Error Handling and Logging for Services

All errors are inspected for severity at the plug-in level and a return code indicating success or failure is returned. In most cases, if no errors occur during runtime, the plug-in returns a flag indicating success. If an error occurs at runtime, the plug-in logs the error and returns a flag indicating failure. At that point, the engine can route that message to any error handler and continue serving requests. Service errors are logged in the sybesp.log file.

Services has five levels of logging that are used to determine what is written to the service's log file. These levels along with other logging features are set in the Orchestration Administration Properties view. See Reviewing and Editing Configuration Properties for the procedure to set the properties.

This lists them in order of severity and amount of information contained in the log file:

The logging levels are set globally but can be overridden at specific service instance levels. Configuring the settings for the specific instance at a service level performs the overwrite.

The Services plugin uses Log4J to assist technical support with analysis of problems.

In sybesp.log, logging and tracing messages are presented in the following format:

Note: The message is also written to the NNSYMessageLog.nml file. See Message Log Contents for information on the log file contents.

You can view instance logs using the Show Instance Log option available for each runtime instance in the Server Explorer.