Multicast sending Ethereum transactions #5452
Assignees
Comments
This was referenced Feb 13, 2025
github-merge-queue bot
pushed a commit
that referenced
this issue
Feb 20, 2025
### Description Add multicast tx submission for Ethereum fallback provider. `EthereumFallbackProvider` will use different strategies to query chain RPCs: 1. If the method is `eth_sendRawTransaction`, it will make a parallel query to all RPC endpoints and return the first successful result. If no success, it will retry 4 times. 2. For other methods, the current functionality does not change: it will sequentially query RPC endpoints according to their priorities and adjust priorities depending on success/failure. It also retries 4 times. ### Related issues - Contributes into #5452 ### Backward compatibility Yes ### Testing Added unit tests --------- Co-authored-by: Danil Nemirovsky <[email protected]>
fmorency
pushed a commit
to fmorency/hyperlane-agents
that referenced
this issue
Feb 20, 2025
### Description Add multicast tx submission for Ethereum fallback provider. `EthereumFallbackProvider` will use different strategies to query chain RPCs: 1. If the method is `eth_sendRawTransaction`, it will make a parallel query to all RPC endpoints and return the first successful result. If no success, it will retry 4 times. 2. For other methods, the current functionality does not change: it will sequentially query RPC endpoints according to their priorities and adjust priorities depending on success/failure. It also retries 4 times. ### Related issues - Contributes into hyperlane-xyz/hyperlane-monorepo#5452 ### Backward compatibility Yes ### Testing Added unit tests --------- Co-authored-by: Danil Nemirovsky <[email protected]>
github-merge-queue bot
pushed a commit
that referenced
this issue
Feb 21, 2025
) ### Description Make only one retrying attempt if we get non-retryable error. At the moment, we are doing 12 RPC calls even if all of them returned a non-retryable error in multicast approach of EthereumFallbackProvider. While we still would like to wait for all providers to return their result when we do multicast, we would like to avoid retrying since it produces a lot of requests with the same failure result. It leads to elevated RPC failure metrics and higher number of alerts. ### Related issues - Contributes into #5452 ### Backward compatibility Yes ### Testing TBD --------- Co-authored-by: Danil Nemirovsky <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The text was updated successfully, but these errors were encountered: