Reduce first waiting time for fuji requests

For now we have a standard waiting time of 5 minutes after a failed request for fuji calls (and trying at max 10 times). Fuji has a request limit of 100 requests per minute.

For some servers with a lot of datasets per publication we are running into that limit and getting a connection timeout. That means, we are waiting 5 minutes where 1 (or 2 to be sure) should be enough.

e.g. for the current KIT/MDC run we have this issue every 5 minutes, which actually means almost for each minute we run the script we have to wait 5 minutes… so this run is actually 3 times longer that it needed to be.