Role of the Shared Cache in Metrics and Event Notification
As part of the metrics and event notification process, the cluster uses the shared cache to:
Register the policy actions configured in, and received as part of, the virtual service definition deployed from
CentraSite. The shared cache provides a cluster-wide view of cached policy actions for the cluster (such as, the Log Invocation action and the Monitoring actions).
Store aggregated metrics reported by the nodes in the cluster. Once all of the metrics for a particular service are reported to the shared cache, the data can be consolidated to minimum, maximum, and average response times and success and failure rates. This data is stored in the shared cache until it is time to report the data to
CentraSite.
Provide fault tolerance for the cluster. Normally, any content stored in the memory of the node as a queued task is lost if the node goes down. However, if the run-time events or metrics are written to the shared cache, they can be recovered.