Specify how long to retain the data that is collected about
resources and the log files that are generated by IBM Spectrum®
Control. By
specifying the number of weeks for history retention, you can control
the amount of data that is retained and available for historical analysis
and charting. The longer that you retain data, the more informative
your analysis, but the more storage space that is required to store
that data.
Data that IBM Spectrum
Control collects about a storage environment is stored in a DB2® database repository. The amount of data that is retained about resources can grow over time, and thus require more storage space for the repository. You can use the History Retention page to modify
the data retention settings according to the monitoring and storage
requirements of your environment. You must be assigned the Administrator
role to modify data retention settings.
- In the menu bar, go to .
- Click Edit to modify the following
data retention settings:
- Capacity history
- Specify how long to retain a history of the capacity data that is collected about monitored
resources. This value determines the amount of capacity data that is retained and available for
historical analysis and charting. The longer that you retain data, the more informative your
analysis, but the more storage space that is required.
- Daily
- Specify how long to retain capacity data that is collected daily about resources. You can retain
daily data for up to 72 weeks and a minimum of 2 weeks.
- Weekly
- Specify how long to retain capacity data that is aggregated weekly for monitored resources. You
can retain weekly aggregates for up to 96 weeks and a minimum of 4 weeks.
- Monthly
- Specify how long to retain capacity data that is aggregated monthly for monitored resources. You
can retain monthly aggregates for up to 48 months and a minimum of 2 months.
- Performance data
- Specify how long to retain data that is collected by performance
monitors.
- Sample
- Specify how long to retain sample data that is collected by performance
monitors. Sample data represents the data that is collected each time
a performance monitor is run. Because sample data is collected frequently,
retaining that data can require significant disk space in the database
repository. The required disk space is determined by the types of
switches, storage systems, and number of volumes that are being monitored.
You can retain sample data for up to 12 weeks.
- Hourly
- Specify how long to retain hourly data that is collected by performance
monitors. You can retain hourly data for up to 24 weeks.
- Daily
- Specify how long to retain daily data that is collected by performance
monitors. You can retain daily data for up to 156 weeks.
Consolidating performance data: Performance data is collected at intervals. An interval
represents the number of minutes over which samples of performance
data are averaged. When performance history is retained, data that
is collected at certain intervals is automatically consolidated, or
rolled up, to higher intervals. For example, data collected at 1-minute
intervals is consolidated into 5-minute data; data collected at 5-minute
intervals is consolidated into 1-hour data; and so on.
Tip: If performance data is collected at 1-minute
intervals, the amount of data that is stored in the database repository
increases significantly. The product stores only 7 days of sample
data that is collected at 1-minute intervals.
- Data for missing resources
- Specify how long to retain data about internal resources that
are no longer detected by IBM Spectrum
Control. You can
retain the data of removed resources for up to 52 weeks.
- If the internal resource of a top-level resource is not detected
when that top-level resource is probed, data about the resource is
removed when the time limit is reached. The internal resource is removed
only from the top-level resource that is probed. For example, if two
weeks are specified, the data for a pool that is missing from a storage
system for more than two weeks will be removed.
- Only internal resources are automatically removed according to
this setting. Storage systems, servers, hypervisors, switches, and
fabrics must be removed manually.
- Alert logs
- Specify how long to retain alerts. An entry is generated each
time that an alert condition is detected on a resource. Any alert
that is older than this value is deleted. You can retain alerts for
up to 12 weeks.
- Job logs
- Specify the maximum number of logs that are retained for data
collection jobs. A log file is created each time that a job is run.
When this number is reached, the entry for the oldest log is deleted.
For example, if you accept the default value 5, and then run a probe
6 times, the log file for the first run is deleted. You can retain
up to 20 logs for a job.
- Click Save to apply the retention
settings.
- Click Restore Defaults to
restore the retention settings to their default values.