Networking: When Good Enough Is No Longer Enough

One of the common customer misconceptions I find is what I call "it's all IT" mindset. Typically, the customer has a small IT department of server and desktop teams that are also expected to run the network since "it's all IT."  While this can work for a while, it will reach a tipping point when good enough falls over.

At this point, I receive calls like this: "Our network is down! We’ve noticed issues recently, but aren't sure why." Or this: "We've had three outages this week! Please help us fix this ASAP!" The root cause almost always stems from the "it's all IT" assumption.

Here is a thought experiment. Suppose you hire a good mechanic, who is great at fixing cars. He does basic maintenance, oil, spark plugs, etc., and can even replace the engine if needed. Now, let's say you ask him to not only work on the car, but design your roads, freeways and traffic lights. This is the "it's all IT" mindset in a nutshell: We have a great team that can fix our cars, so we'll have them design the roads as well. Roads are indeed related to the automobile, but what kind of a transit system would the end result look like? Chances are, you'd end up with a spectacular traffic jam!

Part of the reason for this mindset is that for small to medium-sized businesses, good enough is usually good enough. Often their solution to a problem comes down to “add a new switch.” While this type of organic growth does work, there is a tipping point when the underlying network needs to be addressed.

ethernet.jpg

Here are some common warning signs your network might be near the tipping point:

Many businesses also fall into the "it's all IT" mindset when they view technology investments simply as a cost center. Just as money is spent on freeway design, planning, construction, and maintenance, a network requires resources to remain in a healthy state.

If you've come to the conclusion that your network is nearing the tipping point, what do you do? First you need to do a thorough network discovery. This discovery should include:

With the above information, you can start building your remediation plan.

Rebuilding the network

Once the tipping point is reached, many organizations are tempted to go with a slash-and-burn approach: They will schedule an extended outage and attempt to replace the entire network or a large portion of it within this window. You are better off building a migration plan that prioritizes your high value applications, and a methodical approach to move these applications and departments one at a time. As these are migrated, you start taking the load off the old network and its performance will usually improve, giving you breathing room to finish rebuilding the enterprise.

This measured approach allows you to standardize your design and identify issues missed during network discovery. Planned correctly, many migration outages can be cut from hours down to minutes or seconds with the impact spread across smaller segments rather than the entire enterprise.

To keep your new network healthy, it's imperative to budget time and money for network maintenance. The keys are to include resources for design, documentation, periodic validation, and a hardware replacement lifecycle. Today’s business requires that the network be reliable, fast, and secure. If you can avoid the temptation of "it's all IT," you’ll be on the right path.