I have a proxmox 6.4 pve-pool that accesses an external ceph Nautilus cluster for rbd storage (in an erasure coded ceph-pool). All about 80 VMs are currently *not* using krbd to access their rbds. Recently I "rediscovered" krbd and now I am thinking how to migrate existing VMs so that they will use krbd in the future and if there are any dangers ahead for all the data contained in the rbd images?
Since the krbd option is a storage-pool option, activating it would mean that VMs storing data in this pool that are restarted afterwards would automatically use krbd. VMs that would not yet have been restarted would continue to run without krbd until they are restarted?
This sounds a liitle too simple to work. Any objections? Any problems ahead going this way?
Since the krbd option is a storage-pool option, activating it would mean that VMs storing data in this pool that are restarted afterwards would automatically use krbd. VMs that would not yet have been restarted would continue to run without krbd until they are restarted?
This sounds a liitle too simple to work. Any objections? Any problems ahead going this way?