For Quantity Discount, Please Call1 888 685 3962
- Feb 6th - Thunderbolt is Gaining Traction in the PC World
- Feb 9th - 5 Uses for Your PCI-express Slots
- Feb 18th - 5 Reasons Why Micro Servers Are More Efficient For Smaller Jobs
- Feb 27 - 4 Major Benefits of Home Automation
- Mar 4th - Tips for Making a Green and Environmentally Friendly Data Center
- Mar 12th - The Differences Between Server/Network Racks and Cabinets
- Mar 23rd - Why is The Federal Cyber Security Budget So Large?
- Mar 31st - Hot Aisle Vs. Cold Aisle Containment in The Data Center
- Apr 15th - 5 Tips for Improving Your Improving Your Data Storage Strategy
- Apr 22 - 5 Tips for Choosing The Right Flash-Based Solutions
- Apr 29 - Amazing Facts About Hard Drives
- May 12th - Tips For Developing A Data Center From The Ground Up
- May 22nd - How VDIs Can Improve Your Business Computing
- June 3rd - 4 Obstacles That Are Inhibiting Microserver Adoption
- June 18th - Developing Prospects For Flash Storage In The Data Center
- July 1st - Combining Heat & Power in the Data Center
- Oct. 8th - Things To Consider When Choosing The Right Sever Rack For The Data Center
June 3rd - 4 Obstacles That Are Inhibiting Microserver Adoption
Way back in February, I did a blog about the benefits of using microservers for specialized tasks. The blog focused on how microservers can actually outperform other servers when used strategically and at a lower premium. Unsurprisingly, my blog didnít win the hearts and minds of IT managers around the world and consequently, microservers are still having a hard time catching on. In light of this tragedy, Iím going to go over the 5 main reasons why microservers arenít quite cutting it in the IT beauty pageant.
Processors Ė Intel processors are essentially dominating the mainstream server market, specifically dealing with the X86 framework. AMD has managed to fight the good fight and hold on to a small percentage of that market share, but itís merely a dent. Because of form factor limitations, Intel has released dumbed-down versions of their traditional server chips along with a server-grade version of their infamous Atom chip. While current versions of the Atom chip are exceptionally capable and efficient, the Atom legacy has been scarred by the old days of early netbooks that used less-palatable versions. This has produced skepticism over the chip that may take some time to blow over for veteran IT folks.
Hardware Ė Adapting microservers also means adapting a new hardware form factor in the datacenter. While the concepts are essentially the same as your typical blade server setup, datacenter designers and IT techs will have to look beyond fundamentals and learn take new operational and design considerations to deploy mircoservers effectively.
Workloads Ė This was the focal point of the previous blog on microservers, and an aspect thatís still shrouded in mystery for most IT managers and datacenter designers. There seems to be a universal understanding within the IT community that microservers boast huge rackspace and energy saving benefits, but there doesnít seem to be any cohesive understanding of what workloads they can manage while saving all that energy and space. Because the processors and hardware of microservers are generally less complex due to form factor limitations, theyíre likely not the ideal candidate for highly complex graphical calculations and the like. Microservers are more inclined to perform well within niche areas of computing that are more quantity driven rather than sheer compute-power driven. Tasks like serving virtual desktops and web hosting are often mentioned by vendors as ideal for the little servers.
Virtualization Ė Microservers limit datacenter designers by the physical server architecture. For many companies that have already invested sizeable numbers into virtualization, this could be a mute point. Large corporations like Facebook have been able to prove that massive microserver farms can function as the most efficient way to manage the infinitesimal number of web requests that they do, but for smaller businesses with less design-particular datacenters, this kind of implementation can be tricky.