BSM: Mission Impossible
The Emerging Importance of Analytics in Business Service Management

The Four Basic BSM Challenges

Handling Data Volume
Today’s monitoring tools can track dozens or even hundreds of performance metrics for each device and application in an environment. In complex environments with thousands of monitored elements these volumes of data and alerts can be overwhelming.

Sifting Through Variation
Even in moderately dynamic environments, “normal” performance for any given device can change at different times of the day, week or month. This variability makes manually setting alarm thresholds or developing custom scripts for each device unfeasible. Without automated, scalable means of analysis, even the most experienced IT managers are unable to leverage the collected data to its full potential.

Understanding Complex Relationships
One of the reasons managing enterprise IT is so complex is the deep relationships between systems, applications and devices. It is virtually impossible for IT managers to assess in real time how dozens or even hundreds of the components needed to deliver a critical business service are performing.

Keeping up with Changes
What makes the analytics challenge infinitely more difficult is the fact that environments continue to evolve at an increasingly rapid pace. Devices become obsolete. New devices are added. New relationships are developed. The bottom line is that the other challenges – volume, variability and relationships – change before managers can even get a handle on them.

In the mission to implement business service management, many organizations have made an unwelcome discovery: achieving full, business-focused, end-to-end IT service management is very hard to get right.Why is BSM so difficult? Too much complexity.

“Originally, the focus of the industry was on managing individual devices. As the industry has evolved, however, the requirements have evolved,” explains Dennis Drogseth, Vice President, Enterprise Management Associates.

“Now, instead of managing fairly simple individual devices, the requirement is to manage interdependencies across networks, systems and applications,” he continues. “The capabilities for understanding those interdependencies – to pinpoint a problem that could be anywhere from a datacenter in Chicago to a switch in India; and doing that in real time –- requires a much more adaptive set of analytics than the traditional approach.”

With nearly every facet of a company’s value chain dependent on information technology today, IT organizations often provide and support hundreds – even thousands – of critical IT services. All of them rely on a variety of interrelated components to operate effectively: applications, databases, servers, networks, firewalls and more. For each component there are dozens of individual performance measurements to understand, which can add up to thousands for a given service. This is why advanced analytics technologies are emerging as an important element for successful BSM implementations.

Enterprise Management Associates has been tracking this emerging trend and will publish the first analyst research report on analytic technology adoption in IT management in October 2006.

Facing the BSM Generation Gap

Lacking advanced analytics capabilities, conventional BSM tools are limited in their ability to handle the complex issues and massive infrastructure data generated in today’s enterprise. At the component level, these products allow users to set custom thresholds or write scripts to manage data and alert volumes for each logical element. At the service level, users manually organize devices and applications into groups and establish dependency rules for basic service modeling.

Theoretically this approach enables the BSM tools to monitor the overall performance for each service. In reality there are far too many fluctuating data streams for the human beings to process manually. Consequently, the resulting thresholds and static rules lack the sophistication and granularity to support true business service management.

“Many of these first-generation BSM tools relied on manually designed and maintained business service models populated by component level IT status data, information from synthetic transactions, and other input selected to help approximate IT’s impact on business performance,” says Mary Johnston Turner, Vice President of Ovum Summit. “They generated reports that were more likely to be periodic snapshots than real-time assessments and they were only as good as the business service models used to drive the reports and assessments. Most of these models, which associate a myriad of underlying systems and assets with the top-level business services, had to be manually developed by administrators using complicated design tools.”

“Many first-generation BSM tools failed to live up to their business and IT promises,” Turner adds. “The time and effort required to implement and maintain them caused customers to limit the range of services modeled. Furthermore, the time delay between data collection and performance reporting limited their use to historic trend analysis rather than real-time SLA compliance tracking.”

Taking BSM to the Next Level

“BSM without automation and analytic capabilities becomes a grand vision that will sink under the weight of its own administrative burden,” Drogseth predicts. “EMA forecasts that IT process automation will become a key growth area in the IT industry for the next 5-10 years. Analytics is a critical technology for enabling that, and other types of automation.”

A new generation of BSM tools is addressing the need for automated real-time analysis that can adapt to a complex and ever-changing IT infrastructure.

“Unlike the products of the 1990s, these newer tools have less reliance on manual modeling and historic trend analysis,” Mary Johnston Turner notes. “Instead, they emphasize the ability to dynamically connect IT system health assessments to real-time end-to-end business process and application performance.”

One breakthrough solution making an impact in analytics for BSM is Netuitive. “We leverage the data users already get from their existing monitoring tools by applying advanced multivariate statistical analysis to automatically create self-configuring, self-learning, and continuously adaptive profiles of the devices and services within the enterprise,” explains Nicola Sanna, Netuitive’s President and CEO. “Because our solution is mathematical, our customers can include any data feed – business metrics, infrastructure monitoring data, etc. – for real-time, actionable information about their services.”

Netuitive represents the forefront of a growing trend in BSM – the emergence of second-generation tools with advanced analytics capabilities that are giving new strength to BSM technology. For many organizations, this new functionality is proving to be the missing link that makes end-to-end service management a “mission possible”.

Peter Goldin is a BSM Digest contributing writer specializing in the technology and electronics markets. During 17 years of marketing and PR writing, Goldin has covered performance monitoring and management, storage, networking, security, e-commerce, and a range of other technologies for respected corporations and innovative startups, as well as a variety of publications.

Related Sites

Enterprise Management Associates
EMA is the only industry analyst and consulting firm dedicated to issues of IT Management. The firm conducts comprehensive, in-depth research and analysis on current and emerging concepts, issues, trends, strategies and resources. EMA consults with enterprise IT professionals to assess their organization’s current IT management infrastructure, skills, efficiency and effectiveness. www.enterprisemanagement.com

Ovum Summit
Ovum Summit, a Boston-based division of Ovum, PLC, tracks and enables the IT industry adoption of dynamic computing technologies including virtualization, SOA and ITSM. www.ovum.com