Description of problem: This is a test only bug for QE to verify that RHV can sustain 1000 (or more) disks in a single block based storage domain. The 'rule' today is that we support 300 disks, but we believe we can go much higher than that. On a hefty system (all-flash storage, Engine DB on a SSD or Enterprise storage), please define 1000 disks in a single storage domain, and measure the time and resource usage of the following scenarios: 1. UI - Select 'New' domain 2. UI - Create new domain (several TB - I assume 5-10TB should suffice). 3. UI - Activate new domain 4. UI - Maintenance new domain 5. Create a template, with 5 disks. Note - the first one should have a real OS, with LVs. 5a - qcow2 disks. 5b - raw disks (less of a priority) 6. Run 100 VMs (essentially, 500 disks are connected). 2 hosts should suffice for this. 6a. Serial launching of the VMs. 6b. Launch them in parallel - 10 at a time or so. 7. Stop 100 VMs. 8. UI - 'Manage Domain' 9. While 100 VMs running, clone some VMs. 10. While 100 VMs running, take snapshots. 11. While 100 VMs running, export some disks. 12. Extend storage domain 13. Run the rest of the VMs, so all disks are 'used'.
Also, verification should be completed only when we did FC as well.