Template lose specific datastore

When I create a template (KVM) even I select a datastore, this selection disappear when I instanciate it.
No problem whem i’m admin.
But for my client, when they create a VM from a template. The VM is created but not deploy, so, impossible to run it.
I need as admin to update the VM and selected a datastore.

I used OpenNebula 6.0.0.3 (from debian package).

In addition, the error msg is:

Cannot dispatch VM to any Host. Possible reasons: Not enough capacity in Host or System DS, dispatch limit reached, or limit of free leases reached.

The error refers to the scheduler not finding a suitable host to deploy the VM according to the VM scheduling requirements. Take a look at the what the VM demands and what the hosts can provide. For details about this matchmaking, take a look at the log /var/log/one/sched.log. Increase the verbosity of the log in /etc/one/sched.conf if unclear.

I used OpenNebula 6.0.0.3 (from debian package).

This is a very old version, try using 6.8 or 6.4 which are the currently supported ones.

The log is :

Wed Feb 14 14:21:11 2024 [Z0][VM][D]: Found 1 pending/rescheduling VMs.
Wed Feb 14 14:21:11 2024 [Z0][HOST][D]: Discovered 1 enabled hosts.
Wed Feb 14 14:21:11 2024 [Z0][VM][D]: VMs in VMGroups:

Wed Feb 14 14:21:11 2024 [Z0][VNET][D]: Discovered 1 vnets.
Wed Feb 14 14:21:11 2024 [Z0][SCHED][D]: Match-making results for VM 13:
        Cannot schedule VM, there is no suitable host.

Wed Feb 14 14:21:11 2024 [Z0][SCHED][D]: Dispatching VMs to hosts:
        VMID    Priority        Host    System DS
        --------------------------------------------------------------

but, as admin I force the deployement to my node, it’s OK and no info in sched.log
image

As Well, I will check to test the last installation and go back

You can increase the debug level to 5 in order to see the matchmaking process in detail. Specifically which requirements of the VM are not being fulfilled by the hosts.