Locating the performance problem in the stack

There are many components involved in performing a specific task, which can influence the performance severely. This needs to be kept in mind while trying to locate a performance problem. What needs to be checked will also depend on the type of use-case. If some scheduled jobs activity (reports, imports or exports) is slow, you mainly have to focus at first on Product Master side checks with special focus on the scheduler services log files and activity. But if working with the user interface is slow, then you have to include client side checks as well in your investigations.

Understand the Product Master architecture, see the following:
Architecture of Product Master and Product overview.
A three-tier architecture is used that consists of the following:
  • a web-based user interface for rendering Product Master content in a web browser on the client side
  • a middle tier with the functional modules and services that process user requests and data on the central Product Master server
    Note: The middle tier consists of 6 services which are implemented as separate JVM processes. Each service is responsible for specific tasks and logs debugging or error messages in its own directory under $TOP/logs. The major services are:
    • the application service, (process appsvr?_<hostname>), it handles user interaction with the system.
    • the scheduler service, (process scheduler?_<hostname>), it processes all of the scheduled jobs like imports, exports, and reports.
    • workflow service (process workflowengine_<hostname>), it handles processing of items through the workflow.
  • a database management system, for example, Db2 Oracle, that stores the data
All three tiers, if located on different servers, use the network to transfer data and requests.