Two weeks in trying to sync an Ethereum archive / tracing node.

What's the point of a full node if it takes forever to find and calculate an answer to a query?

@smartbrain full nodes don't need to answer arbitrary queries, let alone efficiently. The point is to validating all state transitions to ensure integrity and being able to answer queries regarding tracked wallets. E.g. you can't efficiently query arbitrary addresses in bitcoind, it's a (linear?) search through the UTXO set, not giving you any historical data.

Or did you just mean syncing/validating is slow in which case I agree that's a problem.

@sebx2a not got that far yet but I want to know how different it is for historical states as that is what they say is very different between full and .. full++ nodes. Trying to sync a tracing node to see what it can do but two, almost three weeks in and as you can see, still nowhere near to finishing.


full nodes don’t need to answer arbitrary queries, let alone efficiently.

AFAIK, you need a fully indexing node to use LN. I know it’s a requirement for Eclair and it was a requirement for CLightning last time I used it.

I personally don’t think the only purpose of a full node is validating. @smartbrain

@verretor @sebx2a From what I read you need a tracing+archival ethereum node to get any reasonable response time for any old state - some said full nodes take hours. I know that I cannot reasonably run any dex with a full node on my 8th gen i7 as it works, but keeps bogging down, so I gave up and just used metamask's infura. So now I'm wondering if things will be significantly faster if I sync it in archive+tracing node. That said, I doubt I can sync, definitely not in one month at this rate.

Sign in to participate in the conversation
Bitcoin Mastodon

Bitcoin Maston Instance