Amazon CloudFront now offers two capabilities to enhance origin timeout controls: a response completion timeout and support for custom response timeout values for Amazon S3 origins. These enhancements provide more granular control over origin response timeouts, allowing you to deliver consistent and reliable user experiences regardless of variation in network conditions or origin performance.
Previously, you could configure a response timeout to control the amount of time CloudFront waits for your origin to send the first packet, as well as the amount of time CloudFront waits for subsequent packets. If your origin times out, CloudFront resets the response timeout and tries again based on the configured number of retries. With the new response completion timeout, you can now additionally configure the maximum amount of time CloudFront should wait for a complete response from your origin across all packets and retries. This allows you to control the cumulative response time for latency sensitive workloads such as media streaming or API calls. When using Amazon S3 as your origin, you can now also set custom response timeout values instead of using the default value of 30 seconds. These capabilities provide you with more control over how CloudFront handles slow or unresponsive origins.
CloudFront supports response completion timeout, and custom response timeout values for Amazon S3 origins, across all CloudFront edge locations excluding the AWS China (Beijing) region. You can configure origin timeouts using the CloudFront console, API, and AWS CloudFormation at no additional charge. To learn more, visit the CloudFront Developer Guide.
Categories:
Source: Amazon Web Services
