|
Business Challenge
There was a need to develop a software monitoring solution for the large enterprise system.
The enterprise software ran on 70 IBM WebSphere Application Servers and DB2 UDB Databases over a large
geographical and organizational area. The software environment was multiplatform and application servers
were based either on Linux, Windows or IBM Z/OS. To implement the monitoring solution, IBM Autonomic computing
toolkit and IBM Tivoli Common Event Infrastructure (CEI) software were selected by the customer.
Software Solution
Axmor's J2EE developers were hired to implement this project primarily because of our extensive knowledge of
IBM CEI and Autonomic Computing, and also because of our previous experience in automate software monitoring
solutions.
As a solution to this project, Axmor designed and developed a Java custom software solution for logging and
monitoring of the customer's enterprise software system. The solution included three main functional components:
- Logging infrastructure that accepts log messages from a distributed environment on a programming level,
as well as diagnosis data from WebSphere Application Servers and DB2.
- Saving monitored logs into the database.
- Data retrieval and analysis in the centralized CEI runtime.
The main requirements for implementation of the monitoring application were the use of specific technologies,
performance targets, and an ongoing functionality increase.
IBM CEI and Autonomic Computing
IBM Tivoli Common Event Infrastructure (CEI) is a server solution that allows for saving heterogeneous diagnosis
data into the common CBE (Common Base Event) format and for their further integrated analysis. Axmor's Java EE
developers implemented a series of components to send data from the distributed application to the CEI server,
monitor logs, and correlate events in the interaction log view. These tasks were performed with the help of tools
available from the IBM Autonomic Computing Toolkit.
Performance and fault tolerance
The customer had specific requirements regarding performance and fault tolerance of the monitoring solution. It
should be able to handle big logs (3-5 gigabytes per day) without failure. Axmor's Java EE programmers developed
performance testing components to automatically validate these requirements and to monitor software operation
stability.
The customer was completely satisfied with the achieved performance measured using several event-specific metrics.
Iterative development and software improvement
The programmed monitoring solution was developed at the same time when the entire enterprise-level software was
implemented by the customer. The Software system was improved and extended over the time, and requirements to the
monitoring solution were also build up and modified. Axmor's developers provided a customer with a flexible
development environment to integrate such inevitable changes into the final solution.
As a result of this project, Axmor Java EE developers created a full-featured Java custom software solution to
perform automated software monitoring using the well-proven components from the leading software vendor.
Back >>
|