You can monitor current system performance using metrics. You can assess the status of the system as a whole, as well as the status of individual servers, dispatchers, and enabled services.
For example, you check the performance metrics and notice that the report service shows a red square indicator, which identifies poor performance. You view the metrics for the report service and determine that the number of requests waiting in the queue exceeds the number that can be processed in a specified length of time. You can then take action to reduce the number of requests waiting in the queue.
Metrics are grouped using the following categories:
session metrics for monitoring the number of sessions in your system
These metrics are collected by Content Manager.
queue metrics for monitoring the ability of dispatchers and services to keep up with requests that are waiting in queues
For example, queue metrics for services may identify that requests are waiting too long in a queue for processing. There are two types of queue metrics. Global queue metrics, which include the JobQueue, SMTP Queue, and Task Queue, are collected by the monitor service. Service queue metrics are collected by individual services.
JVM metrics for monitoring status information such as the length of time a JVM in the IBM Cognos 8 environment has been running and how much memory it is using
These metrics are collected by the JVM.
service request metrics for monitoring processing times, volume of requests, operational status of services, and response times
These metrics are collected by the dispatcher that manages the service.
report service metrics for monitoring report service processes
These metrics are collected by the dispatcher that manages the report service.
The availability of a metric depends on the resource with which the metric is associated. For example, some metrics are available for the system, servers, dispatchers, and services. Other metrics may be available for only a subset of these resources.
You define the threshold values that determine whether the status for a resource indicates excellent performance (a green indicator), average performance (a yellow indicator), or poor performance (a red indicator). There are no default threshold values. If you define thresholds, the values are stored in the content store.
You can also create an agent that monitors the metrics and notifies you when threshold values are exceeded. For example, you can create an agent that sends you an email when a threshold value is exceeded.
When a threshold is exceeded, the dispatcher writes an entry to the logging database.
For information about system metrics, including how to define threshold values and notifications, see the IBM Cognos 8 Administration and Security Guide.