Cloud Computing

I recently had the pleasure of discussing cloud infrastructures with Mark Rivington, vice president of technology at Nimsoft.

Here is the insight he had to offer:

Much has already been written and said about cloud computing. Many use cases have been tested, and the performance and cost benefits are now measurable. The notion that cloud computing will soon become an integral part of all IT strategies is evidenced by the fact that most enterprises have either begun to invest in it or have at the very least moved to the starting gate. Cloud computing initiatives are taking off at such a rapid pace that leading analyst groups such as IDC (News - Alert) have forecast that by 2013 enterprises will be investing upwards of $44 billion in cloud-related IT projects.

There remains, however, one looming question on the minds of IT department managers tasked with taking advantage of the cloud on behalf of their enterprises, which is, “How will I manage an infrastructure that is outside of my physical domain?” After all, no one wants to be left without visibility into how mission-critical resources, applications and business services are performing. Fortunately, there is an answer. 

A new breed of management and monitoring technologies that provide real-time insight into the performance of cloud-based applications and services can enable you to achieve control over external IT infrastructures. As an added benefit, a number of these new technologies simplify the overall IT monitoring task by providing visibility of internal and external IT infrastructures through a single, unified view.

Amazon Web Services (AWS), Microsoft Windows Azure, IBM Smart Business cloud solutions, Google (News - Alert) App Engine and Rackspace Cloud are a few of the leading cloud platform providers that have emerged. Each provides enterprises with well-differentiated features and benefits that address a wide range of enterprise computing challenges. As an IT department head, you might choose one cloud for storage, another for application development and another for building a temporary testing environment. Regardless of your use case, the cloud will provide you with on-demand resources that can respond to your organization’s always expanding and contracting computing needs.

After defining your cloud computing objectives, which will likely take advantage of multiple platforms, you will need to find a way to monitor performance and availability, identify problems and hold the providers accountable to Service Level Agreements (SLAs). 

To achieve control, you could choose to rely on the monitoring and reporting technologies that each cloud provider makes available, but this level of trust could place you in a precarious situation. Will the providers’ monitoring mechanisms be accurate? Will they provide the level of flexibility you need to get answers to your specific questions? Will you and your already strapped IT team be able to keep an eye on multiple and disparate performance and availability data streams flowing into your organization? Without a comprehensive monitoring and management solution that can accurately measure and unify all data streams in real-time, chances are the answer to all of these questions will be a resounding “No.” However, with the help of the aforementioned new breed of IT monitoring and management tools, the answer to all of these questions can be turned around 180 degrees, to a confident “Yes.”

The new breed — as mentioned by Yankee Group (News - Alert) in its From Crisis Comes Opportunity: Yankee Group’s 2010 Predictions report — is providing what’s needed to help ensure the success of cloud computing. With these effective monitoring and management solutions in place, delivered through an architecture that enables performance data streams to be consolidated and unified into a single view, IT managers are finding themselves able to easily and efficiently handle a number of tasks:

*Monitor and manage the availability and performance of multiple cloud infrastructures as well as their internal data centers

*Quickly identify and resolve trouble spots and performance problems before they have any major impact on systems and performance

*Measure the performance of cloud resources vs. internal data center resources

*Hold cloud providers responsible for delivering on their SLAs

*Demonstrate which IT resources perform best in the cloud and which perform best within internal data centers

*Determine which cloud environments perform best for their specific IT initiatives

*Measure usage and compare with expected workloads and performance

At this point in the evolution of enterprise computing, there is absolutely no question that cloud computing is viable and an inevitable fact of IT life — providing measurable, unquestionable benefits. If you are being charged with delivering these benefits to your enterprise, remember to select a monitoring and management solution that will enable you to keep tabs on all of the platforms your infrastructure is extended into, and to select one that can unify monitoring of all of your IT resources, regardless of their physical location.