I'm looking for information and advice regarding usage of Veeam with VMware and a mixed environment of some hosts using vFRC(vSphere Flash Read Cache) and some not. According to documentation from vmware backups performed using vmware snapshots on VMs utilizing vFRC can only be restored to hosts that also have vFRC. In my environment we have production hosts licensed for vFRC and do NOT have it at our offsite recovery locations which leads to this issue. The issue supposedly occurs because the snapshot also snapshot the read cache and references it(even though the read cache is redundant and if in the case of disk failure is not needed, the purpose seems to just be for migrating cache settings and not losing them if you move VMs around, which makes sense), however if you restore to a host without the cache location the VM will fail to start because it cannot locate the referenced cache location.
In a quick and dirty test i was able to confirm that it does indeed fail to start if restored to a host without vFRC:
Failed to lock the file
Cannot open the disk '/vmfs/volumes/<path to vmdk>-000001.vmdk' or one of the snapshot disks it depends on.
What is interesting about this test restore is that it restores as if its running on a snapshot with the disks(including disks not using vFRC) all having additional XX-000001.vmdk extensions even though it is not showing an actual snapshot, neither is the backed up VM running on a snapshot. If I restore to original location it also does not act as if its running on a snapshot or create these files, only if i restore to a non vFRC host.
If I run a disk consolidation, the "snapshot" vmdk's disappear and I can now power on the VM. But does anyone know what impact this situation is having on the state of the VM and the consistency of the backups? I'm assuming these -000001.vmdk files must be related to a redirect to the cache location, even though they don't exist on the running VM. Especially with regards to the fact I am backing up SQL servers with application aware backups, I want to make sure I understand what is going on and if running disk consolidation is a valid work around for this problem. In relation to this would it also be possible to somehow automate/script the consolidation in certain restore scenarios?
Thanks for any input.