17.1 CloudWatch Monitoring
HeatWave on AWS supports AWS CloudWatch for detailed monitoring of your DB Systems and HeatWave Clusters.
These are the prerequisites for using the feature:
- An AWS account for receiving the metrics emitted by the monitoring service.
- Ability to create an IAM role in your AWS account.
Follow these steps to set up or reconfigure CloudWatch monitoring for your DB System (requires logging in to your AWS account):
- Create an AWS IAM role and policy in your AWS account for CloudWatch Monitoring. See instructions here.
- Enable CloudWatch monitoring for your DB System by one of the
following ways:
- Setting the Monitoring configuration when Creating a DB System.
- Enabling CloudWatch monitoring for an existing DB System by going to the DB System Details page and select under the Actions menu Update Monitoring Configuration. Follow then Step 3 in Update Monitoring Configuration.
Verify the list of CloudWatch monitoring metrics you have set up by checking the Monitoring Configuration tab of your DB System (see MySQL DB System Details).
To view the metrics in your AWS CloudWatch account:
- Go to the Metrics tabs of your DB System (see Viewing DB System Details) and HeatWave Cluster (see Viewing HeatWave Cluster Details). Under CloudWatch Metrics, follow the link there to your AWS
CloudWatch console.
Note:
If CloudWatch monitoring has not been enabled for the DB System, you can enable it by clicking Update Monitoring Configuration under CloudWatch Metrics, and then follow Step 3 in Update Monitoring Configuration. - In the AWS CloudWatch console, choose on the left sidebar Metrics > All metrics. On the Browse tab and under Custom namespaces, choose the OracleHeatWave namespace and select the ID of the DB System of interest.
You can also create alarms and dashboards in your AWS account for your metrics. See Related Topics below.
Note:
For high availability DB Systems, metrics are only emitted for the primary instance. On a failover, the new primary is configured to start sending metrics, and the old primary is configured to stop sending metrics. During this time of reconfiguration, the metrics emitted may be inconsistent.Note:
During maintenance, the metric emission might be interrupted.