The biggest Culprit of downtime in a data center is probably human error, and a key reason for increased chances of human error is bad cabling.
What are some of the causes of downtime in a data center? Equipment failure, bad design, or human error? It’s obviously a combination of all but human error is probably the biggest culprit, and not without good reason. Let’s analyze each in more detail.
Equipment failure can be tackled by eliminating the main causes behind it, and by purchasing good quality equipment. Heat for instance, is the biggest enemy of most electronic equipment and IT is not exception. If the temperature of your servers, storage and other equipment is constantly rising then it’s going to degrade their performance, and eventually lead to system failure. This cause can easily be tackled by ensuring proper ventilation and air-conditioning. Apart from this most enterprises usually have a DR and BCP strategy in place to ensure minimum downtime even if there is equipment failure. Then of course, there are the usual technologies like RAID, automatic fail over systems, etc to ensure business continuity.
Next cause of downtime is poor design. If you’ve increased server density, but your data center does not have appropriate ventilation and air-conditioning for it, then you’re obviously headed for trouble. If you’ve not planned for sufficient power backup, then obviously, you won’t be able to add more equipment in your data center. What you end up doing in such cases is force fit, E.G keeping the rack doors open if they’re made of glass if you don’t have sufficient cooling or extending long cables between different racks to extend power and connectivity and so on.
The third cause, human error, is probably a result of all these adds, moves and changes in your data center. These changes could have been minimized if your data center was designed properly in the first place. Poor cabling desing can increase chances of human error considerably. If you have unnecessary cables dangling all over, then they’re likely to obstruct the cooling vents, thereby increasing the temperature in the data center. Unfortunately most data center administrators will look for alternatives to untangling the jungle of cables ( like keeping a fan in front of racks to keep them cool! ).
It’s only natural then, to have more human errors, and greater downtime subsequently. What’s needed is a good design in the first place, that’;s modular, standardized and one that gets the least affected due to changes.
This time, we’ll look at the cabling aspect of data centers. It’s extremely important to ensure a good cabling design to reduce chances of human error. So, whether you should go for cabling under a raised floor, or above the racks is one question. We’ll understand the pros and cons of each in future articles. Plus, we’ll look at the impact of improper cabling on a data center cling, which can be quite significant. Finally, we’ll also look at some equipment you can use to ensure proper cabling in the data center.