re: Virtual Machine Recovery Keeps Improving
Change-block recovery sounds pretty cool, and will especially help with reducing the bandwidth requirements for restoring data from the cloud. It sounds very much like the block-level synchronization commonly used in host-based replication software solutions that create application protection with an active/passive computing pair. (Also supported in cloud DR scenarios.) Change-block recovery is still doing recovery, but as you stated above the recovery process is getting faster. If there is a failure inside the VM image you can get back to an earlier version sooner, but that earlier version will be missing any data changes that occurred between failure point and the restore point. (RPO is affected) There's also the time required to identify the failure, decided on a recovery action, and perform the recovery itself. (RTO affected) For business-critical applications you need more than rapid recovery, you need protection from the downtime in the first place. When downtime and data-loss are not an option, solutions like those from Neverfail Group make a great compliment to backup & recovery tools like these.