Issue #131 - Node Sync Optimization

[ezeike]

Description

This PR updates the node syncing functionality in controller/consensus.go to optimize synchronization and handle unexpected blocks.

Changes Made

• Implements a queue to track both block requests sent and block request responses awaiting processing by the FSM.
• A new type blockSyncRequest was created to track these requests in the queue.
• Block requests are distributed randomly to all known peers in the PeerSet.
• Uses the SimpleLimiter to limit requests per node to avoid rate limiting and reputation slashing.
• A rateScaleFactor setting to allow node runners to increase or decrease their block request rates
• Gracefully handles (ignores and prints warning) when the latest gossiped block is received.

Closes

Issue: #131

[rem1niscence]

this behavior is not strictly related to the controller itself and I see you perform more operations over the queue, why don't convert the queue map into its own type like type queue map[uint64]blockSyncRequest{} and more this method and queuing/dequeueing to methods of that type, that way you can encapsulate all the behavior of this queue there

[ezeike]

We could do this and generally I prefer that approach but in this case there's very little to the queue.

One spot to add, two spots to remove, and it's not going to be used from anywhere else.

[rem1niscence]

@andrewnguyen22 thoughts on this? I honestly ultimately don't mind keeping it as-is but might be a little cleaner to have a custom type