0

Is this true that miners alone cannot do 51 attack even one or a group of miners posses >51% of hashrate of the network?

As far as I know, every block should be accepted by nodes to be added to the blockchain. In other words, nodes are responsible for maintaining the whole blockchain, so they can choose which blockchain is the best one.

If a group of miners decide to perform an attack such as double spending a coin, nodes can block those miners and so their blocks would not be a part of the blockchain. So any party who owns more than 51% of hashrate can not do any malicious thing, unless it controls 51% of both nodes and miners. Is this claim correct?

Improve this question
1
  • I think nodes without economic relevance (or with redundant economic relevance) don't really cause 51% attacks, but eclipse attacks.
    – Mercedes
    Feb 26 at 13:01

2 Answers 2

Reset to default
2

51% attack

My understanding is that the 51% attack only requires more than 50% of the total hashpower. It doesn't require control over a majority of nodes.

Nodes each choose the chain with most work as their active chain. If one miner can produce more work than the rest of the miners combined, they can produce a chain with more work than any alternative chain.

Attack detection and reaction

So far as I know, the broad ecosystem of wallet software is not able to explicitly detect an attack of this sort and automatically apply countermeasures consistently.

the human operator of the software may be able to apply some measures such as block invalidation, but I believe every person operating a wallet or other node would have to do this manually. It is likely that people who choose a wallet for ease of use do not actively monitor bitcoin news and would not react in a timely manner. Since there is no established system of coordination between ordinary users, the network would presumably fragment. At least initially. It seems likely this would develop into a major issue with significant long term technical, community and reputational consequences.

The countermeasure to the 51% attack is not a reactive one but more of a preemptive one - it is considered very difficult for one person or organisation to suddenly amass 51% of the network hashpower.

I guess if the more vigilant members of the Bitcoin community detect some gradual centralisation of hashpower rising towards 50%, the response is possibly to engage more people in diversifying hashpower and perhaps develop new measures that aren't currently automatic.

Improve this answer
10
  • 1
    @AmirrezaRiahi Yes, yes they can. But do they want to? The goal of the protocol was to introduce an economic incentive to avoid misbehaving parties. If that incentive doesn't work (exemplified by the occurrence of an attack), it means the assumptions underlying the system were wrong, and maybe the conclusion is that PoW in general just doesn't work. You can't patch it up by replacing it with the consensus of users overriding the hashrate - if that was ok in the first, why were you using PoW in the first place, as its goal is specifically avoiding the need for trusted decision makers.
    – Pieter Wuille
    Jun 13, 2022 at 17:22
  • 1
    I think the possibility of the users of the system deciding to reject a hashrate majority consensus-valid chain does factor in to the equation, but only to the extent that it is hardly more than a mutually assured destruction mechanism. Yes, if absolutely need be, the users can decided to reject the PoW-decided chain, but it likely be a very catastrophic event (only slightly less bad than the alternative of the attack going through), so much that it becomes undesirable for able parties to even attempt the attack.
    – Pieter Wuille
    Jun 13, 2022 at 17:24
  • 2
    In PoS stakers and nodes are still separate groups. The alternative with just nodes is traditional consensus algorithms like PBFT - but you can't compare these algorithms, as they operate in a very different context (needs trusted parties!).
    – Pieter Wuille
    Jun 13, 2022 at 17:48
  • 2
    The Bitcoin ecosystem (humans) needing to come to consensus to discard the result of PoW seems catastrophic to me. Yes, it may be needed in order to avoid a 51% attacker preventing progress to the chain for example, but it's making it evident that PoW just doesn't work, as its entire point of existence is avoiding the need for humans to decide which version of history is the correct one. That combined with the fact that it inevitably requires political consensus about this decision to be made, which would raise questions about whether it can happen again.
    – Pieter Wuille
    Jun 13, 2022 at 18:04
  • 2
    So to be clear my point here is: when we're talking about Bitcoin nodes, we're assuming nodes which faithfully follow the most-work valid chain, even if that chain is produced by an attacker in some form. While yes, indeed, node operators could jointly decide to discard that result and switch over to another chain, and functionality for doing so even exists in Bitcoin Core (the invalidateblock RPC), I'd argue that if you're talking about a system where this mechanism actually needs to be used and is relied upon, it's not anything like Bitcoin as we know it anymore.
    – Pieter Wuille
    Jun 13, 2022 at 18:10
0

A 51% attack on Bitcoin is an attack on the Bitcoin network carried out by a group of miners who control more than 50% of the network's total computing power. When such a group of miners work together, they can manipulate the network and alter or reverse transactions in their own interest.

Improve this answer
2
  • Well some say they would be detected and expelled.
    – Mercedes
    Feb 26 at 13:03
  • You could perhaps clarify that even with a majority of the hashrate, they can only alter transactions that they sent themselves.
    – Murch ♦
    Mar 2 at 21:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.