What’s the Bitcoin Mempool for?

The mempool or mempool has been used by everyone who has ever used Bitcoin. What is a mempool, then?

Technically, it is not possible. “the” mempool. Mempool is the cache of all valid bitcoin transactions broadcasted to the Bitcoin network that are still awaiting confirmation in a Bitcoin block. Nodes send each other messages to exchange transactions and see if they are available. 

Each mempool can be viewed as a separate island, complete with unconfirmed transaction, configuration variables, and other settings. The default size is 300MB. A minimum feerate is also available, which can be configured. It adjusts dynamically and automatically. When your mempool is full, and new transactions continue to arrive, this option determines which transactions should be removed. Other options are configurable, including the datacarrier size and datacarrier options for transactions containing OP_RETURN. 

The reasons that different nodes run a mempool are different, as well as their needs. However, it’s only through the interaction of everyone running their mempools in sync and interacting together, can those needs be met. 

Consider each mempool like a real pool. They are all interconnected by ground channels. The pool will be deeper in the earth if a mempool has a larger size. The deepest pools are those used by miners, block-explorers and exchanges. All of them have their own reasons for wanting to be aware of all unconfirmed transactions that are waiting to enter a block. To make sure that they are using the best transactions to create their next block. Exchanges to ensure they know about all transactions. Because their whole service aims to display as many details as possible about the blockchain, mempool, and other data. You only need your average nodes to be large enough to hold the highest feerate portion of the blockchain. “mempool.” 

Imagine each transaction is a liquid drop. The higher the rate, the more dense the liquid drop. The drops are sent through channels that connect the pools. Once they arrive at each pool, the drop is copied and sent to the next pool. When pools overflow, the liquids with lower density (lower fees) spill out first. 

One lucky miner will eventually scoop up a small amount of liquid from the pool’s bottom and pour it into the latest glass tank of a long, winding line of tanks filled with liquid that sit there indefinitely (the blockchain). It is a simple way of thinking about the entire system. 

This interlinking of pools has different uses for different users. 

Transactors 

Mempool can be used in two different ways by the users. The first is for users to send their transactions to miners. It is impossible for transactions to end up in blocks if they are not sent to the miners’ mempool. Interlinking mempools, and sharing transaction data, ensure that, eventually, a single transaction will end up on the mempools for all miners. It is important to have a decentralized and robust network in order to ensure that all transactions are eventually sent from the user to the miner, regardless of the changing or fragmented network connections. 

Fee estimation is the second application. It is important to Layer 2 users, who at any moment may have to make sure that a response transaction for an invalid state has been confirmed on time. The feerates of those transactions can be used to estimate fees, but this doesn’t tell us anything about how the mempool is currently performing after the latest block. This does not account for spikes in transactions, opportunistic players flooding the pool, or even the next round of an ongoing transaction surge. It is impossible to estimate fees without a mempool view. 

Receivers

Your node will verify the transaction and block that contains it when you receive bitcoin. You receive bitcoins via broadcast. The payment is then sent into a miner’s mempool. They find a new block and it is distributed to the entire network. Finally, your node verifies and downloads that block. 

But that is not the way it works in reality (unless you disable a node’s memory pool and only run blocks). Each transaction that is stored in your node’s mempool will be validated by the node and marked as valid. The blockheader is all that a miner relays when they find a newly-found block. The node grabs all the transactions that have been pre-validated, checks the header and, if everything passes, the block is sent forward. “compact block” onwards. 

The miners are no longer dependent on centralized relay networks, such as FIBRE (formerly maintained by Matt Corrallo) and Falcon Network. These were once necessary to ensure low latency for block relays to other miners, due to the slow relay speeds across peer-to-peer networks. 

Mines

It is obvious that miners are interested in seeing everything. The miners are profit-oriented entities who want to have the ability to pick the most lucrative transactions from the large set that is available. It is this way that they earn money and maximize profits to keep expanding and remaining competitive. 

Their incentive to acquire any valid fee-paying transaction is so strong that they have, historically, currently and almost certainly in the future built numerous systems. The miners’ incentive to obtain any fee-paying transaction is so great that, in the past, present, and most likely, the future, they built numerous social media systems to let users submit transactions directly to them, instead of through an open peer-topeer network. 

Block Explorers, Chain Analytics, Etc.

Like miners they want to view every transaction in progress that’s been broadcasted to the entire world. While miners directly make money from the transaction by collecting fees, Blockchain explorers or analytics companies can indirectly monetize transactions in the form of a product. 

Although I cannot provide any specific examples, chain analytics companies are known to buy private metadata pertaining to transaction activity. The companies have been operating sybil Bitcoin Nodes, which peer with as many nodes as possible across the network in order to pinpoint who broadcasted a particular transaction. 

The entire business is based on monetizing visual representations of mempool and blockchain data. If they can display more data, it could be more information that is monetizable if there are novel or useful ways of displaying the information. 

Information wants to Flow

There being a unified system of user classification benefits all of these users. “a” Information flows freely between public mempools. It is valid as long as the fee is sufficient to pass minimum relay filters and it does not pose a risk of legitimate denial-of-service or resource exhaustion to nodes. 

If there is no public mempool that works, then the only alternatives for users are centralized solutions and unmanageable chaotic attempts to create fragmented mempools. Each user would have to track each one individually. 

Private relaying transactions can lead to concerns over Miner Extracted Value, manipulation of data fees, or deceiving of users. Bitcoin is going to have these issues if there’s no healthy public mempool. 

This article is a continuation of the previous one. I will also discuss different mempool filter types and their purpose.