E-Handbook: How hyper-converged secondary storage fits in the HCI world Article 3 of 4

Pavel Losevsky - Fotolia


Converged secondary storage for data protection carries risk

Using converged infrastructure for the secondary storage in your data protection plan can be valuable, but you need to weigh the benefits and drawbacks before choosing that path.

In recent years, secondary storage has rapidly gained popularity as a data protection platform. At the same time, converged infrastructure platforms have become increasingly popular. Even so, the use of converged secondary storage might not be the best option for your data protection scheme.

Converged secondary storage is defined as integrated storage within a purpose-built, hyper-converged infrastructure platform that is being used as a backup. Examples of products in the market include Cohesity's DataPlatform, now in version 4, and Rubrik's Cloud Data Management platform, also hitting version 4 this year.

One of the advantages of converged secondary storage that seems to receive much attention is that it tends to be comparatively easy to deploy and maintain. You won't have to worry about selecting individual components, hunting for drivers or optimizing resources to work together because the vendor has done all of that for you.

A second advantage to using such a storage system is that, although vendors often market converged systems for use in small enterprises or branch offices, converged systems can provide impressive scalability up to enterprise levels. Converged systems use clustered nodes. If an organization needs to increase its storage capacity, it can simply add additional nodes to the cluster. There is a limit to the number of nodes that a cluster can accommodate, but most vendors support multicluster deployments.

Although the node-based architecture used by vendors makes scalability almost effortless, it also has its disadvantages. One such disadvantage is that the only way to increase storage capacity is usually to install additional nodes. These nodes contain more than just raw storage. Nodes are commonly equipped with storage and compute resources and, sometimes, network resources. As such, a storage expansion may require you to pay for hardware resources that you don't necessarily need.

One of the advantages of converged secondary storage that seems to receive much attention is that it tends to be comparatively easy to deploy and maintain.

Perhaps the greatest advantage to using the converged type of secondary storage is that the converged platform is made up of performance-matched hardware and software components that the vendor has certified to be fully compatible with one another. This takes almost all of the work out of hardware purchasing decisions because you won't need to select hardware components individually.

Better still, because the hardware and software components are sold as a single unit, support services are greatly simplified. The vendor acts as a single point of contact for technical support, and you don't have to worry about vendor finger-pointing.

Be aware that vendor lock-in is a risk

The no-hassle nature of using a prebuilt product based on components that a vendor has certified to work together, however, can be something of a double-edged sword. One of the drawbacks to using a converged secondary storage system is vendor lock-in. Prebuilt converged storage products are not vendor-agnostic, so you won't be able to add another vendor's storage products to your converged secondary storage stack. While it is true that you can run alternative storage products alongside a converged platform, you won't be able to attach those products directly to the converged platform without putting the converged system into an unsupported state in the process.

A somewhat related disadvantage is that converged systems are a poor choice for use in organizations that require top-notch performance. On the surface, this one seems counterintuitive because not only is the hardware component performance matched to avoid bottlenecks, but the vendor optimizes the entire system in an effort to strike the best possible balance between performance and stability.

The fact that the hardware components in a converged system are certified to work together means that the vendor has invested a lot of time into testing the platform and its components. Because testing is a labor-intensive and time-consuming process, you aren't usually going to find the latest, cutting-edge hardware in a converged platform. Vendors design converged platforms to be stable and reliable, so vendors choose tried-and-true, extensively tested hardware over the latest and greatest performers.

One last issue to consider is the cost of converged secondary storage. At one time, vendors marketed converged and hyper-converged systems as low-cost alternatives to traditional data center hardware. Today, however, converged systems are anything but inexpensive. It is worth considering whether a converged secondary storage platform will be a good fit for your needs or if you would be better off using a more traditional and likely less expensive product.

It is also worth noting that some vendors have begun offering cloud-based versions of their products. This allows you to receive the benefits of converged secondary storage (especially as it relates to data protection) without the upfront hardware costs. The disadvantages to this approach, however, are that the long-term costs can exceed the purchase cost, and bandwidth limitations may make cloud-based data protection impractical.

Dig Deeper on Converged infrastructure management

Cloud Computing
and ESG