create VM failed, get "No valid host was found. There are not enough hosts available", what can I do?

[caojia0715cn]

create VM failed. message is  "No valid host was found. There are not enough hosts available",  what can I do?

[dannchen]

Are you using a single host for the compute cluster? I would assume that you have enough resources.

This happened to me in my nested environment. The solution was to add a second host to the compute cluster.

[caojia0715cn]

How do I add a second host to the compute cluster? all is default,  how to operate?

[admin]

1. You can just add host to existing cluster and extra capacity will be automatically consumed.

2. Or If you want to add new cluster then see the picture below

[LockeGalastacia]

I've got the same issue and i'm not expecting it to be a resource issue actually.

My LAB setup has two compute nodes in a cluster that are barely doing anything, and this message keeps popping up.

[admin]

1. use some syslog server so that you can dump all the logs and search in time windows to troubleshoot.  If you don't have a syslog server handy you can get LogInsight Trial version from VMware product sites. This will help save you lot of time.

2. Log into the controller01 by doing ssh controller01 from OMS (this is the management-server VM in the VIO vapp. You want to search for "TRACE" or other signs in the directory:

sudo ls /var/log/nova/

3. Log into the compute driver VM. If you have more than 1 cluster then there will be multiple of these computeXX VMs. Log into the computeXX VM associated to the cluster where you are seeing the failure. (this is where syslog server help for aggregation).

You essentially want to take a look at nova compute log:

sudo ls /var/log/nova/

...

Or you could just log into the OMS and gather all the logs using

> sudo viocli deployment getlogs

then you will need to provide the UUID of instance you are having trouble with and we could help search in the logs for what happened.

Also provide us some more info: what is this image? format? operating system? details of flavor you are using ...etc

[admin]

Are you using the admin tenant? or a different tenant that you have created?

Also, are you assigning a provider network (i.e. external network) to the instance or a private network?

If the provider network, is it shared?

[Whyknott]

I had this same error a few times. My resolution was these 2 things. 1. I had to change the default quota for that tenant. 2 verify that the meta data for that image is correct. I had some issue with the disk type after an import that caused this error also. Deleting old images that were in VIO from the console also resolved this in a pinch. ( do to the quotas )