AWS Raises CloudWatch Query Limit to 100,000 Rows, Adds API Pagination

1 articles · Updated · InfoWorld · May 18

AWS lifted Amazon CloudWatch Logs Insights result limits to 100,000 rows from 10,000 and added pagination for the GetQueryResults API across all regions.
The change targets large distributed applications, where SRE teams previously had to split outage queries into smaller time windows, rerun searches, and manually stitch results together.
Pagination may be the bigger architectural gain because it lets runbooks, bots, and SIEM ingestion pipelines retrieve large datasets incrementally instead of handling truncated results with custom logic.
Analysts said the update should speed incident investigations and improve operational efficiency, but not materially cut CloudWatch costs because pricing is tied mainly to data scanned.
CloudWatch may now keep more AWS-native troubleshooting inside Lambda, ECS, and EKS, though Datadog, Splunk, and Elastic still offer broader multi-cloud, security, and governance capabilities.

As AWS native tools improve, is the era of expensive third-party observability platforms like Datadog and Splunk coming to an end?

Could AWS's massive log query upgrade inadvertently lead to higher cloud bills by encouraging larger, less efficient data scans?

With 10x more log data, can AI finally automate incident response and eliminate the need for 3 AM debugging calls?