DB2 10.5 for Linux, UNIX, and WindowsHealth indicators
The health monitor uses health indicators
to evaluate the health of specific aspects of database manager performance
or database performance. A health indicator measures the health of
some aspect of a particular class of database objects, such as table
spaces.
Criteria are applied to the measurement
to determine healthiness. The criteria applied depends on the type
of health indicator. A determination of unhealthiness is based on
the criteria generates an alert.
Important: The health monitor,
health indicators, and related components have been deprecated in
Version 9.7 and might be removed in a future release. Health monitor
is not supported in DB2® pureScale® environments. For more information, see Health monitor has been deprecated.
Three
types of health indicators are returned by the health monitor:
- Threshold-based indicators are measurements that represent a statistic (on a continuous range of values) of the behavior of the object. Warning and alarm threshold values define boundaries or zones for normal, warning, and alarm ranges. Threshold-based health indicators have three valid states: Normal, Warning, or Alarm.
- State-based indicators are measurements that represent a finite set of two or more distinct states of an object that defines whether the database object or resource is operating normally. One of the states is normal and all others are considered non-normal. State-based health indicators have two valid states: Normal, Attention.
- Collection state-based indicators are database-level measurements that represent an aggregate state or one or more objects within the database. Data is captured for each object in the collection and the highest severity of conditions among those objects is represented in the aggregated state. If one or more objects in the collection are in a state requiring an alert, the health indicator shows Attention state. Collection state-based health indicators have two valid states: Normal, Attention.
Health indicators exist at the instance, database, table space, and table space container level.
You can access health monitor information through the CLP, or APIs. You can configure health indicators through these same tools.
An alert is generated in response to either a change from
a normal to a non-normal state or a change in the health indicator
value to a warning or alarm zone that is based on defined threshold
boundaries. There are three types of alerts: attention, warning,
and alarm.
- For health indicators measuring distinct states, an attention alert is issued if a non-normal state is registered.
- For health indicators measuring a continuous range of values, threshold values define boundaries or zones for normal, warning and alarm states. For example, if the value enters the threshold range of values that defines an alarm zone, an alarm alert is issued to indicate that the problem needs immediate attention.
The health monitor will only send notification and run an action on the first occurrence of a particular alert condition for a given health indicator. If the health indicator stays in a particular alert condition, no further notification will be sent and no further actions will be run. If the health indicator changes alert conditions, or goes back to normal state and re-enters the alert condition, notification will be sent anew and actions will be run.
The following table shows an example of a health indicator
at different refresh intervals and the health monitor response to
the health indicator state. This example uses the default warning
of 80% and alarm thresholds of 90%.
| Refresh interval | Value of ts.ts_util (Table space utilization) health indicator | State of ts.ts_util health indicator | Health monitor response |
|---|---|---|---|
| 1 | 80 | warning | notification of warning is sent, actions for a warning alert condition are run |
| 2 | 81 | warning | no notification is sent, no actions are run |
| 3 | 75 | normal | no notification is sent, no actions are run |
| 4 | 85 | warning | notification of warning is sent, actions for a warning alert condition are run |
| 5 | 90 | alarm | notification of alarm is sent, actions for an alarm condition are run |