

 |
|
 |
 |
 |
Dynamic Capacity Management for Virtual Environments
Learn More |
 |
Transformational Analytics: Virtualizing IT Environments
Learn More |
|
 |
 |
 |
 |
 |
|
 |
 |
Gartner 27th Annual Data
Center Conference
December 2-5th, 2008
MGM Grand Hotel & Casino
Las Vegas, NV
Learn More
|
 |
 |
 |
 |
|
Green IT
For many organizations, IT infrastructure represents the single biggest opportunity to make change for the positive with respect to environmental impact. The biggest contributor to power consumption and heat in the data center is servers. Servers in aggregate consume huge amounts of power and generate significant amounts of heat. The heat must then be dealt with through cooling, which in turn demands more power. X86 servers have proliferated over the past decade to the point that many organizations simply cannot get enough power into their data centers to sustain the servers they need to deploy in support of the business. The irony is, that depending on the data or research you view, the most commonly cited metric on average X86 server utilization is <10%.
These low levels of utilization are extremely wasteful and inefficient and result in not only wasted financial resources, but greatly impact the environment through needless power consumption.
New technologies, such as virtualization, represent an opportunity to drastically reduce the number of physical devices required to sustain the computing demand of a business. Virtualization enables organizations to move away from a “one app, one box” mentality. Of course the hosting of multiple applications on a single server brings about a variety of new management issues.
- The demands of multiple applications must be met by a single server without impacting performance
- Regulatory requirements and organizational divisions may impact what applications can be hosted on the same hardware
- Service levels, maintenance windows and a variety of other non-technical constraints will shape how applications can be consolidated onto smaller numbers of servers
It’s not as simple as removing servers and taking advantage of consolidation. As servers workloads go up, so does their power consumption. This dynamic necessitates a careful analysis of what the before and after power consumption will be given certain combinations. Different server models, chip sets etc will also impact power consumption.
CiRBA’s Placement Intelligence Technology™ enables you to leverage empirical information from your data center combined with business constraints and hypothetical hardware models to rapidly create “What-if” scenarios that illustrate the detailed impact of any path. The before and after analytics show the impact from a utilization perspective but also the impact of power consumption when moving from one state to another before you make a single change.
|