OPTIMIZED SYSTEMS

New Focus on Workload Optimization

Michael Steinhart | Date: 10-21-11 | Comments

In many data centers, inefficiency runs rampant. Systems are provisioned to handle heavy workloads and peak requirements, but they�re sitting idle when demand ebbs. With comprehensive middleware ensuring optimized resource allocation, companies reap numerous benefits in terms of cost savings and higher productivity.

If you�re trying to run a data center or any kind of IT organization, you know that business demands are going up while budgets aren�t. You have to support more data, more applications and more devices than ever. To do this, you have to wrangle all of your servers, storage, and network systems into some kind of unified and manageable platform.

Virtualization and cloud-style computing can help in this regard, but they work best when you have preset configurations for common workloads. What about applications that fluctuate based on time or market demands? Do you provision extra resources so they�re ready for peak demand at all times? Or do you try to monitor usage and assign capacity manually?

This confusion has led to server sprawl where 85 percent of that compute capacity is idle at any given time and 15 percent of servers are running 24/7 without actually being used for anything.

At the end of the day, you�re still responsible to ensure that critical applications function reliably around the clock, that there�s room to grow, and that you�re not wasting a lot of money, time, and power on resources that are sitting idle. So what�s the right approach to maximizing usage while minimizing waste?

This is where workload optimization comes in and why it is getting more attention these days.

Workload optimization matches application characteristics and service-level requirements to the best platform for the job, be it a cluster, a rack or blade server, a grid environment and so on.

Smart management systems let you optimize resource stacks to meet memory, I/O, scalability, security and storage requirements. Different applications have different needs, and workload optimization ensures that middleware and hardware are tuned specifically for those tasks.

It turns out there are many benefits to having a workload-optimized infrastructure:

Virtualization works faster, so you can increase utilization rates and performance without investing heavily in new hardware, and you can reap a higher and faster return on investment from existing infrastructure.

Applications and databases run faster, so you see an uptick in performance and in business.

Workload optimization allows applications to hold more data in memory, which is valuable for business analytics, decision support and other data-intensive applications.

Your hardware and software are integrated tightly, so there are fewer�if any�resource conflicts and compatibility problems.

So it�s clear that a comprehensive middleware layer can make the difference between inefficient, poorly integrated infrastructures that struggle with virtualization, and agile, successful data centers that match applications tightly to the resource stacks they need for optimal performance and efficiency.