Operating Systems

If there’s one thing certain these days, it’s that workloads in data centers are changing. Not long ago, workload optimization at its purest and simplest was focused around transaction throughput and compressing batch processing windows. Now, the introduction of cloud-based services and a new set of computing requirements that extend to business analytics and even to High-Performance Computing (HPC) is ushering in transformational changes to IT infrastructure and how workloads are viewed.

The ultimate “grand vision” is a day when enterprise IT can monitor, provision, and tune resources without caring whether the resources exist in private clouds, public clouds, hybrid clouds, or even traditional data centers. In this world, there will no longer be a fear of losing visibility of enterprise applications or transactions, no matter how many computing platforms and cloud boundaries they must cross. Technology vendors and IT decision-makers have this vision in view, but the reality is, we aren’t quite there yet. Consequently, today’s tasks for CIOs and senior IT managers are to revisit their internal roadmaps for enterprise computing, and to lay in the best possible courses to meet rapidly emerging business needs. These needs almost certainly will require a cloud computing strategy that addresses the handling of non-traditional transactions such as social networking and the breakdown and analysis of Big Data.

The Road Enterprises Are Traveling

Because enterprises recognize that the dynamics and even the definition of workloads are changing, key IT decision-makers are determining what their infrastructure strategies will be. But just like other technology transformations, there are always two enterprise IT camps: those that understand they must rethink the way they’re defining their mission-critical workloads and workload performance, and those that are simply trying to keep pace with what they already have on their plate, with few extra resources or time left for futuristic thinking.

Regardless of the approaches enterprises take as they look at new workload demands, virtually everyone understands that workloads are becoming more diversified. IT is moving away from jobs strictly structured around transaction processing and batch processing, and are now incorporating (or looking to incorporate) new processing elements such as business analytics, cloud performance, and the handling of Big Data.

Enterprises are also recognizing that many of the tried and proven tools they’ve used to manage workloads over the years in their data centers also must change. In the face of rapid change, selecting tools becomes complicated because vendor competition is fierce and the variety of product offerings can be overwhelming. This has prompted many enterprises to simplify their data center management approaches by consolidating their tools into a few best-of-class solutions. The bottom line for most enterprises at this point is:

• There’s a need to balance the goals of architecture/platform simplification against the value of maintaining infrastructure diversity, so IT can offer the right platform for the job when deciding where to run a growing variety of application workloads.
• There’s a need for more automation in the data center that can manage traditional applications, new cloud implementations, and even Big Data.

The Evolution of Workloads

As enterprises re-assess how they’re running their workloads, these workloads are continuing to move through an evolutionary and transformative process. This evolution is proceeding through three key stages:

1. A focus on production and reliability where transaction throughput and the timely execution of specific workloads in prescribed windows of time are the orders of the day and concentration is on the building of job schedules and runbooks (akin to traditional data center computing)
2. Growing sophistication as sites move into flexibility and adaptability for provisioning virtual servers, often in a private cloud deployment
3. A move into autonomic computing, where automation adapts processes and IT infrastructure to support changes in business computing needs and also introduces new processes at greater levels of integration, such as the IT service desk.

Greater data center automation for workload management, once the exclusive province of only the largest enterprises, is also working its way into smaller companies as everyone feels the need to implement private clouds with on-demand, automated provisioning. One important driver of automation is the pressure sites now face when they’re supporting many different applications and systems from different vendors. This variety of solutions pushes the need for data center integration and automation to where it must happen sooner than later—regardless of business size.

Addressing Cloud and Big Data

As workload transformation occurs, large enterprises and even Small and Medium-Sized Businesses (SMBs) have overwhelmingly embraced private cloud as their cloud implementation of choice. The need for “ownership” over the cloud and to know that security and data protection are sufficient for the business are major drivers of private cloud adoption. But at the same time, there are also isolated applications of public cloud solutions that are occurring, and they must be factored into the IT workload mix.

2 Pages