IT Stability Step 1: Take Human Error out of the Equation

Here’s a common scenario in the world of tape backups: you walk into the server room Monday afternoon, only to find a tape labeled “Thursday” sticking out of the drive. The person who was supposed to change the tapes; maybe she took off early for happy hour on Friday; maybe she went to Vegas for a long weekend and forgot to remind a co-worker to swap the tapes; or, maybe she was busy making money and was just plain too busy to swap the tapes! The real problem with manual tape backups: human error is part of the equation.

Now, tape backups are a easy target to pick on, but the same concept applies to all areas of IT: any time you introduce a human element into your infrastructure, you are introducing a (relatively) likely point of failure. People get sick; people take vacations; people move away; people forget. Machines tend not to do these things. We all know this, and yet, for some reason, most businesses still build their IT infrastructure in a manner that requires some level of regular human attention in order to operate.

A good challenge to help develop your IT strategy is to think through your IT processes and find the steps that require human intervention in order to keep the system running, and get rid of them. Below are some common scenarios to get you thinking:

CRITICAL:	You need a staff member to rotate tapes or an engineer to check backup logs daily to make sure backups are running. This is a bad state to be in. Perhaps it’s time to think about moving to automated online backups or a managed in-house backup solution.
FAIR:	Your shipping department got a virus. Turns out, the AV software expired last month. You do not have a way to manage the software renewal process. This is fixable. Implement a software and hardware inventory management system and you’ll be fine. Alternatively, switch to software as a service options which don’t require regular renewals or updates.
ALMOST:	You have 4 servers in a beautiful rack in a custom-built server room with dedicated air conditioning. Your AC breaks down unexpectedly in the middle of the night. Later the next day a server goes down. An engineer arrives to find that the server room is 20 degrees too hot. In this case, you are almost there. Add a temperature and environment monitor to your server room–configure it to send an alert or gracefully shut down the servers as any indications of “weather” in your server room. You got all the other hardware, might as well spend just a bit more to help keep an eye on things.
EXCELLENT:	If you are at the point where you’re server has to melt down before a real person gets involved, well, congratulations–you have forced the machines to make the errors. Luckily, you’ll probably find that doesn’t happen quite as often, leading to less downtime, fewer headaches, and predictable IT costs.

Of course for small businesses, there will always be a point where having a technical answer for every contingency just gets too expensive. Here’s my advice for that: when you get to the point where automation gets too pricey, that’s the point when you outsource. Find a technology company with the systems and expertise already in place. Work with them to develop a level of service so they can manage what you can’t manage yourself efficiently!

Keep your eyes peeled for my next post, where I’ll talk about the next element to take out of your IT equation: good old fashioned luck.

-Danny G

Jump on the local advertising train ... you can do it!

Windows Live Mesh - the new Foldershare, for FREE!?

Let's talk "Cloud-Computing"!

Small Business Tech Tips

Brought to You by the Folks at Everonit.com