Amazon CloudWatch Cost Optimization
CloudWatch costs are driven primarily by log ingestion and retention. Teams ship everything to CloudWatch Logs by default, retain logs forever, and create custom metrics and dashboards without considering cost. Log ingestion alone can cost $0.50/GB, and high-volume applications can push tens of GB per day.
Common CloudWatch cost waste patterns
Excessive log retention
CloudWatch Log Groups default to "Never expire" retention. Months and years of log data accumulate, charging $0.03/GB per month for storage.
High-volume log ingestion
Debug-level logging in production, verbose Lambda output, and chatty applications can push enormous volumes of logs at $0.50/GB ingestion.
Unused custom metrics
Custom metrics that were created for debugging or experiments and are no longer queried but still being published at $0.30/metric/month.
Detailed monitoring on all instances
EC2 detailed monitoring (1-minute intervals) enabled on instances that only need basic monitoring (5-minute intervals).
How to reduce your CloudWatch costs
Set log retention policies
Save 30-60% easyConfigure retention periods on every Log Group. 30 days for dev, 90 days for staging, and 1 year for production is a common starting point. Never leave retention at "Never expire."
Reduce log verbosity in production
Save 20-40% mediumSwitch from DEBUG to INFO or WARN level in production. Filter out repetitive health check logs. Every GB you do not ingest saves $0.50.
Export old logs to S3
Save 40-60% mediumFor logs you need to retain long-term for compliance, export them to S3 and delete from CloudWatch. S3 storage is roughly 10x cheaper than CloudWatch log storage.
Audit custom metrics
Save 5-15% easyList all custom metrics and check which are still being queried in dashboards or alarms. Stop publishing metrics that nobody uses.
How ZapStack finds CloudWatch waste
ZapStack automatically scans your AWS environment and surfaces CloudWatch cost savings you can action immediately.
- Identifies Log Groups with no retention policy set
- Flags high-volume log ingestion per Log Group
- Detects unused custom metrics
- Estimates savings from retention policy changes
- Recommends S3 export for long-term log retention
Frequently asked questions about CloudWatch costs
Related AWS services
Lambda Cost Optimization
Serverless compute service that runs code in response to events without managing servers.
EC2 Cost Optimization
Elastic Compute Cloud provides resizable virtual servers for running applications in the cloud.
ECS Cost Optimization
Elastic Container Service runs and manages Docker containers on AWS using EC2 or Fargate.