Counter interpretation

Counters collect specific information on a listener for a defined period that is the aggregation time interval. A listener is a specific subsystem or component of an IBM® software product.

There are several types of counters that collect information for the aggregation time interval.

• Event counters

  Event counters record the number of times an operation or activity occurs within the aggregation time interval.

  Events have fields that are associated with them and count real world items, such as the number of database updates. The Dashboard computes and displays the number of events that occurred during each time interval and the corresponding average rate per second that the events occurred.

  • Accumulator

    An accumulator is a counter, which is a defined subordinate to an event. The application adds values to the accumulator that typically represent a measured value that is associated with the transpiring of the event. The listener library automatically computes the total of the squares of the presented values, and the minimum and maximum values that are presented during each aggregation interval. The accumulator accumulates values of actions that happen with an event.

  • Duration accumulator

    A duration accumulator, which measures time, is the most common type of accumulator. This is not Central Processing Unit (CPU) time, it is duration that is measured in nanoseconds. Duration counters record the average elapsed time for a specific event within the aggregation time interval. Because accumulators measure elapsed time for a specific event, they are always subordinate to an instance of an event counter. Therefore, duration counters always have an associated event counter, but an event counter does not necessarily have a duration counter.

    Although this duration is normally captured in nanoseconds, depending on the tool that you use to view the data, it may also be expressed in milliseconds (nanoseconds/1,000,000). For example, you can use the ECM System Manager Dashboard to generate a report to a csv file. If you do not override the default output option, the ECM System Manager Dashboard generates the report in nanoseconds to the csv file. If you override the default output option and select the "Scale Durations"report output option, it writes the duration in milliseconds to the csv file. As another example, when you use the PCH monitor plug-in for Content Cortex containerized deployments, the duration is posted as an integer value in milliseconds.

• Meters

  Meters record the instantaneous value at the end of each aggregation time interval of a value in the server that continuously fluctuates up and down whenever the server is running. The minimum and maximum values during each interval are also recorded.

• Averaging meters

  The averaging meter uses a fixed time window to compute a moving average of values that are recorded by the application. The length of the time window is determined by the application, and it is unrelated to the sampling interval used in the Dashboard.

  There are two types of averaging meters:

  • A simple moving average – in this case, the minimum and maximum values that are presented are the extremes of the values during the time window, which ends at the end of the interval displayed.
  • A rate of progress meter – in this case, the minimum and maximum values that are presented are the smallest and largest values that were recorded since the application was started.

Dashboard exposes counters that are generated by various IBM software products you want to monitor. For each product, related counters and their definitions are listed in the following topics. Use the information for each product to help you interpret the counters that are displayed by the Dashboard.

For information on the IBM Content Navigator Counters, see Monitoring IBM Content Navigator .