CIOZone.com - Professional Network for CIOs and IT Professionals

By Eileen Haggerty, NetScout Systems

Despite the hype around cloud computing technology for reducing costs and infrastructure in global data centers, Michael Eggebrecht titled his CIOZone article of October 27, 2009: “Green is Hot, Cloud is Not: Data Center Survey.” Leveraging data from a recent survey, he reported -- “Cloud computing may be dominating headlines in the IT world, but only 14.9 percent of data centers are currently using the approach, according to a new survey from AFCOM, a trade association for data center professionals. Another 46.3 percent of data facilities have considered cloud computing and opted not to implement it.”

Given the pressure on IT managers to ensure that services are delivered consistently, it seems likely that one of the reasons for low adoption is the concern over lack of visibility and effectively monitoring and managing service delivery performance in a data centers that are early adopters of cloud computing technology.

The same CIOZone article pointed out that “73 percent of the 436 commercial, government and university data centers polled by AFCOM have implemented virtual processing.” Further, “while some cloud computing technologies may be marketing hype, green IT is very much a reality in the data center,” with 71.3 percent of the respondents indicating they had green initiatives underway of which 42.2 percent were “formal” projects within their organizations.

It seems as though we see the following concepts thrown together as a matter of practice these days -- any article on one will invariably include the others -- data centers, green IT, cloud computing, and virtualization. However, what is not often discussed is “visibility,” or management of application services throughout the virtualized data center or across the cloud. In fact, the very definition of cloud computing has been evolving, to the extent that cloud computing has now become a general reference encompassing an organization’s own data center, their private cloud, a public cloud and/or a hybrid cloud. The concept of “outsourcing” some of the IT processes, such as Software as a Service (SaaS) or storage is an accepted part of the cloud computing definition. And virtualization -- be it server, desktop, or network virtualization, is also typically a characteristic of the definitions.

So let’s explore these elements further, specifically two popular forays into the evolution of the data center and cloud computing -- Server Virtualization and SaaS -- so we can demonstrate both the need for and value of visibility into the complexity of cloud computing.

Opportunities and Challenges of Server Virtualization

According to a recent report from Gartner, 18 percent of server workloads in 2009 will run on virtualized servers, and this will grow to 28 percent in 2010 and to nearly half of all servers by 2012. Why?

As server utilization in traditional dedicated server deployments typically fall in the five to fifteen percent range, organizations have been challenged to uncover ways to better use the over capacity, which has the potential to dramatically alter the financial outlay for data center expansion. Server virtualization has brought with it attractive and compelling benefits. Among them:

Better Overall Performance. Cross-server traffic within a virtualized environment has better performance than across a physical data center network

Efficient Physical Server Capacity. Lots of available capacity left unused historically is now consumed by multiple virtual servers

Cost Savings. Reduced capital expenditures with fewer physical servers required to perform the same functions. Operating costs are simultaneously lowered with fewer maintenance contracts

Green IT Initiatives. Energy and power consumption, air conditioning demand and real estate footprint are all reduced with fewer servers required to perform the same basic functions

Ease of Deployment & Movement. Virtualized applications can move either within the same server or from one physical server to another, located in the same data center or in a different one elsewhere in the world. This dramatically complicates the network management team’s ability to efficiently manage the traffic.

Early adopters have recorded stunning successes of virtual server migration projects saving tens of thousands of dollars in capital expenses by avoiding additional server purchases, reduced operating expenses from lower energy and air conditioning demand, and lower maintenance costs.

Unfortunately, as IT organizations leverage application and server virtualization technologies to improve efficiency, reduce costs and ease the environmental impact of compute resources, there is one very real challenge – virtualized servers are the “new black hole.” They lack essential visibility into the behavior and performance of the very applications that have been virtualized. What was once a rack of dedicated servers interconnected via network switches is now consolidated into a single virtual server hosting many different applications with a virtual switch enabling communications between the virtual machines. Movement of servers within the same machine or to different machines in different data centers can occur making performance management and monitoring of virtualized server environments simply a nightmare.

