Availability Guide for Problem Management

Monitoring Event Messages
Availability Guide for Problem Management–125509
4-14
Step 2—Analyzing Application Event Messages
Using EMS FastStart to Develop and Test EMS Event Messages
EMS FastStart is a TACL-based code generator that generates and compiles a number of
source files that are used to simplify event generation and testing. EMS FastStart
enhances development of applications by providing a simple, cost-effective way for
programmers to develop and test EMS event messages.
EMS FastStart uses a configuration file to generate a module that is bound into the user
application. This module, called EGEN, serves as an interface between the user
application and EMS. This enables programmers to generate events in the same syntax
as the user application.
Step 2—Analyzing Application Event Messages
For each application, you need to analyze the event messages, select the important
messages, define their severity, and document recovery steps for critical messages.
Using EMS Analyzer to Analyze Events
You can use EMS Analyzer to examine and analyze information from a variety of
applications: operations, online, point-of-sale, and batch.
EMS Analyzer can read all EMS events generated by your application. Using EMS
Analyzer, you can generate reports to create a profile of the application and evaluate the
number and type of messages being generated.
You can download EMS Analyzer data to a PC, where you can use a spreadsheet tool
such as Excel to create graphs and charts that illustrate event information. Figure 4-2
shows a sample chart of EMS Analyzer data, created with a spreadsheet tool, that
represents a typical week of events generated by operations, point-of-sale, and batch
applications. In this sample, batch applications generated more than 80 percent of the
event messages. One process generated over 15 percent of all batch event messages, and
six processes generated 50 percent of all batch event messages.
Using the information provided in EMS Analyzer reports, you can identify the
application processes that generate the most event messages. This information can help
applications programmers to make minor changes in application programs and thus
reduce the number of event messages. This will allow operators to focus on critical
events instead of having to react to every event.
Note. For a detailed discussion of how to instrument new applications, see “Instrumenting
Applications for Effective Event Management,” written by Jean Dagenais in the October 1991
issue of the Tandem Systems Review.