In the technology boom of the nineties, when networking and data centres were relatively new, using a data centre effectively was pretty straightforward stuff.
Since then, however, all businesses have needed to process more and more data, and managing these kinds of systems has become progressively more difficult.
Factors such as CPU availability, rack space, power and cooling have gradually become a part of the equation, and more factors means more problems.
Here are some of the most common data centre issues every modern business owner should be aware of.
While hackers are a big concern for businesses, one of the biggest threats to modern data centres is often their own energy consumption.
Businesses that are able to find ways to cut down energy consumption will save a significant amount of money, all the while changing the popular perception of data centres being energy guzzlers and damaging to the environment.
In order to reduce your centre’s energy consumption, you first need to make sure you’re measuring it accurately. This kind of monitoring needs to start off with a baseline measurement of the energy which every part of the centre is actually consuming.
To ensure accuracy, you need to cover all the areas, such as your basic IT equipment, your infrastructure for power distribution, adiabatic data centre cooling and so on.
There’s no denying that recurrent system failures have emphasised the need for more proactive response to threats to your data centre uptime.
To help your IT and facility managers control and circumvent these threats, you need to be using an alert engine that will allow you to monitor every single aspect of your data infrastructure.
For the best results, you need to look for a solution with a highly detailed view of your data centre operations, all the while giving you the necessary resources for a drill-down process which can be applied to focus on the smaller details, such as sudden temperature fluctuations on the hot and cold aisles.
Your monitoring system should also cover your generators and their power output. A more comprehensive monitoring system will allow you to pick up on the early warning signs of technical issues, and act on them quickly when they strike.
Countless data centre managers lack the kind of monitoring and visibility resources they need to tell whether their systems are running at peak capacity or not. In the past, a lot of operators have left room for error in order to avoid uptime being interrupted.
This is a practice known as the capacity safety gap, or “over-provisioning” in some circles. Though it may have worked in the past, applying this tactic in 2016 will only waste thousands of dollars’ worth of space. It’s also a great way to waste the power and cooling resources you have.
If this sounds familiar, make sure you set a window for reviewing your current capacity planning. Take the time to pin down where you have any unused physical, power and cooling capacity, and then think of ways to maximise it.