Blog / If It Ain’t Broke, Don’t Fix It

Try with your new car. Anything made after the early 2000’s is purposely designed to be non-serviceable by the backyard mechanic. The engine compartment is stuffed with electronic controls. You can’t begin to troubleshoot unless you have costly fault-code readers and diagnostic tools. They even started using plastic shields over the entire engine bay.  It makes it look clean and modern, but it also hides everything important.  The impression is that there are no serviceable items to worry about.  Of course, that’s not true.  Everything we used to do to maintain the family car still needs to be done; we just let the techs do it – and often only when the Check Engine light comes on.

The same holds true for your IT systems. Most modern server systems hum away in sleek black cabinets with lockable doors.  It looks like it’s made to run forever. Maybe not.

One of our clients that we don’t hear from very often contacted us late 2016.  They were toying with the idea of replacing their aging workstations (6-plus years) with new units. It would be the first serious upgrade they had done for many years. The users were complaining about a general slow-down in processing speeds.  We did some aggressive shopping for them and came up with a very nice unit on sale.  They ordered 32 of them.  This was just before Christmas, and it took a few weeks for them to be delivered to our shop for some in-house pre-configuration work.

They called us about 2 weeks before the installation; their servers had shut down.  It turns out it was due to a power outage.  They had some very nice UPS units (emergency power supplies) installed, but it turns out the batteries had not been replaced within recent memory, and they weren’t able to keep the servers running for more than a minute or two.  We ordered some new UPS units with extended run-time batteries.

On the eve of the workstation installation, they called again.  This time the problems were more serious. They were experiencing very erratic network behaviour.  Two techs were dispatched for some extensive after-hours troubleshooting and emergency repairs.  It turns out the network switches were very old and one of the primary ones was failing – which caused a cascade failure across their entire internal network. We installed a temporary fix and ordered new switches.  We’re hoping they will be delivered this week.

Aside from purchasing replacement parts, this client is faced with many unplanned costs in the form of emergency service calls, expedited shipping, and staff down time; it’s expensive. In a year of uncertain economic conditions,multi-thousand-dollar hits are hard to take.

We like to recommend clients proactively plan and schedule replacement of critical components; servers, switches, firewalls, workstations, laptops and much more.  Most business-class components have a serviceable life of about 5 years.  Some components should be replaced much earlier; UPS batteries for example should be replaced every two years.

Not only do you guard against unplanned and catastrophic failures, you also gain performance and compatibility with the newest technologies. For example, new WiFi standards are available that offer up to a 10-fold increase in speed over systems 3-plus years old, but only if you have the latest WiFi Access points and laptops to take advantage of it.