Newsletters

When faced with problematic IT services or multi-tiered applications, analysts need to zero in on the component causing the problem - regardless of where it may reside within the enterprise. Analysts, using proper capacity management software, can take data from thousands of servers, and slice and dice this information in order to trace issues to their underlying causes. Let’s take a look at monitoring service performance within the SOA framework.

Delve into SOA
A service-oriented architecture (SOA) is a collection of services that communicate with each other; are self contained, and do not depend on the context or state of other services. They work within a distributed systems architecture as follows:

• Component - an individual executable module that performs a single, measurable task.
• Service - a collection of components that provide a unique business service such as an order processing service or a general ledger service.
• Processes - a collection of components and services that are invoked in a particular sequence in order to mirror business processes.

SOA Concept
Let's take a closer look at the concept of a component within the SOA framework. A component is a smaller group of objects working together to provide a system function. For example, a claim, an automobile and a claimant object can work together within a claims component to provide the claim function for a large insurance application. The claim component becomes a black box, however, at the level of a large system function i.e. the claim and automobile objects are not known to any other part of the system, except for the claim component. This means that no other part of the system can become dependent on these objects since they are completely hidden.

Contract design between components, then, is a critical activity in a service-oriented architecture. The difference between a public interface and a published interface comes into play here as an important distinction. A public interface is one that can be used by components within a system. The public interfaces of a component are easier to change, because they are only used by known clients. A published interface, on the other hand, cannot be changed so easily as it is exposed to the network.

It should be understood, that there is no conflict between SOA and IT Infrastructure Library (ITIL). ITIL processes that support the goals of SOA include Service Level Management (SLM) (what services are being provided at what level), Capacity Management (where is the organization heading in terms of growth), Financial Management for IT (to ensure responsible spending is taking place), Availability Management (to keep applications up and running) and Service Continuity Management (for disaster recovery purposes). Any and all service delivery processes in the enterprise must provide high-quality, cost-effective services to the customer. This includes measuring and reporting on performance within the SOA. Thus ITIL and SOA complement each other.

IT Resources Concept
The best way to understand how, for example, TeamQuest IT Services Analyzer fits in with SOA is to begin with the definition of an IT Resource.

An IT Resource identifies a physical or logical component or group of components within a data center or IT organization. Resources are typically computer and related equipment, software, facilities, or organizational. IT Resources are the center piece of the latest TeamQuest solutions. They make it possible to conduct even more powerful analyses and reports.

IT Resources come in a variety of granularities and can represent anything across the infrastructure from individual components to entire systems. They can represent physical or logical components, and groupings of IT Resources. In addition, they can be linked to other IT Resources to represent relationships.

Higher-level IT Resources, can represent all systems. Below that, IT Resources can be subdivided into administrative groups. Say an IT Resource for all AIX systems encompasses three servers. It would have its own name (AIX in this case) within TeamQuest IT Service Analyzer, and each of the three servers would have their own icon and name. One might be called dbA.

Another IT Resource named Services might be comprised of a grouping of two other IT Resources named Order Entry and Payment Processing. These two IT Resources, in turn, have two systems that are associated with them. One of these could be dbA, as mentioned in the previous example. Thus an IT Resource may exist in multiple views and multiple groupings. It is also possible to drill down to even lower-level IT Resources such as CPUs, disks and applications.

Finally, it might make sense to look at the IT Resources discussed from a different perspective. If there was an alarm received by the IT Resource named dbA, TeamQuest IT Service Analyzer quickly identifies what services are affected by that server. In this case, two services are affected.

Complementing SOA with IT Service Analyzer
To log onto IT Service Analyzer requires a logon and password. The user has the choice of either IT Service Analyzer or IT Service Reporter. For the purposes of this paper, only the former will be discussed. Once IT Service Analyzer is selected, it is possible to group IT Resources in terms of system types, i.e., Windows, HP, Sun and AIX servers, for example. If an analyst then wants to see the IT Resources for all UNIX servers or one UNIX server, it is a simple matter of dragging and dropping the appropriate icon in the IT Service Analyzer report area. This is displayed in your web browser. TeamQuest IT Service Analyzer can also be used to analyze SOA components and understand their relationships. Take, for example, a process such as a credit card application. It might have associated services such as Background Check and Authorization.

The Background Check service consists of a number of components such as Customer Look Up, Review Account History, Review Credit History and External Credit Verification. The Authorization service, too, has several components: Set Credit Limit, Assign Credit Card Number and Issue Card.

TeamQuest IT Service Analyzer helps anyone connected with the credit card application process in analyzing every facet of this process, how it fits into the broader SOA picture, and to drill into each component to discern the reasons for any problems with this particular process. Further, IT can better help users understand the relationships that exist between the different aspects of this process, as well as how they relate to other systems and processes. This might prove useful, for example, in identifying bottlenecks that are preventing the reliable delivery of a specific service.

In addition, TeamQuest IT Service Analyzer can be used to define IT Resources and create new IT Resources. Due to its use of rich web technology, it is also possible to drag and drop pre-defined IT Resource types as well as business unit details onto the Analyzer screen. CDB or CMDB data can be imported into TeamQuest IT Service Analyzer via XML format, minimizing duplicate entry and maintenance.

By doing so, an administrator can rapidly build a hierarchy that fully supports SOA. Thus he or she can determine where data resides within the entire credit card application process. This includes which servers store customer data, which server deals with external credit verification, and how other components and services relate to the underlying IT infrastructure. This might point up, for example, that one server is overloaded and that this is causing delays on the credit verification side.

Analysts can drill into any SOA process to either determine the reason for a slowdown, or optimize service delivery. By looking at CPU utilization on the server or servers used by a process or service, for instance, a bottleneck can be quickly identified. By clicking on the screen beside the bottleneck, it is possible to see what applications or services are affected and exactly where the contention exists.

Within SOA, TeamQuest IT Service Analyzer is a fundamental tool in isolating the causes of performance problems. It facilitates the creation of IT Resources for services used in SOA, the organization of those services, reporting on combined/individual services and drilling down into component issues.