Show TOC

Procedure documentationMonitoring Performance Locate this document in the navigation structure

 

When you want to obtain statistical data about the performance of message processing in the Adapter Engine, you can use the performance monitor. Within this monitor, the performance is gauged by measuring the amount of bytes processed over a given time period.

The performance details provided by the monitor are useful for administrators who need to know the amount of time spent by messages at various points of their life cycle in Adapter Engine. These points (known as measuring points) are at the following stages of message processing:

  • Message is put in the queue.

  • Message is retrieved from the queue.

  • Call to a module in the Adapter Framework.

The performance monitor allows you to choose the time period for which you want to measure the message processing. You can also choose whether you want to see information for all processed messages, or only for messages with common parameters such as the sender or receiver component, interface, or quality of service.

Prerequisites

The user account you use should be assigned to one of the Process Integration monitoring roles. For more information about which roles allow you to view monitoring information, see Monitoring Roles.

Procedure

  1. Access the performance monitor in one of the following ways:

    • Open a browser and go to http://[host]:[port]/pimon to access the Process Integration tools. Then choose   Monitoring   Adapter Engine   Performance Monitor  .

    • Open a browser and go to http://[host]:[port]/nwa to access SAP NetWeaver Administrator. Then choose   SOA   Monitoring   PI Performance Monitoring  .

  2. Choose the period of time for which you want to see the performance data.

  3. Choose the aggregation interval.

    If you want to obtain larger amount of details about the performance of message processing in the Adapter Engine you should choose longer aggregation interval.

  4. Optionally, choose Advanced Selection and enter the parameters of the messages for which you want to see the performance data.

    The parameters you can specify are inbound and outbound channel, sender and receiver party and component, interface and interface namespace, server node, quality of service, and direction.

  5. Choose Go.

Result

The system analyses the message processing data and displays the results in a table. The results are grouped into aggregation entities that are displayed in rows. Each entity contains data about message processing performance in a particular aggregation interval within the specified time period. The system displays only aggregation entities where there is data for at least one message.

The system keeps the performance data in the memory (cache), and writes it to the database on each finished time interval. Consider that it can happen that the performance data is not updated immediately, but only after an interval is finished. The performance monitor only shows completed time intervals and reads the data only from the database.

When you select an entity, the system displays information for the messages in the selected aggregation interval in the Message Scenarios table. The messages in this table are grouped by common parameters; each group is displayed in a separate row. In addition to the basic message parameters, for each group you can see information about the processing time, average processing time, average message size, total message size, and total processing time.

You can select a group to see additional performance details in the Message Processing Details screen area as follows:

  • Choose the Module Information tab to see the amount of time spent by the messages in the selected group at the measuring points.

  • Choose the Message Information tab to see the statistical data for the minimum, average, and maximum message size and number of retries.

More Information

You can manage the collection of data using the following parameters of the XPI Service: AF Core service:

  • You can switch the data collection on and off using the profile.performance.runtime parameter.

  • The background job for data aggregation runs periodically. You can control the length of the period at which the system runs this job using the profile.runtime.interval parameter.

For more information, see the “Properties for XPI Service: AF Core” topic in the documentation for configuring Process Integration.