Notes:

• Setup:

  • 3 mev-geth nodes (jason/v1.11-mev branch) and 1 bootnode (to allow peering) on a shared docker network

  • The 3 nodes are mining, with 8 threads, and propagate txs/blocks upon mining as expected (though DAG generation can take a long time)

  • --override.london=0 to force the london hardfork from genesis block

    • Confimed by making sure baseFeePerGas is exposed in the block header

    

  • All nodes have their RPCs (port :8545) enabled + binded to :1111, :1112, :1113, :1114 ports on the host machine (to send txs to them)

  • Also a network dashboard to help debug (making sure blocks are mined, peers are alive), can be visited by opening the random port (assigned by docker after running) in the browser

    

• Running

  • First, build the required docker images

    • ./build.sh

  • Next, deploy + start the containers

    • ./run deploy

  • To stop/start after:

    • ./run stop and ./run start

  • send sample tx (to a random peer RPC):

    • npm run sendTx

run deploy
run start
run stop
run login nodename
run geth nodename
run log nodeName

^nodeName = ethbn/eth1/eth2/eth3

Next steps:

• Migrating to ethers, relevant ethers-io/ethers.js#1685

• Adjust gaslimit in the genesis.json file to test for the following conditions:

  • Each block’s baseFeePerGas can increase/decrease by up to 12.5% depending on how full the block is relative to the previous one: e.g. 100% full -> +12.5%, 50% full -> same, 0% full -> -12.5%.

  • Making sure bundles don't break this^

• Migrating old bundle flow rate script here to stress test the basefee + how tip must be adjusted to get included