A transaction propagation delay causes "onchain coins could not be sent" error about boltz-backend HOT 6 CLOSED

Have the transaction index enabled txindex=1

https://docs.boltz.exchange/en/latest/deployment/

The boltz backend explicitely requires the txindex to be enabled. It's not about propagation of the transaction but about the this.chainClient.getRawTransactionVerbose which calls the Bitcoin Core RPC and requires txindex

from boltz-backend.

What I meant is I have LND with neutrino and Bitcoind with txindex enabled on different servers. When boltz-backend calls the Bitcoin Core RPC this.chainClient.getRawTransactionVerbose, it returns the error because the transaction that is broadcasted from LND is not yet propagated to the Bitcoind.

from boltz-backend.

Ah I see. Interesting setup you got there. I guess we could work around that propagation issue.

But the real question is: why would you want to do that in the first place? 😅
Why run LND with neutrino if you could just run it alongside your Bitcoin Core?

from boltz-backend.

The instance I run LND has limited disk space. That's why I have to run another instance for Bitcoind. I added sleep function, but it does not solve the fundamental issue. Do you have any suggestions?

from boltz-backend.

Do you have any suggestions?

That code snippet you sent in the first message assumes that the transcation was broadcasted through the Bitcoin Core node Boltz is connected to.

To fix that we have two options:

1. query the raw transaction that was sent not through Bitcoin Core but with the LND gRPC
2. you connect your LND to the Bitcoin Core running on the other instance

The first option could even be implemented as a fallback. But I'd prefer the second one, because I wouldn't use Neutrino in producton. What do you think?

from boltz-backend.

Closing, since we don't want to optimize for Neutrino backend usage.

from boltz-backend.