The Merge is a protocol upgrade that is being rolled out on all Ethereum networks being kept around. This procotol upgrade will be rolled out on Mainnet. Ethereum validators will need to take some measures and some actions to make sure their staking machine will work during this upgrade and that they will keep running well after the upgrade. Tests are already underway and the Merge upgrade has been applied to testnets already.
This guide is meant for people who are managing and maintaining staking machines. It will explain which steps or actions you should take in order to be ready for the Merge.
Here are the clients that are ready for the merge on Mainnet and their version. You must update your clients to these versions or later before the Merge on Mainnet. You must update your clients before September 6th 2022 at 11:34:47am UTC.
- Geth: v1.10.23 (or later)
- Nethermind: v1.14.0 (or later)
- Besu: v22.7.1 (or later)
- Erigon: v2022.08.03-alpha (or later)
- Teku: v22.8.1 (or later)
- Lighthouse: v3.0.0 (or later)
- Prysm: v3.0.0 (or later)
- Nimbus: v22.8.1 (or later)
- Lodestar: v1.0.0 (or later)
There are some risks your staking machine will stop performing its duties if you are not well prepared. There are some risks your Ethereum validator will miss some rewards if you are not well prepared.
The Merge on Mainnet is happening in a few different steps. The first one will happen on September 6th 2022 at 11:34:47am UTC. It's a fork called Bellatrix to prepare for the Merge. The final and definitive step will happen around September 15th 2022. The uncertainty around the exact date and time depends in part on the current miners and the overall Ethereum network hash rate.
Make sure to follow the official Ethereum blog and watch for announcements there. You can also subscribe to this newsletter.
- Make sure you are running your own execution client.
- Make sure your Ethereum clients are updated.
- Make sure you are using the new configuration options for your Ethereum clients.
- Make sure you have a configured JWT token file for the engine API.
- Make sure you have a configured fee recipient address.
- Make sure you have a configured MEV solution if you want one.
Running your own execution client is going to be required after the Merge for an Ethereum validator. Some Ethereum validators are entirely relying on public infrastructure like Infura or Alchemy for their execution data. If you are in this case, you will need to install and configure your own execution client.
If you are running your own execution client and using a public infrastructure provider as a fallback endpoint, that fallback endpoint will not work after the merge. Using multiple execution clients after the Merge is going to be slightly more complex. This is out of scope for this guide.
The 2 main concerns for an Ethereum validator when running your own execution client is disk usage and bandwidth usage.
If you are not currently running your own execution client and if you used a guide or a tool to setup your staking machine, you should first check to see if that guide or tool has a section on how to setup an execution client. If you don't know where to start, check out the client setup tools and guides. Our support can also help you here.
Syncing an execution from nothing can take a few hours or a few days. Make sure to install and configure your execution client at least a week before the Merge to be safe.
It is not uncommon for an execution client to use more than 650GB to store its data on Mainnet. If your disk size is less than 2TB for storing everything, you should consider migrating and replacing that disk with one which is at least 2TB in size. You could possibly still manage to run an effective Ethereum validator with only a 1TB disk but that is going to require a lot of efforts which might not be worth it for the price difference of a new disk. Alternatives include adding another a new disk to your machine if it has an available slot or running your execution client on another machine on your local network. Make sure to get in touch with our support if you are interested in those alternatives.
Pruning your execution client database is a good strategy for managing your execution client disk usage. If you are using Geth, you should prune it regularly and you should prune it just before the Merge. Geth currently only offers offline prunning that takes a few hours to complete. You should know that having an offline execution client after the Merge will result in your validator failing to perform all its duties meaning that you will lose rewards during that time.
Other execution clients have different strategies for managing data growth. Some are leaner in terms of disk size usage to get started with and some include automatic online pruning. You should consider running a minority execution client.
I suggest you implement monitoring and alerting to help you find out when you are low on available disk space.
An execution client will use a lot of bandwidth. Reducing the number of peers is one way of reducing bandwidth usage. It comes with some risks and the con of potentially being out of sync at the wrong time and making your Ethereum validator lose some rewards. The default configuration option for the number of peers should be your baseline value.
