Wake Up CDOs, It's Time to Transition to Storage Virtualization

Data storage can feel unwieldy and hard to manage, with multiple physical devices complicating its usage even further. Instead of utilizing a combination of hard disk, optical media, SSD, NAS and tape, all of your media should be brought together into virtual storage.

This allows the function of the data to be prioritized away from the physical hardware of the storage and managed from a centralized console. Managing it at the virtualization layer will simplify and streamline the process, making it easier to put resources into action.

Benefits of storage virtualization

As organizations grow, they face many challenges and expanding IT needs is one of them. Using storage virtualization has a number of benefits, including:

• Improving flexibility and agility
• Reducing downtime during data migration
• Increasing scalability of attached storage resources
• Responding in real-time
• Accessible by virtual machines
• Cost savings and investment protection
• Can add storage intelligence
• Manage multi-vendor storage devices with a single dashboard

Businesses are now generating and collecting more data than ever before, which offers huge potential, but storage solutions must evolve alongside these changes. Optical media can no longer be relied upon for large-scale file storage or back-up, and phasing it out is a way of future-proofing your organization.

Storage virtualization in cloud computing

Combining virtualization technology with a cloud computing environment can have a powerful effect. Data used in the cloud is controlled by management software layered over the virtual resources to ensure infrastructure, platforms, applications and data can all be manipulated easily.

Adding an additional layer to automate repeatable processes and instructions reduces the need for human interactions, making the cloud more of a self-service tool. Implementing a virtualized cloud server solution improves your server infrastructure without the need to buy and maintain physical hardware.

How to implement storage virtualization

While storage virtualization looks like a single storage pool, it runs on multiple devices and can still be implemented even if there are different vendors and networks involved. The visualization engine will do the hard work and identify available capacity across the storage arrays and media, before aggregating it, managing it and presenting it to all applications.

The software intercepts storage system I/O requests to servers and the engine assigns physical requests to the virtual storage pool and can then access requested data from its physical location. This replaces the need for the central processing unit (CPU) to deal with the request and return the data to storage.

When the computer has completed the process, the engine sends the I/O from the CPU to its physical address and updates the virtual mapping accordingly. Storage admins can then manage multi-vendor arrays from a single browser-based console.

Storage virtualization in varying situations

Your options for storage virtualization will depend on what infrastructure your organization owns. Then, you must take three things into consideration:

Block or file?

Block-based storage virtualization is the most commonly used and works by abstracting the storage system’s logical storage from its physical components, which can include memory blocks and storage media.

File level virtualization requires NAS devices to pool and manage each appliance and makes utilizing multiple NAS devices easier. It can also aid the migration of data from legacy infrastructure.

Host, network or array?

Locating the virtualization engine in a host-based computing component makes sense in virtual machine environments and online applications.

Storage area network (SAN) owners will most likely opt for a network-based storage virtualization solution that can run across Fibre Channel or iSCSI networks.

Array-based storage virtualization is useful in storage tiering and it’s been around for a while, as Redundant Array of Inexpensive Disks (RAID) levels are essentially storage virtualization by another name.

In-band or out-of-band?

In-band storage virtualization works when the engine is between the host and the storage, with both I/O requests and data passing through the virtualization layer.

Out-of-band virtualization places agents on individual servers to send their storage I/O to the virtualization appliance and can help to avoid bottlenecks.

How storage virtualization can save you money

The main principle behind storage virtualization is reducing server hardware, which in turn reduces capital and operational costs, as well as freeing up space that isn’t required to host infrastructure. Network Computing has found that some storage virtualization users have seen server hardware reduced by as much as 80%, and while 40%, 50% and 60% reductions are more likely, it still represents a more efficient approach to housing your data.