1

I do not know all of the details of how ChainLink's nodes and oracles work (hence why I am on here) but I did a little research on my own. According to a Gemini blog post by Cryptopedia, chainlink uses an aggregating contract so that

if five nodes deliver one answer from a weather sensor and two other nodes deliver a different answer, the Chainlink Aggregating Contract will know that those two nodes are faulty (or dishonest) and discard their answers. In this manner, Chainlink nodes can validate data from a single source.

So it seems ChainLink's oracles are based on a majority rule type system. So, if an oracle holds a substantial amount of LINK and is reputable enough, could it influence/control/change the number generated in a specific call for ChainLink's VRF? Would it be considered a 51% attack?

I understand it usually would not be worth it because of all the LINK they'd have to have staked would be devalued. (I think?) However, if the number was to be generated for a lottery and the lottery was a few million dollars in ETH (or let's just say "worth" manipulating the number), could it be possible?

Improve this question

2 Answers 2

Reset to default
3

In Chainlink VRF, they are generated by one specific node, not by multiple nodes. There is no 51% attack because there is not a aggregating contract for Chainlink VRF.

Maybe your next question is how can I trust the random number provided by a single node? The answer is that you don't need to 100% trust VRF node, because you can verify it. The full name of VRF is Verifiable random function and the random words generated by the algorithm can be verified by seeds input by the requester.

In the scenario, the seed is block hash in the future block that cannot be decided by VRF node. With the seed, VRF node can use its private key to generate random words which can be verified with seed and node's public key by anyone.

Hope it helps!

Improve this answer
1
  • 1
    Yes, if you control 51% mining power AND chainlink private key, you can theoretically control random words. It is much harder than control one of these 2.
    – Frank Kong
    Commented Feb 8, 2023 at 1:40
2

If you are referring specifically to the VRF, where a random number is generated then no it is not susceptible to a 51% attack hence the Verifiable nature of the function. For more arbitrary pieces of information that are already hard to verify such as the weather yes, that may be possible.

Improve this answer

Your Answer

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

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