Is it time to rebuild your network from scratch?

By Mic Geoghegan | March 20, 2017

Here's how 90% of the business networks I've seen have been built. Tell me if this sounds like you! You started with a fairly small, unsophisticated network. You probably had a single server, maybe even a workstation that was utilized as a server. You have a basic switch that you got from BestBuy or Amazon, probably DLink or Linksys. You had a basic firewall, probably SonicWall. As things progressed, you added to the network, maybe a bigger switch, maybe another server, maybe you integrated some kind of cloud services, but generally, your network wasn't intelligently designed, it was pieced together over the years as your needs changed, your network changed. And, it probably works...ish. Most of the time it's fine, when it fails though, it's a little bit of a game digging through old configurations, services that weren't properly removed, switches under peoples desks, a mix of old and new cabling, maybe a random wireless router you added.

Does that sound familiar?

It probably does, because that's how most business networks look! Eventually you reach a point either in size or sophistication that you decide it's time to rip it all out and start from scratch. Well designed business networks require less upkeep, maintenance is easier, and configurations are standardized according to Microsoft or manufacturer best practices. Intelligent networks are easier to maintain and much easier to fix.

So when is it time?

Most of the time, customers wait until they have a major disaster and are looking down the barrel of tens to hundreds of man hours to repair, and that's the point that they decide to just start from scratch. At that point you're either spending a lot of man hours to put it back the way it was or the same amount of man hours to redesign it, so the math is easy.

A better approach is to segment your network into different pieces, and upgrade/redesign them individually. Generally I recommend starting with the infrastructure project. IT infrastructure, in this instance, refers to the network hardware. Switches, access points for wireless, firewalls, cabling (if necessary). Since that is the backbone of the network, I recommend standardizing that first. After that core services is the next logical place to start. Look at what you're using and have a systems architect draft a few proposals for redesigning it. If you're using a local email server, maybe it's time to look at cloud based messaging. If you have all of your servers and services on premise, it might be time to look at using a cloud based, private or public, hosting facility.

Many of these options offer a financial tradeoff.

The upfront cost of migrating is higher, but the follow on maintenance is much lower than you've been paying. Once your core services and infrastructure are up to spec, it's time to look at ancillary services. You might consider moving your phones to a cloud based VOIP provider to increase user mobility. It may be time to look at printers and copiers and see if there's better deals on newer units that require less maintenance and represent less downtime and operating cost. Finally, the last step is generally the user experience. Are you still running Windows 7? Do you have workstations that are older than 3 or 4 years? Is there an intelligent design to your workstation refresh program?

These are all end user considerations that will streamline and control your costs.

Many customers wait until all 25 of their computers are 5 years old and then have a massive project to refresh everyone. This represents both a significant upfront cost and a major user interruption. A more intelligent design is to select a standard rotation timeline (somewhere between 3 and 5 years is standard) and then divide your workstations by the rotation timeline, and that gives you the number of workstations you should be refreshing each year. By staggering the rotation you have a consistent hardware cost for end user hardware and you spread the disruption over many years, instead of a few months.

In the end a well-designed network represents less downtime, simplified maintenance, and less operating costs. When you decide to pull the trigger on executing is ultimately up to you, but I always recommend you make the decision yourself before a major failure makes the decision for you. Need help with network maintenance and server administration services? We are always here as your trusted IT consultant to make sure things are done right the first time!

Tags: Managed IT