Not a surprise really, in the partial survey results released by Managed Objects most IT executives expect a few hours of downtime this year. The interesting piece of information is that there is a disconnect between senior executives and IT executives. 70% Senior executives expect no meaningful failures while 54% of IT executives expect no meaningful failures.

While those numbers are for big companies (>$1 Billion in revenue) the same pattern exists in smaller companies. It’s important to recognize that failures are going to happen, they happen to companies of all size and your IT infrastructure is critical to your business.

…Among the smallest organizations, a slightly smaller slice, (88 percent) saw IT uptime as critical to revenue generation…

“It’s clear that regardless of the size of a company, executives and IT leaders alike realize how important technology has become to generating revenue,”…

Smaller organizations are more at risk, they don’t usually have the system redundancy and don’t have the daily ongoing maintenance of larger companies. So what are you going to do about it.

Here are some things for you to think about.

If your server(s) went down for a week what would it cost you? Do you have a way to ensure that if your server(s) went down you would be back up in the same day or even within a week? If your server(s) completely failed how long would it take you to be back up? If your office burned down would your business burn down?

So, what can you do on a limited budget to mitigate some of this risk? In a small shop it’s difficult, typically a small shop has a small IT budget and not a lot of in house expertise. Hopefully, you have been working with an outside IT firm that can get replacement equipment into your office within hours.

Some simple steps you can take:

Use multiple servers. I am not talking about multiple $10,000 dollar servers but two or three $2,000 servers. When you set them up you pick a few jobs for each one to do. If you buy one big server you have a single point of failure. To avoid a single point of failure, you should at a minimum have two servers.

One server can handle your PMS system, another your email and a third your file sharing. What this does is allow you to rely on using your two remaining servers if one server completely fails. If your PMS server goes down, you restore your backups to your file server and keep going until a replacement comes in.

When you put together a file server use a Raid Array for the disks. It will provide you fault tolerance on your hard drives for a pretty low cost. You can buy 4 drive raid controllers for pretty cheap these days.

Test your nightly backups. Do a restore of your backups onto another machine and make sure the data is good. I have seen many firms who think they are backing everything up, only to find that the tapes they have are bad and nothing has been getting backed up. You should try to do this once a quarter at least.

Simulate a failure and document the process. After you have tested your backups and you know the data is good. Try to bring your PMS up on the file server, try to bring your email up on the PMS box etc… Make sure that if one of your servers goes down you know how to bring it back up on another machine. You don’t want to figure this out when you revenue depends on it.

Store a set of backups off-site. If your office burned down, could you be back up within a week? Certainly not if your data burned up with your office. Generally, keep the last week of backups in your office. After that get them off-site. In the event of something catastrophic you would only lose the data from the last 5 days.

Failures are inevitable, however, with a little planning and practice they don’t have to be catastrophic.