Concepts

The following topics provide information about the components and concepts behind IBM webMethods BPM Optimize business activity monitoring. It describes the architecture of the Optimize implementation, and the processes by which data is collected and fed into the Optimize application. In addition, it discusses the interaction between Key Performance Indicators (KPIs), operational data, business processes, and rules in analyzing and displaying collected data. Finally, it describes the tasks that system and business users perform and the roles that users can be assigned.

Optimize supports various levels of Business Activity Monitoring (BAM) depending on customer needs. Business activity monitoring enables you to analyze real time business metrics information, including system performance, the volume of business activity and its responsiveness, serious errors that may have occurred, and other key performance indicators (KPIs). Aggregate analytics working in parallel with instance-level data provide actionable information that you can use to eliminate problems, increase efficiency, and take advantage of business opportunities.

Optimize uses special data collectors to gather data from other webMethods components that is used to support business activity monitoring. Note that there is no specific Optimize interface; rather, Optimize is an implementation of webMethods functionality that is available through the standard My webMethods user interface.

Two different implementations of Optimize are available to webMethods customers: Optimize for Infrastructure and Optimize for Process. These implementations are explained below.

• Optimize for Infrastructure

  This implementation enables you to monitor system data in real time. System data consists of data about equipment and applications, such as queue length and whether a managed webMethods component is online or offline. You can monitor individual objects as well as the overall status of your system. Optimize for Infrastructure contains two components: Infrastructure Data Collector and Optimize. Optimize for Infrastructure uses the Infrastructure Data Collector to collect system data.

• Optimize for Process.

  This implementation enables you to monitor business processes, in addition to managed objects and system status. You can track key performance indicators (KPIs) that relate to operational factors such as margin, revenue, customer satisfaction, and inventory levels. Optimize enables you to compare actual performance with rules you establish. If a monitored item goes out of compliance with a rule, Optimize alerts you so that you can address the problem. In addition, you can use Six Sigma to analyze your business processes. Six Sigma is a disciplined methodology for improving business process performance by eliminating defects. Using this methodology, you can measure the number of defects in a process and systematically determine how to eliminate them. Optimize for Process also includes business visualization tools that provide at-a-glance mashups of Optimize monitoring information.

The following table describes the five main functions that the Optimize implementation performs.

Each subcomponent can reside on its own separate host, or all subcomponents can reside on the same host.

Typical Optimize implementation:

The Installation and Configuration checklists have been moved to Appendix A in Configuring BAM.

Data collected from many parts of the enterprise infrastructure is fed into the Optimize environment and stored. KPIs that represent information vital to the business are defined and applied to the data. These KPIs are analyzed in real time in order to learn normal behavior and detect abnormal behavior within the environment. Rules are applied to these KPIs and their analyses in order to define conditions that represent problems and opportunities. Violations of these rules can be reported to the appropriate people and automated corrective actions can be taken. Further detail can be provided about a particular violation to determine the true root cause of the situation. Optimize also allows you to correlate information from disparate parts of the enterprise, such as the system and infrastructure, business processes and business data, and services. In addition, it provides analytic mashups, which are graphical overlays of correlated data such as business process information overlaid with real-time statistics.

Optimize collects data from other webMethods applications, through the Infrastructure Data Collector, or through the Web Service Data Collector. The following diagram illustrates how Optimize collects, defines, stores, and analyzes data.

The Infrastructure Data Collector is the primary method for collecting infrastructure and SNMP data. The Infrastructure Data Collector uses the Monitoring API to pass operational data and monitoring model data to the Analytic Engine. The Monitoring API is designed as a programmatic interface to the systems monitoring functionality of a business integration platform. It communicates over JMS, utilizing a JMS queue, to provide secure and reliable communications between Optimize and the Infrastructure Data Collector.

The Infrastructure Data Collector lets you identify webMethods component resources that you want to monitor and analyze. Using the Infrastructure Data Collector, you can monitor the following:

• Adabas (Deprecated) or Natural (Deprecated) Component assets
• Apama (Deprecated) Server assets
• ApplinX, EntireX, Adabas SOA Gateway (Deprecated), Com-plete (Deprecated) or Natural Ajax (Deprecated) Component assets
• Broker Server assets
• Event Routing Server assets (deprecated)
• Integration Server assets
• IBM My webMethods Server assets
• MashZone NextGen (Deprecated) Server assets
• SNMP Component assets
• Terracotta Server Server and Terracotta Server Array assets
• Universal Messaging assets.

For more information about discovering and monitoring components, see Discovering Assets and Monitoring Components.

The Web Service Data Collector is installed automatically with Optimize. Web services provide an interface to Optimize through which client programs can send operational data about applications or equipment in your enterprise (such as databases, printers, or disk drives), process data (such as the start and end time of each step in process instances), or business data (such as customers, order quantities, and revenues). Optimize then displays the data in the form of system KPI instances for operational data or business KPI instances for business process data.

