Computer scientists from the Ethereum Foundation and Stanford confirm three new attacks on Ethereum’s new POS network as part of ETH 2.0. They submitted their new findings in a paper and posted the paper for public view.
Many technologists and cryptographers have concerns about Ethereum‘s switch from Proof of Work (POW) to Proof of Stake (POS).
The three attacks work on the POS consensus mechanism and chain reorganizations. Crypto lingo shortens the term to chain reorg.
What is a Chain Reorg?
Both POS and POW blockchains use consensus. The consensus mechanism ensures that everyone uses the longest chain. It’s part of what makes a blockchain immutable.
But what happens if two miners mine different blocks at the same time? Assuming both blocks are valid, then the miners and nodes have to figure out how to order them. After all, it is a chain. They need to be one after the other.
So how do they decide this? After coming to a consensus that both blocks are valid, they do a chain reorg. They reorganize the chain to put Block A ahead of Block B and accept both blocks.
The name for this is a reorg because it is a reorganization of the chain. By deleting all parts of the chain that have Block B ahead of Block A, Block A is now confirmed to be first in the longest chain. Nodes play an important role here in helping to make this reorg smooth and not noticeable to the chain’s users.
The Attacks on the Ethereum POS Network
The attacks are:
- increase in POS validator profits during a short-term chain reorg. It delays consensus decisions too.
- uses something called ‘adversarial network delay’ to delay consensus decisions indefinitely
- combines both 1 and 2 to allow an adversary with a small stake to exercise much larger influence on the network. This adversary can cause long-range chain reorgs despite their small stake
We are going to try to simplify what these problems are.
For attack #1, increasing the profit of one POS validator to the detriment of others is bad for the network. Delays are bad too.
For attack #2, adversarial networks involve machine learning. With this type of delay, the adversary uses the delay to exploit the network. The exploit, according to page 1 of their paper, uses of strategic voting. They vote to vanish a small number of validators. The problem is with not enough validators voting, it stalls the consensus mechanism. They cannot confirm which is the longest chain. So this stalls the entire network.
For attack 3, they combine the ‘best’ of both previous attacks. The adversary, who holds little to no stake in the network, holds great influence over the network. They can stall it or even create a long-range chain reorg affecting many previously mined blocks.
As the programmers say in the abstract of the paper, “Honest-but-rational or ideologically motivated validators could use this attack to increase their profits or stall the protocol, threatening incentive alignment and security of PoS Ethereum”
And we know everything in crypto works on aligning incentives or no one does anything. That’s the beauty of a free market.
Potential Effects of These Attacks
This paper does two good things for ETH users like us. First, it tells us there is a problem. Secondly, it proposes some possible solutions.
But we don’t know that these solutions will ever get to a vote, let alone implemented to solve the problem.
So here are some questions you need to ask yourself. They might even affect your investment decisions.
- Is holding ETH riskier based on this new information?
- Could information about these attacks kill the bull market that led to an all-time high?
- Does this make using Ethereum-compatible platforms like for DeFi riskier than using other chains?
- Could this delay the rollout of ETH 2.0?
- Will those who committed ETH for POS staking be waiting longer to get started and get paid?
What do you think could happen? Ping us with your comments below
Also, join us on Telegram to receive free trading signals.
Find out more about the blockchain and crypto space on the Altcoin Buzz YouTube channel.