The opening chapter of a white paper, “Cloud-native Data Protection for VMware,” available from Druva is titled “The Current State of VMware Data Protection.” While IT has seen a lot of innovation over the past decade, it seems that the pace of change in data protection is slowing. As a result, many vendors are starting to find it difficult to differentiate as they begin to match each other’s feature sets. It leaves one to wonder if innovation in VMware backup is dead.
Innovation in VMware Data Protection?
In the early days of VMware, the data protection process wasn’t very innovative. Most backup solutions protected a virtual server like they did a physical server. VMware, after a couple of stumbles, got its act together when it introduced its data protection API, VADP. Vendors were able to leverage the VADP API to bring features like change block tracking to market. Building on that they created game-changing features like the ability to instantiate a virtual machine on the backup device. Today though, VMware innovation is stagnating. The cloud offers a new way to innovate VMware data protection, but most vendors settle on using the cloud as a digital dumping ground.
What is Cloud Data Protection Innovation?
Most vendors are still in their infancy in leveraging the cloud. They use the cloud for a disaster recovery copy of the backup dataset; some can even tier older data sets to the cloud. A few vendors are using cloud computing to support cloud-based recovery, also known as disaster recovery as a service (DRaaS).
Clients should look at leveraging cloud storage and DRaaS as cloud table stakes, not innovations. Using the cloud to innovate data protection means vendors using all of the available cloud resources. Cloud storage, for example, is a multi-tiered affair with different price points between them. Data protection software needs to leverage all the available cloud storage tiers to help control costs and meet customer’s performance expectations.
For vendors to leverage cloud computing for more than DRaaS presents a challenge for data protection vendors. Most backup software applications were designed to run self-contained on a single server. They can’t leverage the scale-out compute of the cloud to dynamically increase backup capacity on demand. A cloud data protection solution should be able to scale to as many nodes as is needed to meet a performance demand and then scale-down when the performance demand passes.
Another area for innovation is innovating how efficiently the software executes DRaaS. Many vendors force their customers to wait for a VM to be transformed from a VMware image to a cloud hypervisor image before it is ready for use. The problem is the transformation process can take four hours or more, which misses the RTO requirements for most businesses. Another option is to run the VM in a VMware instance, like VMware Cloud on AWS.
In both recovery cases, the data protection vendor that started the DR process may not be able to protect the now native cloud application. Depending on the disaster, an organization may have to run cloud versions of its applications for a considerable time. While running in the cloud, production data is being added and changed but not protected. Having data unprotected in the cloud is just as risky as having unprotected data on-premises.
The cloud is the next most logical innovation incubator for new data protection capabilities. The problem is that leading the way in cloud innovation first means leveraging the foundational elements of the cloud. Capabilities like multi-tiered storage, scale-out computing, and the protection of native cloud applications are at the top of the list — vendors with these elements already in place have a distinct advantage over competitors that do not.