The evaluation of the file collector resulted in the following:
The file-collector currently cannot be scaled across multiple instances. This depends on the implementation in memory whether a file has already been fetched. The query as to whether a file has already been set on the data router is not synchronized. So it can happen that on several instances at the same time a file is downloaded several times and is set in the data router. See also the evaluation in the attachment.
The implementation must be adjusted so that the data-file-collector is scalable. The download must be able to be synchronized across multiple instances.
In addition, the following circumstance would have to be solved: If files cannot be downloaded (e.g. no longer exist), then the data-file-collector gets stuck when its 200 threads are occupied with it. Then it only continues after an expired...