To help S3 account holders monitor the API operations that are performed on their objects, the service provides a mechanism called Server Access Logging to generate log files detailing the requests made against their buckets. This logging mechanism is intended to provide similar information to the access log files produced by standard web servers. These logs can help you to monitor the usage of your S3 resources, particularly when these resources are publicly available. By reviewing the log files, you can judge the popularity of particular resources and the cost incurred by serving them. There is no cost for using the Server Access Logging feature, beyond the standard fees for storing the generated log files.
The Server Access Logging mechanism is a beta feature of the S3 service. Neither the timely delivery nor the accuracy of the log documents is guaranteed. The logging operates on a best-effort basis and is not intended to provide a reliable record of activity suitable for billing or auditing purposes.
Server Access Logging must be explicitly enabled for each bucket you wish to monitor; it is not active by default. When you enable logging for a bucket, S3 will track the requests performed against that bucket, or against the objects inside it, and will write the details to log files. These log files are periodically saved as objects in a bucket of your choice.
It is important to note that the log files are not updated dynamically and are only written ...