This content is part of the Essential Guide: What data storage virtualization looks like today
Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

Virtualizing storage for a flexible approach to disaster recovery

Virtualizing storage can expand your disaster recovery options, giving you the ability to replicate storage on products from different vendors.

If you put much stock into what VMware and Microsoft are saying, then you would believe everything in the data center will soon be controlled by software or virtualized. They may be right, since the advantages of virtualized storage can bring new levels of flexibility to the data center. In this article, I'll explain the advantages of, and expose the challenges to, virtualizing storage.

One of the most interesting use cases for virtualizing storage is for disaster recovery. Imagine if you could leverage storage virtualization to bring up a warm data center by replicating storage between remote locations using storage products from two different vendors. Virtualizing storage solves the technical challenge of desperate storage solutions, allows you to use a storage array built from generic x86 hardware at the warm data center and brings advanced capabilities – such as snapshots – to older platforms.

Virtualizing storage for production

To virtualize production storage, a product must first attach natively to the production storage area network (SAN) as a client node. This means that whatever connectivity approach you use in the production system must also exist on the hardware used to virtualize the storage array.

The next requirement is the ability to "see" the volumes to be replicated. This is probably the most complicated part of using storage virtualization for replication. When you are replicating between two arrays from the same vendor, you can use their products to do block-level replication. However, this isn't an option when using a storage virtualization product as a proxy.

One option is to use the storage virtualization product as a production controller. You can provision all available physical storage to the virtualization platform and then provision the production storage from the virtual storage platform. This works when you want to create a single virtual SAN from a pool of disks on different arrays. But, in our case, we are looking to take an existing production setup and replicate it without impacting production.

The most common approach I've seen is to leverage a product that can mount the operating system volumes as read-only targets. This means your production SAN needs to support zoning the LUNs or volumes to multiple hosts. In this case, you'd better be sure that it's read-only and that your underlying application will not be adversely affected by the dual reads from production and replication. This is where native Windows and Linux approaches fall short for most environments.

Replication requirements

Replication isn't overly complicated. The basic requirements are that the product is able to do continuous or scheduled replication over a wide area network. Normally, this will be block-level replication, possibly with the ability to throttle replication based on schedule or available bandwidth.

This is another area where both standard Linux distributions and Windows Server don't have the needed capability.

SAN controller performance

In order to be viable, your virtualized storage must meet your performance needs in case of a disaster recovery event.

Most commercial storage virtualization products will offer the performance you need. The limitation will be the physical storage subsystem and communication transport just as with traditional non-virtualized approaches.

The administrative portion will be a bit more complicated. You need to consider whether a product has the access controls and interface requirements you need for disaster recovery. It's important to note that administering the storage in a disaster recovery scenario will be different, as you are using a different controller than production. It's also important to take application configurations into consideration.

Storage virtualization product options

While native Window Server and Linux distributions can meet some requirements of our disaster recovery scenario, both operating systems lack the capability to see volumes beyond their native file systems. The replication and administrative capabilities needed in production are also missing from general-purpose operating systems. The products listed below might fulfill our needs. This isn't an exhaustive list or an endorsement of a specific vendor, but is meant to give you an idea of the landscape and a basic note about features.

  • HP StoreVirtual is Hewlett Packard's virtual storage product formerly marketed as LeftHand. It's a very mature and robust platform. Like many vendors, HP sells a hardware appliance to create your virtual SAN or provides the software to build your own appliance.
  • FalconStor is a software-first product similar to HP's StoreVirtual. FalconStor has positioned itself to fill this niche specifically for disaster recovery, backup and continuous data protection.
  • While not positioning itself as a virtual SAN provider, Actifio offers many of the features associated with virtual SAN technology. The company focuses on the concept of reducing the physical footprint of data by providing virtualized copy of data for disaster recovery, test and development.
  • VMware's vSAN generated a lot of attention. It is a robust product that allows you to meet many needs. However, the primary drawback to this product is that it's only for VMware environments. There's no simple way to provision and replicate non-VMFS volumes from your production-legacy SAN.

Dig Deeper on Virtual server backup and storage

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.