Announcing increased quotas and improvements to CloudWatch Logs Anomaly Detection and Pattern Analysis

Amazon CloudWatch is excited to announce improvements to its log pattern analysis and anomaly detection features. First, CloudWatch Logs Insights pattern and diff commands now use named tokens to make the results easier to read. Second, the default quota for Log Anomaly detectors has been increased from 10 to 500 per account.

CloudWatch Logs Insights customers use Machine Lowered (ML) powered commands to aggregate logs into patterns, enabling thousands of logs to be condensed to few lines for analysis. Now, the pattern and diff commands make it eaiser to analyze your log data by parsing and naming fields according to the data type. For example, a field containing an ARN values will now be named ARN-1, a field containing an IP address will be named IPV4-1, etc. Using the named patterns, customers can easily identify and inspect common fields that occur in their logs such as Request IDs, HTTP response codes, and more.

This feature is now supported in all AWS Regions where CloudWatch Logs Anomaly Detection is available.

To get started with the pattern and diff commands, visit the CloudWatch Logs Insights Query Syntax Guide or select “Query help” from within the CloudWatch Log Insights console page. See the documentation to get started with CloudWatch Logs Anomaly Detection.