Opportunities and Challenges of SaaS

According to another Gartner report, “Market Trends: Software as a Service, Worldwide, 2009-2013 - Update,” released in November 2009, SaaS is projected to grow more than 17 percent from roughly $6.4 billion in 2008 to more than $7.5 billion in 2009 -- despite the pressure to cut costs in virtually every other area of IT spending. The report further projects SaaS worldwide revenue to total $14 billion by 2013 for enterprise application markets. Again the question must be asked “why?”

As capital budgets tighten, the ease of deployment and rapid return on investments (ROI) of SaaS are making pay as you go solutions more attractive. Further, SaaS on-demand vendors have made their services more attractive by enhancing security and enriching the actual services with feature rich partner alliances. Some of the more successful ventures thus far are Salesforce.com, a CRM solution, and Webex, a collaboration technology.
The challenges presented by SaaS from a performance management perspective are:

• How will the user’s experience be managed if the service is out-sourced?
• What should the expectations be for performance and response times, once the application service has been outsourced?
• How can IT managers even gain visibility and measure performance?

Despite all savings and benefits of either server virtualization or SaaS, if any of the key critical business applications that deliver revenue for the business degrades or slows, IT staff must troubleshoot, identify, diagnose and restore high quality service for that application. Depending on the business, the application at issue and the length of time the service is disrupted, all the savings realized by an organization could evaporate without effective, unified service delivery management solutions to quickly resolve the problem.

Assuring Application Service Delivery -- From Data Center to Cloud

Added to concerns over lack of visibility and effective monitoring and managing of service delivery performance, the many steps involved in implementing private cloud, hybrid cloud and public cloud alternatives may be negatively impacting adoption rates. IT executives may have very real concerns over whether unified tools or solutions are actually available to manage application services throughout all those steps.

The good news is that there are approaches and solutions available to help IT staff assure application service delivery in their data centers today that will protect initial investments right through to tomorrow’s cloud computing environments. Global organizations have been managing the application performance in data centers using a technologically mature combination of passive, packet-flow based monitoring and sophisticated, in-depth application-level analysis. As the applications shift from individual hardware servers to virtualized servers, or to SaaS solutions, unified service delivery management solutions provide the very visibility into the virtualized server platform and SaaS applications necessary to protect application service delivery.

Leveraging packet-flow data from within the virtualized server platform and using an embedded virtualized performance management data source or software agent, essential metrics can be collected from the virtual machines to analyze and report on intra-server traffic, both in real-time and historically. Further, standard packet-based monitoring and continuous recording data sources strategically deployed to capture SaaS user traffic can add both response time analysis and forensics troubleshooting capabilities that can be shared with third party vendors for collaborating on problem resolution and best practices for optimizing business service delivery.

Generally speaking, these solutions incorporate the data from virtualized servers and SaaS applications to perform unified service delivery management tasks common-place in most IT organizations including:

• Accelerated problem resolution;
• Intelligent early-warning analysis;
• Infrastructure optimization and planning;
• Protection of user experience, and
• Network and application performance management

Complete end-to-end service delivery management solutions for performance assurance in your organization with SaaS applications and / or virtualized server environments should provide:

• A packet flow-based solution for in-depth decode analysis to troubleshoot end-user impacting degradations;
• Both real-time and historical monitoring and trending metrics between virtual machines for optimization and planning;
• Early warning, automated behavioral analytics to isolate the movement of virtual machines, and
• Complete end-to-end views of intra-server activity, inter-server packet flows and data center-wide application traffic, including any and all SaaS applications, to protect overall service delivery of essential business applications.

Whether or not your organization is moving in the direction of private, public or hybrid cloud computing, or adding technology like virtualized servers and/or SaaS, unified monitoring of the application’s performance now and making sure that the organization is getting the most from its upcoming investments should be common sense. And when it comes to ensuring performance - visibility and transparency rule.

Comment on this article

Only registered users can write comments.
Please login or register.