Tech Primer

Blade Servers

What are blade servers?

A blade server is a thin, modular electronic circuit board, containing one or more microprocessors and memory, that is meant to run a single, dedicated application (such as a Web server). IDC defines a blade server as a computing system that includes processors and memory on a single board. Many such boards can be easily inserted into a single rack mountable unit. One product offering, for instance, permits the installation of up to 280 blade server modules in multiple rack units or rows of a single cabinet.

Blade servers share a common high-speed bus and generate less heat, saving on cooling costs and make it possible for designers to cram them into a smaller form factor. Power, cooling and network access services are not contained on the server blade. These resources can be shared among a collection of blades that are, in turn, accessed through the backplane of the chassis. The power and bus connections are part of the cabinet that houses a collection of blades. Blades can be easily installed and removed and are smaller than rack-optimised servers. Blades may be general-purpose servers or they may be tailored and pre-configured for specific data centre needs (e.g., as security blades with firewall, virtual private network [VPN], and pre-installed intrusion detection software).

Are there variations in blade servers?

A blade server usually comes with an operating system and the application (programme) to which it is dedicated. Individual blade servers come in various heights, including 5.25 inches (the 3U model), 1.75 inches (1U) and even ‘sub-U’ sizes. (‘U’ is a standard measure of vertical height of a rack unit equal to 1.75 inches.) There are many standards for connecting blade servers and these include compact PCI (cPCI), PCI-X or Fast Ethernet. Blades are largely identical, but can be configured with more disks or processors to serve specific functions, such as a NAS or a firewall.

Are blade servers more efficient than standard servers?

The reason for greater efficiency in blade servers is its ease of management. The management of multiple blades can be performed with a simple interface. Moreover, blade servers can be easily scaled-up when compared to traditional servers.

Do blade servers help reduce operational costs?

By providing greater processor density and taking up less floor space, server blades help cut costs in a data centre. Blade servers also require less power, fewer power supply units, thus reducing capital costs. This is because power is provided to many blades by shared power sources built into the blade chassis. Cooling requirements for a cabinet of server blades will be simpler to maintain than separate cooling systems for racks and pedestals. Cabling systems will be simpler to deploy and maintain. All these cost advantages are driven by the basic design of a server blade that emphasises resource sharing.

Where are they likely to be used?

Large data centres and Internet Service Providers (ISPs) that host websites are among companies that are most likely to buy blade servers. A blade server, also referred to as a high-density server, is typically used to take over the duties of a cluster of servers that are dedicated to a single task, such as file sharing, webpage serving and caching and encrypting and Web communications using SSL. Blade servers are also popular for streaming audio and video content.

Do server majors sell blades?

Leading server vendors such as Sun, Dell, IBM and HP—all offer blade servers.

For more information on blade servers check out the following link:

www.serverwatch.com/tutorials/article.php/10825_2194221_1/