How Statistics Work
Each application or cache server that joins the cluster can collect and archive statistical data for analyzing system performance.
GemFire statistics can be enabled for a cluster, for an application, for a server, or for a region. Statistics gathered for a cluster, an application, or a cache server are saved to a file and can be archived, whereas region statistics are transient and accessible only through the API.
Set the configuration attributes that control cluster, application, or cache statistics collection in
gfsh or in the
gemfire.properties configuration file. You can also collect your own application defined statistics.
When Java applications and servers join a cluster, they can be configured via the cluster configuration service to enable statistics sampling and whether to archive the statistics that are gathered.
GemFire statistics use the Java
System.nanoTimer for nanosecond timing. This method provides nanosecond precision, but not necessarily nanosecond accuracy. For more information, see the online Java documentation for
System.nanoTimer for the JRE you are using with GemFire.
Statistics sampling provides valuable information for ongoing system tuning and troubleshooting. Sampling statistics (not including time-based statistics) at the default sample rate does not impact overall cluster performance. We recommend enabling statistics sampling in production environments. We do not recommend enabling time-based statistics (configured with the enable-time-statistics property) in production environments.