How not to get stung by storage virtualization: Part 1

Storage virtualization isn’t just the buzz term du jour. More and more organizations are transitioning antiquated legacy storage environments into more flexible and agile infrastructures that can help reduce waste, outages and labor, maximize operating efficiency and performance – and, of course, save a lot of money. In fact, by some accounts, for every 12 terabytes of usable disk capacity, there is the potential to save $1 million in net operating expense.

That’s a lot of savings. But, like other types of virtualization, this one isn’t all mom and apple pie either. Transitioning to a virtualized storage environment comes with some risk and potential disruption to your organization, which I’ll review in this post and the next. But just like climbing a mountain, once you understand and mitigate the risks, reaching the top is that much more rewarding.

Risk 1: Failed implementation

As with any upgrade or conversion, there’s a chance that things won’t go exactly as planned at some point. Once the virtual storage system’s abstraction layer is in place, only the virtualizer will know where the physical data is, so you’ll need to have a back-out procedure in place in case something goes wrong and the implementation fails. It may be time-consuming but the alternative – not knowing where your data went – is obviously worse.

Risk 2: Failure to communicate

Interoperability is key to the successful implementation and control of any virtual system. After all a virtual storage system can include a mix of potentially dozens or hundreds of different physical storage controllers, hosts, operating systems, software and hardware components. If they can’t all speak the same language or if the companies that make these various components can’t offer you coordinated support when you need it, it’s safe to say chaos ensues.

The level of interoperability required will differ based on the implementation your organization chooses. For instance storage controller, switch-based and network-based virtualization each have different and increasing levels of interoperability requirements. Be sure to speak to potential vendors about the ability of their components and support to work together.

In Part 2, I’ll give a brief overview of three more storage virtualization risks: complexity, meta-data management, and performance and scalability.

About Andy Thomas

With over 16 years of experience as an IBM Business Partner, Andy has worked in the open systems, server virtualization, and storage consolidation areas.