Depending on the size of your enterprise, you can distribute the load by deploying Web services to systems that host applications and processes you want to monitor. Web services gather the summarized data and send them to the Analytic Engine, which in turn stores the data in the database and displays it in the My webMethods user interface. In addition, the Analytic Engine runs algorithms on the data to produce a variety of diagnoses. These diagnoses enable Optimize to automatically and continually compare summarized data with a baseline of historical behavior and indicate when types of data are deviating from that baseline.

If you have IBM webMethods Integration Server and Designer, you can build Integration Server services that use Web services to gather data for Optimize. For example, you could do any of the following:

• You can use Integration Server services to gather data about webMethods components. Suppose you are using webMethods Broker (deprecated). You could create an Integration Server service that uses Web services to create system KPI instances for types of data in certain Broker documents, listens for those Broker documents, gets the appropriate field values from the documents, and uses Web services to push the data to the created system KPI instances in Optimize. For more information, see Configuring and Using the Web Service Data Collector.
• You can use webMethods adapters with Integration Server services to gather data about resources in your enterprise. Suppose your enterprise includes an SAP system and you use the webMethods SAP Adapter to connect that resource to your webMethods components. You could create an Integration Server service that uses Web services to create system KPI instances for types of data relating to the SAP system, uses the SAP Adapter to query the SAP system for the data at regular intervals, and uses Web services to push the data to the created system KPI instances in Optimize.
• In Optimize for Process, you can use Web services to gather data about business processes in your enterprise. Suppose you want to monitor a business process. You would use Web services to push data about each process instance to Optimize as the process instance executes. Typically, you would push data about the start and end of each step in the process instance and about process run-time events that might indicate that an error or a warning occurred during process execution. You could obtain that data from a log file, from a database, or from the business process code itself.
• Also in Optimize for Process, you can use Web services to gather data about your enterprise's business transactions. If you are monitoring business data, you would use Web services to push data about each transaction instance to Optimize as the instance executes. Typically, you would push data about customers, products, order quantities, and revenue.

You can monitor quantitative business and system data by defining key performance indicators (KPIs). KPIs are quantifiable measurements that reflect the critical success factors of an organization. Optimize uses KPIs to monitor data for exceptions and trends and to help you answer questions such as the following:

• What is the queue length for this operating system?
• How many orders over $10,000 have completed for this process? (Optimize for Process)
• What is the revenue for the Southeast Region? (Optimize for Process)

For example, when using Optimize for Process, in an order management process, you might define KPIs for how many orders were received, their dollar amounts, and whether they were processed successfully. You can also define KPIs that compare volume by order channel or cycle time by customer type.

In Optimize, KPIs consist of the following things:

• Dimensions, such as customer, region, dealer, or sales person
• Event maps, which associate business data, such as dimensions, measures (measurable and monitorable numeric quantities such as revenue), and transactional attributes, with a particular business process
• KPI definitions, which associate a measure with a dimension (a means for breaking out your data into slices), such as revenue by region or queue length by host
• Hierarchies, which determine how dimensions are related. Optimize uses these relationships to aggregate and organize data on My webMethods business process and system monitoring pages.

Optimize for Process can analyze and display status information for the following two types of business processes:

• webMethods-executed processes (processes that are executed and tracked by the IBM webMethods BPM Process Engine).
• Externally-executed processes (processes that are not executed and tracked by the IBM webMethods BPM Process Engine). For externally-executed processes, Optimize has no knowledge of the rules and data that determine whether these transitions are valid. Therefore, for these processes, Optimize does not recognize invalid step transitions based on branch or split conditions and does not detect or report errors when such conditions occur.

You can use IBM webMethods Designer to design business process models. For more information about using IBM webMethods Designer to set up business processes and process steps, see the IBM webMethods Designer online help. For more information about setting up and identifying those business processes you want Optimize to monitor, see Configuring Business Processes.

Optimize evaluates data against rules you define. A rule identifies a significant condition under which a process or resource becomes problematic, requires attention, or is of interest. A rule can be configured to alert users or take programmatic action when the condition occurs. Rules can also be configured to send user alerts when the process or resource returns to compliance. You can define KPI rules, event rules and threshold rules.

When you define a rule, you identify the type of data to which to apply the rule (for example, Broker queue length). By default, Optimize then applies the rule to that type of data on all processes or applications (for example, all Brokers in your entire environment). However, you might want to apply the rule to that type of data for specific processes or applications only (for example, for certain Brokers only). In this case, you create a rule filter that identifies only the specific processes or applications. When you define the rule, you associate the rule filter with the rule.

During rule definition, you also identify the users to whom to send alerts when the rule is violated, or you can specify other actions to perform when a rule violation occurs (for example, send an SNMP alert to an SNMP manager or invoke a Web service).

Optimize has two types of users: administrators and business analysts. Note that “business analyst” is a type of user that exists only in the context of Optimize for Process.