Today, I delved a bit deeper into how mempools works and discovered a lot of new information.

It turns out there is no single global state of the mempool. Each node initializes its own mempool.

This happens approximately as follows:

1. The node initializes the mempool (via the New() method) and has its own RPC server that listens for incoming transactions from wallets and dApps.

2. The node calls the Add() method when transactions arrive and immediately distributes these transactions across the subpools of the main mempool.

3. When creating the mempool in the New() method, the node launches loop() which listens for blockchain events and filters the local mempool when needed, possibly discarding some transactions. This is another way of synchronization. 

4. The Pending() method of the mempool returns transactions from all subpools in the pending status, which I don't really like. As an MEV searcher, I would like to see transactions in all states, unloading all subpools.

layerzero

Another important update from Bryan

The CEO of the LayerZero project continues to publish his thoughts on the distribution of drops and, in part, collect feedback from the community.

I decided to make a short summary of the main points:

Entrepreneur with a crypto focus, building my own agency

andre

@andrecryptod