AlexColocation can be an excellent option for companies looking to manage their infrastructure. For the uninitiated, however, data center (DC) jargon can be a real doozy. Look no further as we'll explain some of the basic DC-talk to help you look and sound like a true expert.
U's, 1/2 Cabs and 1/4 Racks Are All Units of Measurement
In the data center world, a “U” is referred to as a rack unit. One U equals 1.75 inches high and is the standard unit of measurement used to determine how much space a cabinet or rack is needed to house your equipment. Each U simply represents a slice of a single row inside a cabinet.
Server cabinets and racks are ultimately synonymous for one another to describe the metal enclosures that are designed and dedicated to house server equipment and cabling.
A good tip is to plan for additional U space in the future, so consider purchasing a few extra U’s if you think you might need to scale up in a few months time. It could be well-worth the investment when first deciding how much space you will need for your rack.
You might also be wondering why some racks have square holes in them and some don’t? Those holes are called universal holes and allow for clipping cage nuts to be added and removed quickly and easily. They allow for more flexibility and provide less wear-and-tear as opposed to the stripping of the typical tapped or threaded hole patterns.
With dozens of variations in shapes and sizes, server racks also come in:
- Portable racks built for the on-the-go IT solution.
- Network racks are built a smidge shallower for networking equipment.
- Two-post (also known as a Telco rack) are used for tighter spaces with more patched-panel type hardware.
- Seismic racks are sturdy enough to provide earthquake protection.
- Cable network racks are for those with large cabling bundles.
- And dozens of many other niche and specialty racks are available for specific purposes.
What's the Difference Between a Quarter Rack and 1/2 Cab?
Racks and cabinets or cabs are again synonymous but are often divided into sections:
- A half (1/2) cab simply translates to a cabinet that was split into two sections.
- A quarter (1/4) rack contains four split sections.
- A full cabinet uses the entire casing as one whole space.
For example, an "8U 1/4 cab" translates to a cabinet with four sections, one of which can house 8 servers.
Why Not Just Charge Me (per U) for That Amount of USpace?
Pricing per U usually involves using a shared cabinet option, which will always run the risk of unauthorized access because it’s like leaving your home unlocked for anyone to rummage through your private belongings. Unless, of course, you’re okay with that.
Put it this way: how would you feel if you lived in an apartment with zero walls or doors to protect your privacy? The same concept applies to the data center environment.
The half cab, quarter cab, and third cab models are there to separate rack space in order to help with security purposes. Even if you only use two servers housed in a quarter cab and keep the rest of the space clear, you wouldn’t want to share that access with strangers, would you?
Additional bells and whistles for U's, 1/2 Cab and 1/4 Racks
Part of the server rack ecosystem also contains a channel casing inside each rack cabinet to organize and cover your cabling and power, but more importantly it’s provided to protect those cables from accidental snipping or loosening's that can be caused by your upstairs or downstairs neighbors while they are configuring their own rack space.
Every cabinet includes some form of a combination lock, usually programmed by the customer themselves. Vendors often include access for themselves as well, just in case you need remote access to turn off a port or to immediately put out any fires or fix any incidents that could affect your valuable equipment.
Colocating your hardware is an efficient and easy option for managing your infrastructure. With options such as the First Right of Refusal (the legal version of calling “dibs") which allows you to reserve cabinet space completely free of charge, or in some cases at least until someone else is vying for it.
Power and redundancy
Data centers offer not only a secured physical location, but also a secure digital environment including firewall, disaster recovery, and backup systems; all completely within a redundant infrastructure.
When discussing matters concerning power or what is often referred to as amperage is determined by the load or the amount of power your equipment will need in order to function properly. This is usually determined very early during a consultation meeting to accurately pinpoint just how much power is required for the equipment. Most amperage information is typically found on the back or bottom of the hardware.
In most cases, you will see a data center designed around using a hot/cold aisle principle. What a hot/cold aisle does is that it maximizes the effectiveness of having hot air and cold air distributed efficiently through a system of optimized ventilation. IT equipment generates a ton of heat and nearly all IT equipment have fans incorporated into them so it can pull air into the unit and circumvent the hot air out of it, thus prolonging hardware life.
Many data centers are also backed up by resilient N+1 redundancies. What this means is that every piece of equipment required to run a part of the facility has a spare component that is available and fully operational in case of an outage, or even worse, an attack affecting an original piece of the equipment. This gives complete peace of mind for data center users as providers can ensure that your hardware is covered, backed up, and can be duplicated so that it always has seamless service available.
Safety and security semantics
So there you have it. Many of these data center terms are simply derived from defining the multiple layers of security that are vastly involved to just enter the front door and then some. If you like to know more, please contact us today!
