Description
Container Insights collects and analyzes more detailed CloudWatch information about ECS cluster and service CPU, memory and other operational metrics. Insufficient observations can delay incident response and performance analysis.
If the cluster setting is omitted, an account default can apply. Verify the effective state and distinguish Container Insights from basic ECS metrics or other monitoring that may still be available.
Potential impact
- Resource use and performance problems can be harder to identify promptly.
- Diagnosis of task failures and recovery can be delayed.
Remediation
Specify Name: containerInsights and Value: enabled in ClusterSettings. Consider supported enhanced mode if more detail is needed. Review costs and necessary permissions, then verify CloudWatch dashboards, metrics and alarms in operation.
Examples
Before
Resources:
ECSCluster:
Type: AWS::ECS::Cluster
Properties:
ClusterName: MyCluster
There is no cluster-specific setting. Check the account default and actual cluster state.
After
Resources:
ECSCluster:
Type: AWS::ECS::Cluster
Properties:
ClusterName: MyCluster
ClusterSettings:
- Name: containerInsights
Value: enabled
This explicitly enables Container Insights. Verify actual metric collection and the alarms needed for operational response.