usually you may catch a general exception and throw another one that's caught up the call stack. In this case i think it may be useful for logging additional causes that are not going to be obvious with just the stacktrace(?)