Blade servers are appreciated for their compactness, ease of deployment and administration but also their versatility. In 2007, the adoption of the blade server architectures has continued at a rapid pace under the impetus of the three major market leaders such as HP, IBM and Dell, but also because of the renewed dynamism of Sun Microsystems. According to IDC, the share of blade servers in global shipments of servers, for the first time, exceeded 10% by volume in the third quarter of 2007. As explained by the research firm at the time, the blades are the market segment that has experienced the strongest growth with a jump in value of more than 41%.
The success of blade servers is largely due to the expansion of the uses of this class of servers. If the first installations have been confined to the field of computing, scientific or financial, with emblematic examples with the “General Company of Geophysics and SocieteGenerale”, the blades are now widely appealing for large consolidation projects, virtualization or centralisation of applications.
Another factor boosting the sales of blade servers is that they are no longer limited to only x86 world. IBM with its blades and Power Cell, with its HP Itanium blades, blades with Sun Sparc, all started to attract the interest of their installed base of UNIX, particularly in the context of renewal of entry-level servers. The blades are a way for them to lower prices and thus to stop the progression of x86 servers in their market.
If compactness is the advantage of blade servers that is most frequently cited, then let me tell you that it is not the only one. One of their greatest assets is the sharing of critical components within the chassis, such as components of IO. This simplifies deployment and administration but also maximizes the density by limiting almost 80% of the space usually dedicated to connectors. The density of blade servers has the advantage of allowing a significant reduction in physical footprint compared to rack servers (vis a vis a good data cabinet). There is also a certain ease of deployment compared to traditional rack servers. The essential connections (power, network storage and Ethernet) are performed at the reception of the frame. No other connection is needed to be made when installing a blade server in the chassis.
The main advantages of blade servers are to be sought in simplicity of deployment and exploitation. These qualities are in addition to advantages in terms of physical concentration of blades, although we should properly size the power supply and cooling aspects to avoid problems. The centralisation of administration is also a great asset of this type of servers. They all come with a centralized management tool, which, among other things, easily allows images to clone systems, allocate or withdraw resources to different blades. The chassis architecture also facilitates rapid replacement of a defective blade by a reserve which improves system availability. The tools for rapid deployment of images supplied with its server systems and the possibilities of remote supervision with maps of Directors ILO (Integrated Lights Out) integrated into the chassis and blades are additional reasons why blade servers are so popular.Posted in Uncategorized