The Business Benefits of Blade Computing

How much and in what way blade computing can benefit businesses of all sizes depends as much on the core ideologies of this bunch of technologies as it does upon the skillful selection and blending of the individual blade units incorporated into every solution at both the hardware and software levels.

Being constructed from highly-specialized information processing and storage units; designed for high density plug “n” play implementations where future expansion, serviceability, versatility, scalability and flexibility are considered to be priorities blade computing offers enterprises of all sizes many benefits. A lot of which are not immediately apparent as the actual blade units are carefully secreted away in self-contained housing units known as blade enclosures.

The Blade Computing Model

Distributed centralized specialization and resource rationalization are the fundamental tenets of the blade computing model. This means that blade enclosures such as the HP C7000 depicted in Figure 1 can be deployed in geographically dispersed locations to provide an organization with a distributed specialized central processing and servicing functionality less prone to organization-wide single-point-of-failure scenarios.

Unlike the bastion host, standard rack mounted servers or even traditional clustered servers the blade computing model generally uses more specialized units dedicated to performing fewer more specialized computing functions and processes.

To this end; many blade units are shipped with various combinations of different hardware and support systems normally associated with computers and server-class devices absent. This is because those specific functionalities are provided by the blade chassis/enclosure.

Having the blade enclosure provide many of these integral functions rather than having them duplicated by each individual blade component type saves considerably in the numbers of components incorporated into the manufacturing of each blade unit. This as you can imagine provides substantial cost savings including transport charges. In other instances other specialized blade units will be dedicated to these “missing” functionalities.

In this way many systems that have been massively duplicated and hence often grossly underutilized in the more traditional server and computing models are not to be found in the same massive numbers in the blade computing scenario. This has resulted in blades that are specialized for processing, network, storage, Input / Output (I/O) and memory subsystems.

“A blade for every function” is the phrase that comes to mind. The IBM HS20 Blade server pictured in Figure 2 has provision for two compact high-performance 2.5” SCSI hard drives.

If you need more processing power then add a blade that is loaded with multiple multi-core CPUs and oh-la-la, more processing power to brag about is yours. Similarly if it’s more memory for those graphics rendering intensive applications that you require, then add a memory blade. You can even get “general purpose” blades.

Task and component specialization along with purpose-driven blade design are but two of the ways that blade computing provides greater economic rationalization and more efficient use of computing resources at all levels within the blade system.

Space and Energy Savings

The idea behind this approach is that through removing many of these over-duplicated under-utilized components considerable space savings are achievable. In addition; removing the need for at least one power supply unit (PSU) per server results in a considerable reduction in the numbers of power supply units (PSU) built into or supplied with each individual server unit.

This not only saves considerable space it also reduces the overall excess thermal energy produced during normal and stress level operating conditions. The less floor space occupied by your data center, communications, networking and computing infrastructure the less cooling it requires and the less rent you have to pay.

The result of reducing the amount of excess (waste) thermal energy production during processing is that substantial savings can be delivered on cooling solution energy requirements. This also serves to deliver the “green points” or the carbon production reduction contributions that companies are now being asked for by government, environmentalists and the general public alike.

Everyone who has placed their hand at the top rear of a switched-on ATX tower PC is only too well aware of the amount of heat that the PSU generates. Factor in the fact that the PSU is the most common component of all computers to fail and you see the sense in replacing thirty or forty cheaper PSUs with two to four more robust and reliable units.

From a big picture perspective; the fewer more expensive higher performing more efficient power supply unit solution actually has a far cheaper setup cost than the former multiplicity of PSUs solution. The HP Blade Enclosure pictured in Figure 1 only uses two superior Uninterruptable PSUs and this particular system has 16 blade servers installed.

Through the wholesale removal and replacement of older technology PSUs with dedicated leads from DC units of superior performing, less failure-prone, higher quality redundant sets of PSUs blade computing can indeed deliver considerable energy and accrued downtime loss savings.

Remember; that traditionally, it has been PSU issues that contribute nearly 75% of all IT service-related site visits. Because; datacenter downtime is the most expensive of all outage types experienced by enterprise today, reductions in this area immediately deliver considerable fiscal savings.

The Blade Enclosure

The big thing in blade computing is the engineering technologies encompassed by the blade enclosure. This is what makes it all possible. Here are some of the support features that the blade enclosures deliver:

Unit Housing – Each blade enclosure is capable of providing housing for multiple blade servers at far higher densities than traditional rack mount implementations. In fact; up to three times the number of units in the same 3D space.

Specialized Support – Blade enclosures provide housing and specialty support for specialized blade units e.g. compute, memory, storage etc.

Connectivity – The blade enclosure provides the specialized interfaces by which each individual blade units are connected (plugged into) to the chassis. As IBM make a big point of in their advertisements this can be done as a hot install without the need for attention to cabling.

The reason for this is that each blade enclosure is engineered to be preconfigured and “pre-wired” ready to accept additional post-implementation reconfiguration regardless of whether they be removals, additions or just rearrangements of the blade units housed within each enclosure.

Utility Services – The blade enclosure is usually responsible for the provisioning of redundant utility services such as electrical power and cooling for all units contained within each enclosure. I say generally since some implementations even use external power supply systems that are shared among multiple blade enclosures for additional cost and space savings.

