opendnp3 multidrop efficiency #119
Description
Investigate the feasibility of improving opendnp3 multi-drop performance.
At the moment, outstations sharing the same channel block each other when they're waiting for confirms. This results in an underutilised channel. AFAIK, nothing in the DNP3 standard precludes outstations from sending messages while another one is waiting for a response on the same channel. The existing behaviour is likely just a result of how synchronisation is implemented in the stack; specifically, the stack uses the same executor (asio strand) for a channel and it's outstations.
Changing the executor architecture might be very invasive/difficult, so it would need careful assessment whether it's feasible and/or a good idea.