Finalize Native Threading Model as it affects DMCHandler()

[fuxingloh]

What would you like to be added:

Finalize Native DeFiCh/ain threading model when implementing class DMCHandler for outgoing RPCs from the native side. This might be a non-code change, but it's open for discussion with no resolution in mind.

Why is this needed:

While on the native side, the minter runs on single-threaded at the moment. The current model limits all node operators to run on a single thread. While this is perfectly fine and DMC can keep the same threading model. Coupling model with DMC side where block creation can be more expensive it might create less than ideal performance for DeFiChain network.

That being said, DMC doesn't have any compute-intensive module (e.g., no staking model). It might be trivial and a non-issue. We could also optimize DMC to always "prepare" a block and finalize it on Native side without having to compute at "mining (staking)" time.

[fuxingloh]

The outcome is to assess and understand what we should do there. Should we change anything? Does it matter? Does the threading model on DNC affects DMC?

[mambisi]

Thanks @fuxingloh I will look into it a give feedback