I've been using the s3 logs to track downloads for a podcast (mp3 files) I currently have 23 files I actually care about and I only need to know information about GET operations (download requests being fulfilled.)
This works fine as is, but I'm also getting additional log files generated from asking for the files I care about. Is there a way to configure the bucket logging policy such that only the GET requests are recorded? Even better, can I limit based on some type of url request pattern? (e.g. *.mp3)
The default configuration is giving me 10k+ extra files telling me that I asked for the log files and/or that a log file was created (very meta!) Downloading and filtering out all these extra files takes a significant amount of time.