 15:34:55 | mriedem | https://github.com/openstack/nova-solver-scheduler | |
| 15:35:18 | mriedem | alternative to filter scheduler with pluggable filters/weighers for custom optimizations | |
| 15:35:21 | mriedem | long before placement | |
| 15:35:34 | zigo | mriedem: If you don't mind the comment: it'd be nice, next time, to include the CVE number in the patch description. I'll talk about it to Tristant. | |
| 15:35:44 | leakypipes | mriedem: you lose your conversation rights with leakypipes for bringing up solver scheduler. | |
| 15:36:48 | mriedem | zigo: the CVE number didn't exist when i wrote the patches | |
| 15:36:49 | mriedem | sorry | |
| 15:36:59 | leakypipes | mriedem: I'm just kidding :) | |
| 15:37:05 | mriedem | i was saying sorry to zigo | |
| 15:37:07 | mriedem | :) | |
| 15:37:13 | mriedem | not the leakiest of pipes | |
| 15:37:19 | leakypipes | well, I retract my kidding, then :P | |
| 15:37:55 | mriedem | leakypipes: if it makes you feel better, they also want this in k8s | |
| 15:37:59 | mriedem | so sharing the load | |
| 15:38:15 | leakypipes | mriedem: heh | |
| 15:38:16 | mriedem | openstack can still ruin k8s yet! | |
| 15:38:42 | ayoung | mriedem, well one way we can catch up to Kubernetes is to fix our RBAC | |
| 15:39:18 | ayoung | and on that note, can we please merge https://review.openstack.org/#/c/384148/ as I've been chasing this now for over a year. The tests run right now, and I'd like to avoid more breakages | |
| 15:39:22 | mriedem | ayoung: will fixing our rbac get me to a thinner slimmer openstack for the edge? because that's all anyone cares about | |
| 15:39:38 | ayoung | mriedem, absolutely | |
| 15:40:00 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: instance_mapping/request_spec records exist even after the instances are archived https://review.openstack.org/515034 | |
| 15:40:02 | ayoung | mriedem, it will help with all members of U2 | |
| 15:40:13 | artom_ | mriedem, we should start with an openstack with sunglasses for bono, then we can think about the edge | |
| 15:40:29 | ayoung | Start with the Drummer. He's the sensible one | |
| 15:40:36 | mriedem | https://www.youtube.com/watch?v=UJXqnYCWW7Y | |
| 15:41:50 | ayoung | Anyone, fixing this bug requires getting changes into absolutely every project in OpenStack, and it is not a fun thing to chase | |
| 15:42:48 | ayoung | its coming up more and more, we are getting dupes of bug filings for it, and the requests for RBAC changes that cannot be done without proper scoping are piling up | |
| 15:43:10 | ayoung | I have a T-shirt for whomever finally pulls the trigger on it | |
| 15:45:17 | zigo | Now I get this when building Nova 16.0.3: http://paste.openstack.org/show/626628/ | |
| 15:45:17 | mriedem | ArchiFleKs: this is the bug https://github.com/openstack/python-openstackclient/blob/master/openstackclient/compute/v2/server_group.py#L58 | |
| 15:45:33 | mriedem | ArchiFleKs: osc is hard-coding the choices for the server group policies | |
| 15:46:45 | zigo | This really looks like a broken test, rather than an issue in Nova itself. | |
| 15:47:44 | openstack | Launchpad bug 1732938 in python-openstackclient "Unable to create server group with soft-anti-affinity policy using microversion 2.15" [Undecided,New] | |
| 15:47:44 | mriedem | ArchiFleKs: https://bugs.launchpad.net/python-openstackclient/+bug/1732938 | |
| 15:49:22 | mriedem | test_validate_volume_no_mtab_entry | |
| 15:49:22 | mriedem | zigo: yes, although the test is mocking out that call | |
| 15:49:25 | mriedem | oops | |
| 15:49:30 | mriedem | @mock.patch.object(psutil, "disk_partitions") | |
| 15:50:04 | mriedem | zigo: oh that's not fixed in pike | |
| 15:50:05 | mriedem | sec | |
| 15:50:24 | openstackgerrit | Matt Riedemann proposed openstack/nova stable/pike: Fix quobyte test_validate_volume_no_mtab_entry https://review.openstack.org/521118 | |
| 15:50:27 | mriedem | zigo: ^ | |
| 15:51:19 | zigo | mriedem: Wonderful, that's very helpful for me, and much better than blacklisting the test. | |
| 15:57:28 | mriedem | easy bp win https://review.openstack.org/#/c/450636/ | |
| 16:01:34 | ArchiFleKs | mriedem: thanks a lot i'll follow that | |
| 16:05:23 | bauzas | mriedem: indeed, will look into the vmware log change | |
| 16:37:04 | cfriesen | mriedem: are you planning on updating https://review.openstack.org/#/c/520660/ to query the volume image metadata? (Using _get_bdm_image_metadata() maybe?). If not would you like me to take a crack at it? | |
| 16:37:48 | mriedem | yeah planning on it, was just looking at something related to that | |
| 16:42:23 | fried_rice | Is the multi-node business (e.g. in resource_tracker _init_comput_host) *only* ironic? | |