Cloudfront high % misses

Question

A customer mentioned that they are seeing increased latency from Europe as compared to US.
They have:
CloudFront -> **S3 origin (in us-west-1)**.
Requests are over **HTTPS** using **signed URL**.
Using Default Cache Behavior

Looking at their Cache Statistics Reports, we noticed that Europe had 52% Cache Miss in the last 2 weeks.

1) I am trying to see if there is a way to validate that the high % misses are valid. i.e, requests were made for different objects and so didnt get cache benefit.

2) The objects are modified only once a week, hence the customer wanted to check if increasing the TTL will help. Since the requests are signed URL I don't think it will help, but wanted to check here.

Thanks!

Accepted Answer

Customer should enable access log on their CF distribution: https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/AccessLogs.html

Analyzing the log through AWS Athena or others analytic tools will help customer to know which object or edge location contributed the highest miss rate. If origin is S3, internally we can use snf https://s3prod-finder.amazon.com/snmFinder/snmFinder to tell the object lifecycle. And edge tool to get the insight of each CF request: https://edge-tools.amazon.com/

Cloudfront high % misses

Contenuto pertinente