Sometimes, a technology revolution is so powerful that its disruption cascades through multiple markets. VMware’s server virtualization has triggered that “once in a generation” upheaval. What began as a response to processors shifting from faster cores to more cores has changed how we design storage, networks and… now data protection.
From Backup to Data Protection
Relentless data growth has driven traditional backups to the edge. Each day, full backups and restores have become less practical. The rapid deployment of VMware brought the crisis to a head. In addressing challenges with backup and recovery, VMware has shifted the market from legacy server-centric backup to infrastructure-centric data protection services:
- Scaling Backup: To lessen the strain its consolidation put on the traditional backup infrastructure, VMware added Changed Block Tracking. This has become the catalyst for the new wave of versioned replication solutions (a.k.a. “CDP for the masses,” “snap & replicate without the vendor lock-in”, or “full backups in fewer than 5 minutes”).
- Scaling Recovery: When consolidation pushed full recovery windows to the brink of viability, VMware helped transform the approach to large recoveries. First, with Changed Block Recovery, customers can run a near-instant rollback of any corrupted VM. Second, with the encapsulation of the VM, customers can also gain “instant access” to any protected VM. Unlike the horrific Bare Metal Restore experiences, trying to bring up new hardware, customers simply boot a VM from their protection storage and storage vMotion it at their convenience. No more sweating through downtime during full restores.
In doing so, VMware has become the catalyst for shifting the conversation from backup to data protection: versioned replication converges backup (multi-version retention with granular recovery) and disaster recovery (replication with near-instant bulk recovery).
From “My Way” to Protection Services
VMware has been at the forefront of the transformation of the backup team into protection service providers. When the backup teams could not provide sufficient protection services to the VM team, the server virtualization team went rogue. They deployed their own versioned replication protection solutions. They found that, not only did they like the performance, but they also liked the visibility and control of protecting and recovering VMs.
The backup team can no longer offer traditional “backup as a black box.” The VM and application administrators will not blindly trust the backup team. Too often, they have discovered batches of unprotected VMs or out-of-date backup copies. Even when things work, they’re frustrated by the overhead of interacting with the backup team to set up, modify or recover a VM backup.
Therefore, the protection team has to connect to their services to the VM customers via their preferred interfaces. As a result:
- vSphere integration: Whether it is native backup building into vSphere (vSphere Data Protection) or a management interface that integrates with the central backup application, VM admins expect visibility into protection via their interface.
- Automated, Policy-Based Protection: To improve the “time to” for bringing up VMs, teams expect the protection team to, by policy, automatically begin protecting new VMs.
VMware (and application) administrators drove the backup teams to open up their interfaces and to directly connect to their customers via their preferred interfaces – like any successful service provider must.
From Data Recovery to Environment Recovery
VMware has been helping disaster recovery become more automated. As discussed, instant access to the VM backup enables the team to boot a VM immediately after a disaster; this is much faster than provisioning a new server and restoring all the data. As any customer will tell you, however, most interesting applications comprise multiple VMs and the glue that ties them together. Furthermore, the application itself lives within a broader environment – network, security services, etc. They need environment recovery, not just VM recovery.
With the vision of “The Software-Defined Data Center” and the delivery of vCloud Director, VMware is making strides to encapsulate entire customer environments, to enable protection teams to up-level their services:
- vApps: Traditionally both the application and protection teams could not track the relationships between different data sets in the environment. With vApps, however, the protection team can manage both the protection of the VMs and the dependencies among them. This will become a baseline feature for all protection.
- Multi-Tenancy: Multi-tenancy demands that customer services be segmented, forcing more encapsulation of the environment. With vCloud Director, the protection team will need to evolve to support scale-out expansion on-demand, with API-driven (versus UI-driven) management.
- Hybrid Cloud: As VMware blurs the lines between on/off-premise environments, protection teams will need to understand those distinctions (to ensure you have your data and metadata in the appropriate places).
In other words, as disruptive as VMware protection has been over the past few years – the new work with vCloud Director and vCloud Hybrid Services will raise the level of services (and expectations) yet again. This will challenge protection teams and vendors to keep up, but it will also present opportunities to solve problems that have eluded us for decades. How can we shift from simply backing up data to protecting entire application environments? It began with the shift to data protection, continued with the evolution to becoming a service provider and is now leaping into true cloud services. Now is the time to start planning for your future… the revolution isn’t going to wait for you.
Got questions? Ask me at our Meetup, Wednesday, 1:55PM, Booth #1207.