Multiple "endpoints" can be configured. Each endpoint handles incoming requests for a specific URL, queues them, and then sends them to another configured URL. An endpoint has a "worker" which is responsible for reading the files from the queue and processing them. Workers can process multiple files in parallel as configured (by default 2 at a time). Workers can be paused/unpaused or repopulated via a RESTlike endpoint. Pausing a worker will stop it from processing files from the queue, but the endpoint will continue accepting requests and writing the files to the queue. Repopulating a worker will cause it to refresh its queue from the files on the filesystem. This is useful when manually adding files to or removing files from the queue.
0 commit comments