Upgrading Backup is Too Little, Too Late
One of the big old players in backup software recently announced that they will “blur the lines between backup and archive,” by automatically aging backups to an archive. If your perspective is that of the old-school backup mindset, this probably sounds like progress.
Unfortunately, this is just a baby step towards something that already exists: a unified platform that provides continuity, disaster recovery (DR) and archiving, obviating the need for backup at all.
We call this Three Phase Workload Protection. “Workload”, not “data”, because what’s being protected is the entire server: OS, application, configuration and data. It’s a super-set of data protection. Lose a file? No problem, mount a saved workload as a disk and retrieve the file. But lose a whole server? Also no problem – spin up the saved workload as a virtual server and get back to work.
The key to the Three Phase approach is that the primary (production server) is snapshotted frequently, and the snapshots then automatically proceed through the phases, with no manual intervention needed.
Why are Continuity and DR different? In our view, it’s about what you need to recover, and the best place to recover from. There are three things you can lose:
- A server
- The whole data center
The first two can and should be recovered locally, from a saved workload that’s on your LAN. No big, hairy failover needed. Self-service, fast, no latency. We call that “Continuity”. (In the case of a data recovery, that’s “backup”, which is simply a feature of the Three Phase approach.)
Backup has been relegated to feature status
If you lose your whole data center due to blackout, hurricane, flood, or what have you, you need to failover to a remote facility. That’s Disaster Recovery.
When the workload images get old enough that they are no longer if use for continuity or DR, but still need to be retained for regulatory or policy reasons, they are automatically moved to archival storage. And of course, old workload snapshots can be automatically discarded based on policy.
Sure, we exaggerate when we say that “backup is dead.” It isn’t, it’s simply been relegated to feature status in the much more comprehensive Three Phase Workload Protection model.
- » VMworld 2019: Going big on Kubernetes, Azure availability - and a key ethical message
- » VMware stokes VMworld fires with Pivotal and Carbon Black acquisitions
- » An analysis of Kubernetes and OpenStack combinations for modern data centres
- » What enterprise IT teams can learn from Google Cloud’s June outage: A guide
- » Why it continues to make sense for IT ops to move to the cloud: A guide