As Beats a Data shippers, and they suppose to ship files, metrics, and so on, from a machine, is it a good approach to run beats with docker, as in this way beats is containerized?
I have currently the issue that I want to ship log files from an application, and if I install filebeats with docker, I have to provide somehow the log to the container. Is it a good approach to do something like this with docker, or should I normally install Filebeats, configure and run it on the machine without a container?
I don't think there is a strict rule, but I would use the same approach for your application as for the Beats — either containerize both or neither. That will also help you keep the expectations and setups aligned: logging to a file vs stdout and how to collect that.