Actually

This project demonstrates a two-way cross-chain handshake between two smart contracts deployed on different networks. A relayer service listens for events and sends transactions across chains to complete the handshake.

How to run the relayer fill the .env file presnt in ./go-relayer directory with the following variables: SEPOLIA_RPC=https://eth-sepolia.g.alchemy.com/v2/cuJGeZ26qywahhkli6epc2XqIOoZJhFF RPC_B=https://arb-sepolia.g.alchemy.com/v2/AJAOv3C3nutHF-2duNghabWWPgzD1hxw RELAYER_PVT_KEY=your_private_key CONTRACT_A_ADDR=0xddA72c449E9e8Cf5b440859c6AA78d8E1803FAc3 CONTRACT_B_ADDR=0x8c7A0E5A49f6f4ca7AA8EE10933Ab5E7bCE85FeF CHAIN_IDA=11155111 # sepolia chain id CHAIN_IDB=421614 # arbitrum sepolia chain id build and run the relayer: go build ./cmd/relayer ./relayer how to initiate handshake manually any user with a funded sepolia wallet can initiate a handshake by calling initiateHandshake on contract a.

export CONTRACT_A_ADDR=0xddA72c449E9e8Cf5b440859c6AA78d8E1803FAc3 export SEPOLIA_RPC=https://eth-sepolia.g.alchemy.com/v2/cuJGeZ26qywahhkli6epc2XqIOoZJhFF export YOUR_PRIVATE_KEY=your_private_key_here export SESSION=

cast send $CONTRACT_A_ADDR "initiateHandshake(uint256)" $SESSION
--rpc-url $SEPOLIA_RPC
--private-key $YOUR_PRIVATE_KEY this will emit a syn event, and the relayer will automatically complete the handshake across the two networks.

1. Persistence Layer Currently if the relayer restarts it forgets which sessions it has already handled which could cause duplicate transactions.

Solution: Integrate a simple LevelDB instance to store:

Last processed block numbers Session IDs that have already been relayed 2. Dockerization & CI Pipeline Right now the relayer requires manual environment setup.

Solution:

Add a Dockerfile and docker-compose.yaml to simplify deployment. Add tests to validate the entire relayer flow during CI builds. 3. Transaction Retry with Exponential Backoff If a transaction fails due to issues like RPC errors, nonce gaps, or mempool reorgs, the relayer currently just logs the error and moves on.

Solution: Use a retry mechanism such as go-retry or implement a simple exponential backoff system to retry failed transactions automatically