SAN Plans: Flirting With Disaster

When we heard about all the millions of dollars being wasted by companies buying storage networking technology they dont need -- or, in some cases, forget they’d even acquired -- we decided to see if we couldn't help our readers avoid some of the tragic stories we’ve been hearing about lately (see Millions 'Wasted' on Storage Nets).

A prime example: Relera Internet Centers, flush with $300 million in funding at the start of the year, proceeded to build hosting centers across the U.S. containing rack upon rack of EMC Corp. (NYSE: EMC) storage arrays... As if there weren’t enough empty data centers scattered across the country already! According to F***edcompany.com, Relera has closed its doors.

Then there was the law firm that recently forked out $700,000 on a storage system that wasn’t even being used, according to Jason Rabbetts, co-founder and commercial director of U.K. storage consultancy firm Source Consulting. The company bought a massive disk repository from EMC and forgot about it. Each department assumed it was being used by other departments when, in fact, it was standing idle.

The reverse happened at another company known to Source Consulting. It lost close to $15,000 an hour when its systems went down for two days without sufficient backup, according to Rabbetts. The problem? “The company had 300 more servers than it thought it had."

We asked Source Consulting, IBM Global Services, and KPMG Ltd. to come up with 10 ways to avoid disasters like this. Here’s a synopsis of the results:1. Get Advice. Just make sure it’s from independent experts that have no axe to grind when it comes to platform, storage, and software vendor choice. And ensure your advisors have proven track records!

2. Don’t just throw money at the problem. Make sure you thoroughly understand what you wish to achieve -- not simply from a technical standpoint but also considering the business demands that are driving those technical changes. Sticking a multimillion-dollar EMC Symmetrix system in to do file sharing is probably overkill.

3. Calculate current Total Cost of Ownership (TCO). A primary motivator for a change in storage strategy is to reduce costs. However, very few people establish what their current TCO is and therefore have no fiscal barometer to either justify new investment or measure the success of the new project.

4. Make sure your new TCO model is lower than your old one. Apparently, many companies have been caught out on this one.

5. Avoid “finger in the air” assumptions. Everyone wants "bigger and faster," but where is it best applied and how big does it need to be? By understanding exactly who is using what capacity, where, when, and how, you can design a scaleable solution from an informed position based on fact rather than assumption.6. Conserve the appropriate architecture. Throwing the baby out with the bathwater is not the answer. Many organizations will have significant investments in underutilized equipment and software, much of which can be redeployed.

7. Avoid complexity. With so many new vendors, new products, and new technologies it is very tempting to over-engineer a new storage solution. Don't be tempted. If you build in complexity then you build in cost and inflexibility -- not things you wish to burden your business with in times of changing technology and economic unpredictability.

8. That line about nobody getting fired for buying EMC... It’s rubbish. You will not only get fired should it all go wrong, but no one will ever employ you again for being such a dumbass.

9. Establish a storage team. It must be made accountable and responsible for the future of this project. (Read: someone to blame when it all goes wrong.)

10. If it does all go belly-up... Make sure it's Labor Day weekend when the fruits of your labor turn rotten. They will have forgotten about it by the time everyone gets back. Really.— Jo Maitland, Senior Editor, Byte and Switch http://www.byteandswitch.com