As per the subject, how does the S3 output react when the internet connection is interrupted?
- Are all temporarily files saved until the connection comes back up?
- If a connection fails during transmission, will the output resend the failed file once the connection is back up?
- Different topic but how good is the gzip compression? Is it comparable to the file output .gz compression?
I will be sending logs over slow and unstable connections. Timeouts and interruptions are expected. The current strategy is to have logstash output to a file and upload to S3 from there but that has some downsides as well.