Join Log in
Score:0
Wodel avatar Server

Can openstack compute be deployed as nested kvm VM for testing purpose?

sa flag
Wodel

Just for testing and learning purpose, can openstack compute nodes be deployed as VM with nested-kvm?

I tried to do so with Train version of openstack, but I got this errors :

2021-08-17 14:48:51.039 10 WARNING nova.scheduler.client.report [req-92f040e7-3845-4e71-bade-b248e7752407 2697a416f4374ec4b5104b1a6f07763f 210ce11740904ed4a5f693a7c41cb709 - default default]Failed to save allocation for 0c5d829c-7f3e-4d07-afb0-2307515ac6b8.

Got HTTP 409: {"errors": [{"status": 409, "title": "Conflict", "detail": "There was a conflict when trying to complete your request.\n\n Unable to allocate inventory: Unable to create allocation for 'CUSTOM_BAREMETAL' on resource provider '6d3724fc-6f13-4588-bbe5-56bc4f9a4f87'. The requested amount would exceed the capacity. ",

"code": "placement.undefined_code", "request_id": "req-6d6bd59b-adc2-4891-a014-d02b387f7355"}]}

And I get this error also :

Failed to provision instance fc40457e-4b3c-4402-ae9d-c528f2c2ad30: Asynchronous exception: Node failed to deploy. Exception: Agent API for node 6d3724fc-6f13-4588-bbe5-56bc4f9a4f87 returned HTTP status code 404 with error: Not found: Extension with id iscsi not found. for node

Regards.

53
1 + 3
Michael Hampton avatar
cz flag
Michael Hampton
Sure, but you need enough resources to do that. It appears you do not have them.
1
Reply
Wodel avatar
sa flag
Wodel
Could you explain more? every node has 16Go of RAM and 8vCPU, which type of ressources I don't have?
0
Reply
Michael Hampton avatar
cz flag
Michael Hampton
But how much does the host VM have?
0
Reply
Score:0
avatar Server
cn flag
Mircea Vutcovici

When you see an HTTP error, like Got HTTP 409, this means it was reported by the client side. You need to see it from the server side of the REST call. This could be heat, mistral, or something else. Usually haproxy is used to dispatch the call to a backend. In your case the client was nova.scheduler.client.report, this means that you need to enable debug in nova.conf and restart nova-scheduler. nova-scheduler can be a service or container (systemctl, docker ps, podman ps).

After you figure out which backend service returned HTTP/409, you can enable debug in it.

All openstack services are using OSLO library that allows to enable debugging in the same way. Just add debug=true in the [DEFAULT] section of the config file. E.g. If it's containerized:

[root@controller-0 ~]# crudini --get /var/lib/config-data/puppet-generated/nova/etc/nova/nova.conf DEFAULT debug
Parameter not found: debug
[root@controller-0 ~]# crudini --set /var/lib/config-data/puppet-generated/nova/etc/nova/nova.conf DEFAULT debug True
[root@controller-0 ~]# crudini --get /var/lib/config-data/puppet-generated/nova/etc/nova/nova.conf DEFAULT debug
True
[root@controller-0 ~]#
0 + 2
Michael Hampton avatar
cz flag
Michael Hampton
This sounds like a lot of work that could be avoided by simply reading the error log entry.
0
Reply
cn flag
Mircea Vutcovici
nova scheduler needs to be in debug mode to show which host was filtered out by which specific filter. And it doesn't forward this information in the REST answer.
0
Reply
ธงชาติไทย
Elon Musk
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.