Previous Topic: Control Data CollectionNext Topic: Select Data by Classes


Options for Collecting Data

You have many options for setting the data collection parameters in a collection definition. Some of these options are different for the CPD (primary) collection definition than for the alternate collection definitions. The following table describes the parameters and their default values:

Parameter

Default Value

Schedule

Specifies the weekly schedule for data collection for all nodes in a cluster.

7 days per week, 24 hours

Beginning

Specifies the time when the data collection process is to start collecting data. The primary Performance Agent will not serve data for use by alternate or Real-time Performance Agents until the BEGINNING time has past.

For the CPD, the default value is the date on which you installed the software. For alternate collection definitions, the default is today's date.

Ending

Specifies the date and time on which the data collection process is to stop.

January 1, 2999

Nodes

Specifies the list of the nodes on which data is to be collected

For the CPD definition, the default list includes all the nodes. For alternates, you must specify nodes.

Hot files queue length

Determines whether hot file data is collected for the CPD collection definition. The value is a minimum threshold for the queue length of a disk that triggers the recording of hot file data.

For the CPD, 0.33. Alternates do not collect this data.

Interval

Specifies how often the data collection data is summarized into a set of data records for reporting.

For the CPD, fixed at 120 seconds. You can specify this for alternate collection definitions.

Delete after

Specifies the retention period for how long files are retained in the database before being automatically deleted.

7 days for the CPD 99,999 days for alternates

Classes

The Classes parameter specifies which daily data is recorded. You can choose from process, I/O, and system metrics. See Selecting Data by Classes in this chapter.

All classes

Include or Exclude Lists

Specify inclusion and exclusion lists to include or exclude data on specific processes, PIDs, devices, or volumes.

Created by user

Data base directory path

Specifies the path to the location for data files for an individual node or entire collection definition.

PSDC$DATABASE Can be specified by individual node

Working set

Specifies the working set quota (WSQUOTA)for the process that executes the Performance Agent. The number must be greater than 1024.

2048 Can be specified for a collection definition or by individual node

Free space

Specifies a threshold value that is the minimum number of free blocks needed on the database disk for recording data. When the number of free blocks drops below this minimum, a Performance Agent will hibernate (stop recording data). Once the free space becomes available, data recording will resume.

2500 Can be specified by individual node