Rack Mount Computing

Google+ Pinterest LinkedIn Tumblr +

The adoption of the rack mount computing platform form factor saw a dramatic increase in the density (both potential and realized) of computing and networking infrastructure and devices. This is because the rack mount computing form factor permitted the “stacking” of devices which meant an increase in the number of computing and networking devices per unit of floor space.

Standard Rack Mount Computing Form Factor

The standardized rack mount computing form factor consists of a rack that has a number of standard spaced post mounting points. The rack itself may be open, partially enclosed or fully enclosed for additional security. A fully enclosed lockable rack that is securely fastened to floor and/or walls does have the additional benefits of severely restricting physical access by unauthorized persons to those components housed within it.

Standard Rack Mount Computing Components and Implementations

Standard rack mount computing implementations can be custom designed and built using various types and configurations of 1U or multiples of 1U components. These include networking devices such as switches and routers as well as power supply units, networking and communications interfaces, servers, user interfaces, consoles/terminals and storage devices (hard disk and optical drives) etc.

The Standard 1U Unit

The standard 1U unit has the following minimum dimensions 19″ (482.6 mm) wide and 1.75″ (44.45 mm) tall. The most common rack mount computing form factor platform is based around a 42U configuration. This means that a 42U rack is capable of housing a maximum of up to 42 individual 1U units. As noted above these 1U units may be servers and networking devices etc.

Rack Mount Computing Component Limitations

When incorporating devices that are themselves physically multiples of the base 1U unit a 42U rack will not surprisingly house considerably fewer than 42 individual components. Another limitation of the rack mount computing platform is that very few self-contained servers can be accommodated into the mere 1.75” (44.45 mm) 1U height of the standard rack mount computing form factor.

This limitation is further compounded whenever a monitor is included into the rack as it will consume considerable space such that even fewer other components can be included as well. However; all is not lost as the rack mount computing form factor does offer raised security levels when used as an enclosure for multiple servers.

Non-Standard 1U Components

Although; the 1U minimum unit size applies to all units that are to be mounted into the rack, including servers, redundant power supplies, networking devices etc, non-standard 1U components can be accommodated.

Oversized Components – “Oversized” components that are multiples of the 1U form factor dimensions can generally be housed in the standard 42U rack without too much difficulty. For example many server-grade enterprise-class redundant power supply units (PSUs) are indeed, very much “oversized”. Fortunately these essential components tend to come in physical sizes and form factors that are absolute multiples of 1U and hence can generally be installed into a standard 42U rack with little difficulty.

Undersized Components – When we come to “undersized” sub-1U components however; it is a far different story. In these cases the general “rule of thumb” precludes these components unless special purpose mounting braces/brackets are used when physically introducing the sub-1U component to the standard 42U rack.

Rack Mount Computing Form Factor Value Added Features

As with all things; the basic standards and standards compliance driven bare bones minimum feature sets and capabilities delivered by all standards compliant rack mount computing form factor solutions are in themselves fine; especially for open standards-based technologies like the rack mount computing form factor platform.

However; it is the additional “value added” features of any product, service or technology that will set it apart and distinguish it as being THE “top of the class/top shelf” must have, no matter the cost product. You will find that “in going the extra mile” by incorporating above standard features, capabilities and attributes will always translate into superior sales and service.

Additional Rack Mount Computing Platform Features

Other elective features/options that may be built into a production environment rack mount computing platform implementation to improve its service, usability and user friendliness include:

Sliding Rails – The addition of slide rails allowing the equipment to be slide in and out without the necessity of disconnecting it from the rack chassis is a truly welcome value added security feature. Positional Locking Slide Rails – Simply by incorporating the capacity for the sliding component to be lockable in both the in and out positions definitely makes a technician or network/systems administrator’s life much easier as this gives superior stability while contributing in no small way to preventing accidental drop-age during routine service.

Server/Device Mounted Locking Pins – Here is a little extra that enables a component unit (server, router etc) to be “dropped” into place without the necessitating for fiddling with screws. This one is a big favorite with me and administrators for obvious reasons.

Rear Mounted Handles – Great for pulling and pushing servers and other devices in and out of the rack chassis. It certainly saves a lot of cable pulling and the ensuing damage that causes.

Rear Mounted Cable Tray – Somewhere to tuck cables neatly out of the way. Not only does this help prevent “cable knotting” but it also gives the added protection of securing cables against accidental displacement. In combination with lockable sliding rails the rear mounted cable tray enables the server equipment to be still operational while it is extended out of the frame; a must for “hot swap” capable devices.

Indicator Lights – An often overlooked aspect of rack mount computing component servicing procedures is that the actual component identification and positive component identification confirmation of those devices currently being serviced can consume copious quantities of an onsite technician’s service time.

Furthermore; this can be compounded considerably whenever a service procedure requires what essentially adds up to concurrent direct physical access to both the front and rear aspects of said device/component. Continually swapping between the front and rear aspects of a rack mount device; especially when fully loaded racks are concerned, serves only to muddy the device identification waters even further.

To overcome these issues many rack mount chassis manufacturers now incorporate device identification LEDs at both the front and rear of the rack assembly. Note that these identification LEDs are built into the chassis rather than the device itself.

Rack Mount Computing Device Naming and Labeling – Because of the difficulties associated with device identification and confirmation whenever service personal are concurrently servicing rack mounted devices from both the front and rear aspects it is essential that you also have a secure, clear and unambiguous naming and labeling mechanism in place right from the “get-go”. Again the labels will need to be applied at both the front and rear aspects of those devices housed in your rack. Most rack and rack chassis manufacturers do make provisions for an easily implemented and maintained rack component labeling systems.

KVM Switch Support – Racks with a KVM switch installed allow an administrator to directly interact with all devices housed within that rack using just the one keyboard, mouse and monitor. This is because the KVM switch provides the necessary support that makes it possible for all components of the rack to share these devices.

Out of the Rack

Standard rack mount servers can generally exist outside of the rack and function within a network environment with just the addition of a power cord and appropriate network cable.


About Author

Leave A Reply