I’d work through the dependency chain instead of immediately blaming the host.
1. Template
2. Service Offering
3. Disk Offering
4. Network
5. Primary Storage
6. Host capacity
7. Host state
8. System VMs
9. Management Server logs
10. Hypervisor logs
The exact error message is important.
For example:
No suitable host → investigate capacity/placement.
Unable to create volume → investigate storage.
Template unavailable → investigate secondary storage/template.
Network allocation failure → investigate networking/IP capacity.
Host disconnected → investigate the host/agent layer.
The key is to identify which dependency failed first.