Overview

The PrimeAlert Reporter module maintains a database of historical values for monitored Sun MC objects. When combined with the PrimeAlert WebPortal interface, users can access data, reports, and process lists from any host with a Solaris Sun MC or Windows PrimeAlert Agent - all this from any computer with a web browser. A command line utility has also been provided to request report graphs or data from PrimeAlert Reporter.

Note that PrimeAlert WebPortal version 1.2 or above must be installed in order to use PrimeAlert Reporter. WebPortal need not be installed on the agent host, but it must be installed on a host which can communicate with the agent host. See our web site for more information about PrimeAlert WebPortal.

PrimeAlert Reporter allows you to:

• Store the value of any Sun MC object at regular intervals to a database.
• View the stored data as a graph or list.
• Summarize the stored data in meaningful reports, which can be created using generic or custom templates.
• Export the stored data into standard file formats that are compatible with spreadsheet and database applications; and,
• Automatically capture snapshots of the Solaris system process list every 5 minutes; these snapshots are stored for up to five years.

Figure 1 - A Report Generated By PrimeAlert Reporter

Bins and Data Cascade Types

Reporter stores historical data for all loaded objects into a database. This data is kept in bins of varying time durations, and time resolutions. Bins of shorter duration have a finer time resolution and bins of longer duration have a coarser resolution. Data is cascaded from bins of shorter duration and finer resolution, to bins of longer duration and coarser resolution.

This data is cascaded using 4 different algorithms to provide the maximum amount of information for data far into the past. These algorithms are the Average, Minimum, Maximum and Last data cascade types, described below:

The data bin with the finest resolution is the Last 24 Hours bin. This stores the value every 2 minutes. The Last Week bin has a resolution of 20 minutes. Every 20 minutes, data is cascaded from the Last 24 Hours bin into the Last Week bin.

Using the Average algorithm, the last 10 values stored in the Last 24 Hours bin are averaged, and the averaged value is stored in the Last Week bin.

Using the Minimum algorithm, the last 10 values stored in the Last 24 Hours bin are inspected, and the minimum value of those 10 is stored in the Last Week bin.

Using the Maximum algorithm, the last 10 values stored in the Last 24 Hours bin are inspected, and the maximum value of those 10 is stored in the Last Week bin.

Using the Last algorithm, the last value stored in the Last 24 Hours bin is stored in the Last Week bin.

This process is repeated when data is cascaded from the Last Week bin into the Last Month bin, etc.

Cascaded data for each of these 4 data cascade types can be retrieved. When viewing data in bins of long duration, the Average data cascade type will show smooth data, as it has been averaged over many individual measurements. The Last data cascade type will show noisier data that represents a single measurement. The Maximum and Minimum data cascade types will show the largest and smallest measured values that were measured at 2 minute intervals, regardless of the time resolution of the bin. For example, when viewing the bin for the Last 5 Years, the time resolution is 24 hours, so the minimum or maximum measurement made in that 24 hour period will be shown.