Transformer related functionality is one aspect of power supply that does generate considerable excess thermal energy as a by-product and having this particular power supply function external to the blade does reduce the ambient temperature within each blade enclosure considerably thereby delivering additional energy consumption and fiscal savings.

Supporting Infrastructure – It is the job of the blade enclosure to provide infrastructure support such as communications, networking and various interconnects for all blade units housed within it as well as blade enclosure-to-blade enclosure interoperability.

Today this tends to be through the utilization of fiber optic networking technologies as they deliver the greatest performance, least excess thermal energy production and the greatest immunity to noise, Electromagnetic Interference (EMI) and environmental aberrations. Fiber optic networking and communications technologies are also more resistant to interception than copper and RF technologies.

Physical Security – The blade enclosure provides the physical security for those elements housed within it. Compartmentalized accessibility means that to physically access any individual blade component one need not necessarily be granted automatic physical access to every other blade unit housed within each blade enclosure by default.

Platform Management – The enclosure also either links to an administrative unit or contains an administrator interface that allows for authorized authenticated administrative functionalities to be performed upon all units contained within each enclosure. In circumstances that require it; remote administration is also possible.

Blade Unit Manufacturer Variation – Different manufacturers do have different ideas of what should be included with each blade as well as what should or should not be provided with the blade enclosure. However, it is all of these elements in combination that make the blade computing platform.

Blade Computing Benefits

The benefits of the blade computing platform include:

Minimum Size Limit Restrictions Removed – Removal of the 1U minimum size limit requirement which the traditional rack mounted server platform is stuck with. This permits greater freedom and initiatives for blade system designers and increases the overall versatility and adaptable flexibility of the blade platform as a technology and each blade enclosure and blade unit as an enterprise asset.

Higher Unit Densities – Instead of the 42 unit maximum of the standard rack it is not unusual for blade systems to hold up to 100 or more individual blade units per blade enclosure. These higher densities translate into greatly reduced floor space being occupied by the data center which in turn reduces the amount of cooling required thereby producing greater energy cost savings and a far more rounded, greener, less carbon producing computing solution.

Specialization – Through blade unit specialization further additional significant size reductions are achieved and will no doubt continue to do so. These all combine; a little by a little, eventually producing massive savings organization-wide. Larger enterprises will obviously derive greater initial savings here. The rest of us will benefit from the domino effect this produces.

Improved Energy Efficiency – The “Greener IT” motivation is becoming an ever increasing influence in technology today and will undoubtedly continue to do so in the future. Not only does blade computing reduce an organization’s energy costs it also represents a much “greener” computing initiative.

Organization-Wide Consistency – Through organization-wide component consistency blade units can deliver a more uniform organizational computing environment throughout a geographically diverse enterprise.

Scalability – Enterprise scalability issues are greatly simplified with blade computing. If you require more processing power then you simply install additional compute blade units into select blade enclosures throughout the enterprise or data center.

Redundancy – For mission critical functionality and the truly paranoid blade enclosures do allow for varying degrees of redundant multiplicity of systems in highly customizable configurations and arrays.

Future-Proofing – The concept of using the blade enclosure to provide infrastructure services to multiple blade units allows much greater freedom in future proofing an organization’s information technology and infrastructure investments. Even when fully loaded upgrading can mean swapping out lower performing or lower capacity blade units for better performing higher capacity units. Once again real estate savings are a strong driving factor here.

Recyclability – Blade units swapped out of their enclosures can be reused in other enclosures or even in blade enclosures at different geographical locations. In this way upgrading can become more of hand-me-down oriented which does prolong the useful life-expectancy of IT resources considerably. Through reusing units already manufactured considerable carbon production savings can be realized.

Flexibility – In today’s economic climate business requires its processing facilities and solutions to be built around genuine flexibility from both the hardware and the software perspectives more than ever before. Here is one area where blades can excel by using general purpose blade units for general purpose processing tasks and specialized blade units to deliver specialized processing services where required.

For example; web page retrieval and database searches do not require the advanced vector processing that many rendering applications demand. Through the blending of special purpose blade units with general purpose and storage blades all of these computing tasks can be performed within the same enclosure.

The additional administrative flexibility offered by blade computing allows the system administrator to selectively and electively assign and reassign resources between applications/customers. Physical configuration changes can be made on the fly without the need to power down every unit within the enclosure or to attend to laborious cabling issues.

Virtualization – Another technology that the blade enclosure does lend itself to with remarkable agility is the implementation of multiple virtual computing environments. The advent of massively available mainstream multi-core microprocessor technologies have played a large part in the expansion of the deployment of virtualization technologies hosted on blade server “farms”. Once again this is all about the rationalization of information technology resources usage.

Distributed Centralized Processing – Blades are particularly suited to a distributed central processing ethos. This makes them especially desirable for Web 2.0 applications, Software-as-a-Service (S-a-a-s) and cloud computing.

Thus web hosting and cloud computing service providers can distribute their resources more evenly on a usage basis. It is most profitable to have your facility in as close a proximity to your market, customers or user-base as possible without paying for highly prominent real estate or the rents these sites ask. Ideally as close as possible to the physical location of the Internet backbone would be the most desirable of all locations for a blade center.

Additional Information – Check out ServerWatch.com if you would like to find out more about various specialty servers and new developments in the server industry.