Share This
By Katherine Chalmers
May 2006
The 7 Steps to Business Service Management
The new imperative in IT management is the ability to manage the technology infrastructure within the context of the business services that depend on it. Whether called Business Service Management (BSM) or end-to-end IT Service Management (ITSM), the goal is to understand how the health of individual components within a service’s ecosystem relates to users’ quality of service (QoS) and the company’s business requirements. The quest for BSM can be difficult. Not only is understanding the interdependencies between system components confusing, but battling organizational barriers to cut across internal technology silos can be formidable.
So how does a busy IT organization begin the quest? One of the most popular best practice frameworks is the Information Technology Infrastructure Library (ITIL), which outlines extensively what technology departments should do, but not how to do it. Several vendors offer proprietary implementation maps based on their own products, but at BSM Digest, we’ve taken a vendor-neutral approach to the problem and adapted this simple seven-step framework to ease the way.
Step 1: Inventory
Identify and catalog the elements in your infrastructure
If you don’t know what you have, how can you manage it appropriately? For most companies, the first step toward business service management is a thorough accounting of the assets (or configuration items) in their infrastructure. New innovations in automated discovery technologies are making it easier to identify and catalog elements within complex infrastructures. Many companies have opted to federate this inventory along with configuration information and basic component dependencies into a configuration management database (CMDB), one of the fundamental pillars of ITIL.
Step 2: Instrument
Properly instrument elements for continuous monitoring
Once you have an inventory of all the configuration items in the environment, the next task is to make sure they are all properly instrumented for continuous monitoring. Monitoring technology has evolved dramatically from basic availability management – simple up/down notifications – to more elaborate performance management products that can track hundreds of key performance indicators for each individual system. Plus, a new generation of end-user monitoring and transaction products are making it easier than ever to understand user experience and quality of service.
Proper instrumentation requires two phases of implementation. First is the deployment of instrument applications and distribution of their monitoring agents (or the new-breed agentless monitors). Second is the active administration of baseline and threshold settings for every monitored metric on each configuration item. Traditionally administrating monitoring applications has been very time-consuming, but new BSM tools are automating many of these previously manual tasks.
Step 3: Model
Map services and applications to the elements that support them
Comprehensive inventory and instrumentation allows you to begin pulling together the elements of your infrastructure in ways that make sense for managing your services. Within the modeling process, IT mangers begin to map applications to the network elements that support them and the performance dependencies between them.
In the same way that a doctor analyzes multiple biologic measurements – temperature, heart rate, blood levels, cholesterol, etc. – in context to one another to assess how a patient’s organs are performing and thereby diagnose his overall health, a network manager can model groups of system elements to diagnose the health of key business services. This process of modeling services is fundamental for creating service level agreements tied to business requirements.
Step 4: Analyze
Understand normal behavior patterns and detect performance anomalies
Analysis techniques and technologies allow IT managers to understand the normal operating behaviors of the systems and the services they support. Accurate analysis enables them to establish policies and event rules that notify them of conditions that portend service degradations. As the culmination of the Inventory, Instrumentation, and Modeling steps, effective analysis allows the timely detection of system and service anomalies and the immediate assessment of their impact on quality of service. Until recently, analysis has been a time-consuming process bottleneck for many organizations, but new automated real-time analysis technologies are streamlining this step.
Whether done manually or automated with new BSM technologies, timely analysis is a critical step for successfully managing IT infrastructure within the context of critical business services.
Step 5: Visualize
Gauge your environment’s end-to-end status in real time
A new breed of BSM tools is emerging that finally makes it possible for CIOs to maintain real-time dashboards that track the end-to-end health of their critical business services. These applications provide easy-to-read views of configuration items and consolidated event management consoles to facilitate administrators’ troubleshooting responsibilities. As infrastructure environments grow more complex, the importance of having simple user interfaces for visualizing them increases. IT operations teams and their business customers need intuitive ways to quickly assess service health and pinpoint developing problems before they impact service quality.
Step 6: Troubleshoot
Identify and resolve problems quickly
When performance anomalies and service degradation occur, system and application administrators must effectively manage incidents and analyze their root causes. This goal is only achievable if the flow of incoming alerts from the monitoring infrastructure is small and qualified. While managing events has historically been a slow, hand-picking process, new innovations in dynamic thresholding and multi-variable data correlation are significantly helping increase the accuracy and manageability of system alerts.
Once the true problems are identified, system administrators still need the ability to rapidly drill-down to the source of the problem and trigger the appropriate remedial actions. Root-cause analysis remains the biggest challenge of Incident Management. Fortunately, vendors are now introducing tools that offer accurate root-cause analysis capabilities via integrated end-to-end fault and performance management, advanced analytics, and rule-based techniques.
Effective reduction in Mean-Time-To-Repair (MTTR) relies on accurate problem description, routing to the appropriate technical support tier group, and prompt resolution. Building a known incident database based on past experience and letting administrators consult these known incidents during troubleshooting can drastically reduce MTTR. This requires diligent incident post-mortem sessions where subject-matter experts record the symptoms of the problem, spell out its root-cause, and explain how it was resolved. A tight integration with trouble ticketing systems using intelligent filtering and routing capabilities is also a contributor to efficient incident and problem management.
With each innovation, the industry is moving closer than ever to its long-term objective of automating problem diagnosis and trouble ticket generation and its vision of autonomic, self-healing systems and automated service provisioning.
Step 7: Plan
Enhance IT and business objective synchronization over time
Ultimately, the goal of every IT organization is to minimize the number of problems affecting systems and services, not just addressing incidents in an effective manner. There are two ways to meet this objective: proactive monitoring and proactive capacity, change, and configuration management. Effective planning is the key to achieving both.
Proactive monitoring uses short and medium term forecasting capabilities to identify performance degradations before they become incidents, and then promptly alert on them. System and service behavior modeling technologies typically include predictive engines that allow you to forecast the occurrence of a problem with a high level of accuracy. However, this technology will only be efficient if the back-end support processes are capable of analyzing and reacting fast enough.
Proactive capacity management refers to the ability to accurately analyze the infrastructure’s current load, identify trends, and take the necessary configuration management steps to increase computing capacity on over-loaded systems or implement efficient load-balancing.
Our Prediction
Given the accelerating pace of innovations in BSM technologies and companies’ commitments to achieving end-to-end IT service management, we predict that managing technology within the context of business service requirements will soon become business as usual, not an arduous quest.
Our 7 Steps to BSM interactive diagram illustrates key tasks within each step. In future BSM Digest issues, we will be featuring informative articles and expert interviews to help you successfully navigate each step.
