Software AG Products 10.11 | Administering Integration Server | The Statistics Log | Viewing the Statistics Log
 
Viewing the Statistics Log
The statistics log contains statistical information about the consumption of system resources and Integration Server resources. Integration Server saves the statistics log to the following directory: Integration Server_directory \instances\instanceName\logs.
Unlike other logs, you cannot view the statistics log under the Logs menu in Integration Server Administrator. However, much of the statistics log is also displayed in Server > Statistics page or on the Dashboard.
Depending on the value of the watt.server.stats.logfile.csv parameter, the statistics log is either a comma-separated values file (csv) or hexadecimal. When the statistics log is a csv file, the field values will be displayed as hexadecimals. The statistics log contains the same data regardless of the format.
Keep the following information in mind when viewing the statistics log:
*The request counts (Request Average Time, New Requests, and Completed Requests) include top-level requests only. A top-level service request is a service invoked by a client request, a trigger, or a scheduled task. If a top-level service invocation results in other services being invoked, the entire service invocation chain is considered to be one request.
*A service being invoked as the circuit open service for a circuit breaker is not treated as a top-level service request.
*A service being invoked as the resource monitoring service for a trigger or as the document resolver service for exactly-once processing by a trigger is considered to be a top-level service.
*By default, Integration Server does not capture data about internal services in the request counts. This is true even when the internal service is invoked as a top-level service. Internal services are those in packages whose names begin with Wm. You can change the services excluded from statistics using the watt.server.stats.packages.exclude server configuration parameter.
*When counting JVM thread usage by Integration Server, the server includes system threads used by Integration Server only. The count does not include any other threads started by the JVM or any processes started by the JVM.
The following table identifies and describes the information written to the statistics log, including the fields for the csv format and the column names used in the hexadecimal format.
CSV Field Name
Hexadecimal Field Name
Description
Time
Time
Timestamp at which Integration Server wrote the log entry to the statistics log.
Total Memory
TotalMem
Total amount of memory available in the JVM that runs Integration Server.
Free Memory
FreeMem
Amount of free memory available in the JVM that runs Integration Server.
Current JVM Threads
CurT
The number of JVM threads started by the Integration Server in this polling interval.
Max JVM Threads
MaxT
Maximum number of JVM threads used by Integration Server from the time Integration Server started.
Total Sessions
SSNs
Total number of active sessions running in parallel on Integration Server at the end of the polling interval.
Max Sessions
SSNx
Maximum number of sessions running in parallel on Integration Server at the end of the polling interval since Integration Server started.
Session Average Time
SSNAvg
Average length of time, measured in milliseconds, for a session that completed during the polling interval. The session average time is calculated by dividing the sum of the length of all sessions that completed during the polling interval by the number of sessions completed during the polling interval.
Running Service Threads
REQs
Number of actively running service threads for top-level service requests in the polling period. Service threads that are spawned by requests from top-level services are not included in this count.
Max Running Service Threads
REQx
Maximum number of actively running service threads since Integration Server started. Service threads that are spawned by requests from top-level services are not included in this count.
Request Average Time
REQAvg
Average time, measured in milliseconds, for a completed request during the polling interval. The request average time is calculated by dividing the total duration of all requests completed during the polling interval by the number of completed requests in the polling interval.
New Requests
StartReq
Number of requests started during this polling interval.
Completed Requests
EndReq
Number of requests completed during this polling interval.