understanding how Cloudfront invalidation batches are created

Question

We have a lambda function which runs everytime any user uploads an image. It processes the image, creates multiple sizes, saves them to a S3 bucket and then invalidates a cloudfront which acts as a caching layer over the S3 bucket.

Since the lambda is run per image, I would be expecting that a CF invalidation batch that gets created would only have the different paths related to that image, but in practice I have observed that the invalidation batch sometimes has path for more than 1 image.

is this something that lambda / Cloudfront handles internally where it somehow clubs requests originating from different lambda runs in the same invalidation batch? is this expected behaviour?

Answer

Hello,

According to the [AWS Documentation](https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/Invalidation.html), to invalidate files, you can specify either the path for individual files or a path that ends with the ` * `wildcard, which might apply to one file or to many, as shown in the following examples:

For one image, you specifically mention the name of the image to invalidate; **Example:**  `/images/image1.jpg`

Double check your invalidation to see if you have not used a wildcard such as  `/images/image* ` or `  /images/*` which is a path for more than one image.

Please note that you can include the ` * ` wildcard at the end of the invalidation path, and using the wildcard invalidates multiple files such as all of the files in a directory or all files that begin with the same characters. 
  
Please refer to this AWS documentation to [Specify the files to invalidate](https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/Invalidation.html#invalidation-specifying-objects).

I hope this helps.

understanding how Cloudfront invalidation batches are created

Relevanter Inhalt