CHACTL Command Reference

The Oracle Cluster Health Advisor commands enable the Oracle Grid Infrastructure user to administer basic monitoring functionality on the targets.

chactl monitor

Starts Oracle Cluster Health Advisor monitoring for all instances of a specified RAC database or for the cluster, using the current, specified, or default Oracle-supplied model. The command stores monitoring status and automatically includes newly detected instances; use -force with -model to replace the active model without first unmonitoring the target. Common errors indicate that the CHA service is unavailable, the database is unconfigured, input is invalid, or the target is already monitored.

chactl unmonitor

Stops Oracle Cluster Health Advisor monitoring for all instances of a specified database. The command requires the database unique name and reports when the database is not currently monitored.

chactl status

Displays Oracle Cluster Health Advisor monitoring status for all targets or for a specified cluster or database. Targets with positive status are shown by default, while explicitly specified targets may also show Not Monitoring; the -verbose option includes contained nodes or instances and their executing models, and the command can report when the CHA service is offline.

chactl config

Lists monitored targets and the model assigned to each target. For a specified database or cluster, it reports whether monitoring is enabled and identifies the configured model; multitenant container databases and clusters also expose configuration data status.

chactl calibrate

Creates and stores a target-specific Oracle Cluster Health Advisor model using selected calibration periods and optional KPI ranges, improving sensitivity and diagnostic accuracy for the operating environment. Database KPIs include CPU, disk reads and writes, throughput, and database time per call; cluster KPIs exclude database time per call. The -force option overwrites an existing model, while invalid time ranges, duplicate model names, or insufficient calibration data produce errors.

chactl query diagnosis

Reports abnormal problems, causes, incidents, and recommended corrective actions for monitored cluster nodes and databases, either for the current state or within a specified time range. Results are time-compressed and grouped by component; normal targets are omitted, recurring incidents are consolidated, and options can limit output to a database, cluster, interval, or HTML file. Messages distinguish normal operation, missing data, and targets that were inactive or unmonitored.

chactl query model

Lists available Oracle Cluster Health Advisor models or displays metadata for a named model. Detailed output can include target type, version, operating system, calibration target and date, calibration ranges and KPIs, validation status, source data, and monitored targets.

chactl query repository

Displays Oracle Cluster Health Advisor repository capacity and usage information, including maximum and available retention time, entity counts, and allocated and total repository size. The command applies only when GIMR is configured.

chactl query calibration

Displays calibration statistics for a specified database or cluster, optionally filtered by time ranges, KPI thresholds, and sampling interval. Output includes sample counts, filtered-data percentages, and descriptive statistics and distributions for supported CPU, I/O, throughput, and database-time metrics; without a time range, all collected calibration data for the target is shown.

chactl remove model

Deletes a named Oracle Cluster Health Advisor model and its associated calibration data and metadata from the repository. A model cannot be removed while it is assigned to monitored targets, and the command reports an error if the specified model does not exist.

chactl rename model

Renames an Oracle Cluster Health Advisor model while preserving all links associated with it. The source name must exist and the destination name must be unique in the repository.

chactl export model

Exports a specified Oracle Cluster Health Advisor model to a file for external storage or transfer. The command requires the model name and output file path and is available only when GIMR is configured.

chactl import model

Imports an Oracle Cluster Health Advisor model from a file into the repository. If a model with the same name already exists, the -force option is required to overwrite it; the command applies only when GIMR is configured.

chactl set maxretention

Sets the maximum retention period, in hours, for Oracle Cluster Health Advisor diagnostic data. The default and minimum retention is 72 hours, although storage constraints can reduce retention for all targets; monitoring stops if retention falls below 24 hours.

chactl resize repository

Resizes the Oracle Cluster Health Advisor repository tablespace according to the current retention period and requested number of hosts and database instances. The operation can fail because of insufficient disk space or data exceeding the requested size, and it is available only when GIMR is configured.
Read More: chactl resize repository