Rename Filebeat s3 input to aws-s3 input

[kaiyan-sheng]

This issue is to track the work for renaming s3 input to aws-s3 to match the naming convention for other inputs in Filebeat. I couldn't make it backwards compatible so this issue is to track the work for 8.0 as a breaking change.

[elasticmachine]

Pinging @elastic/integrations-platforms (Team:Platforms)

[P1llus]

Just wanted to add that it might be useful to consider the fact that we want to/is looking at the possibility to add a S3 input without SQS, similar to Logstash.

If we are renaming the S3 input already, and don't want to potentially rename it twice, maybe we could name it aws-s3-sqs ?

That way, if we decide to add S3 input without SQS later down the line, the name does not need to change again.

[masci]

Let's say we had elastic/beats#18205 implemented, this to me would look like an input that can be configured either polling (current missing feature) or event-driven: in this regard, SQS become an implementation detail supporting one of the two operational modes, not sure we should add it to the name of the input. Staying generic seems more future proof.

[P1llus]

I don't mind @masci approach either, I just remembered that is how it was named in the Logstash implementation, and earlier discussion of this feature was mentioned as being implemented as a separate input, if we want to combine them into one, then indeed there is no need to add anything to the name itself!

[exekias]

++ on @masci comments. Also @kaiyan-sheng, having that this input is in beta, I don't think renaming would be a breaking change. Perhaps we can rename + GA at the same time in the next release?

[kaiyan-sheng]

++ using aws-s3 input name but include both with/without SQS setup.

@exekias Sounds good to me to rename + GA it in 7.12. We just need to make sure all the filesets that are using this input get adjusted as well and in aws integration.

[masci]

GA release tracking issue #489