Different running time using same partners

Dear Prof. Bonvin,

I am using the haddock webserver. The past runs lasted 2 hours, while the last runs using the same molecule partners are running since more than 12 hours. Is there a way to know whether there has been an error or it is normal that computation time vary so much?

Thanks in advance

It seems we have a computational infrastructure problem which is delaying the runs.
Be patient. The responsible units have been notified.