Geth is currently the supermajority execution client. It is used to build almost all blocks right now. This is not currently a big issue but it will become one after the Merge. You should consider running a minority execution client. We have some great execution clients that work well for an Ethereum validator that should be considered:
There are new specification changes that were added recently that will be needed for the Merge. If you have not updated your Ethereum clients regularly, you will need to update them before the Merge.
If you used a guide or a tool to setup your staking machine, you should first check out with that guide or tool on how you can update your Ethereum clients. If they don't have any section on how to do it, you can check with each client or get touch with our support.
The communication between your consensus client and your execution client is going to change with the Merge. They will use this new Engine API which will be served at a port independant from the JSON-RPC API currently in use.
You are likely going to need to use a new configuration option for this endpoint or at least update it on your consensus client beacon node. The default port for the Engine API is 8551 and the default port for the JSON-RPC API is 8545. You might not need to change any configuration option on your execution client, but you will likely need to modify your beacon node configuration. Here are the common values you will likely need to add or modify for each consensus client assuming your execution client engine API endpoint is at http://localhost:8551:
- Prysm beacon node:
--execution-endpoint http://localhost:8551 - Nimbus:
--web3-url=http://localhost:8551 - Lodestar beacon node:
--execution.urls http://localhost:8551 - Teku:
--ee-endpoint http://localhost:8551 - Lighthouse beacon node:
--execution-endpoint http://localhost:8551
As usual, when changing the configuration for your consensus client, you will need to reload this configuration and probably restart the client.
The new Engine API used to communicate between the execution client and the consensus client requires authentication which is provided with a JWT token stored in a file. There are various ways to configure this. Here is a simple one.
Create a JWT token file in a neutral location and make it readable to everyone. We will use the /var/lib/ethereum/jwttoken location to store the JWT token file.
$ sudo mkdir -p /var/lib/ethereum
$ openssl rand -hex 32 | tr -d "\n" | sudo tee /var/lib/ethereum/jwttoken
$ sudo chmod +r /var/lib/ethereum/jwttoken
Each execution client will need a different configuration option for the JWT token file. Here are the common values:
- Geth:
--authrpc.jwtsecret /var/lib/ethereum/jwttoken - Erigon:
--authrpc.jwtsecret /var/lib/ethereum/jwttoken - Besu:
--engine-jwt-secret=/var/lib/ethereum/jwttoken - Besu on mainnet:
--engine-rpc-enabled --engine-jwt-secret=/var/lib/ethereum/jwttoken - Nethermind:
--JsonRpc.JwtSecretFile=/var/lib/ethereum/jwttoken
As usual, when changing the configuration for your execution client, you will need to reload this configuration and probably restart the client.
Besu requires
--engine-rpc-enabledwhen no TTD has been configured for the network. Once mainnet has an announced TTD, this flag is no longer necessary
- Prysm beacon node:
--jwt-secret /var/lib/ethereum/jwttoken - Nimbus:
--jwt-secret=/var/lib/ethereum/jwttoken - Lodestar beacon node:
--jwt-secret /var/lib/ethereum/jwttoken - Teku:
--ee-jwt-secret-file /var/lib/ethereum/jwttoken - Lighthouse beacon node:
--execution-jwt /var/lib/ethereum/jwttoken
As usual, when changing the configuration for your consensus client, you will need to reload this configuration and probably restart the client.
After the Merge, validators will replace miners in building blocks and getting the associated rewards. The validators will receive the tips from each included transaction on their proposed block. For this, they will need to configure a fee recipient address. This will represent the address where tips are sent to when your validator propose a block. Those fees will be available right away for any transaction.
This address can be configured at different layers depending on which client you are using. It can be at the execution client layer, at the consensus beacon node layer, it can be at the consensus validator client layer and it can even be for each different validator key. Here are the common values to configure this on the consensus client. Make sure to replace the 0x0000000000000000000000000000000000000000 address with your own Ethereum address that you control where you want to receive the transaction tips.
- Prysm:
--suggested-fee-recipient 0x0000000000000000000000000000000000000000 - Nimbus:
--suggested-fee-recipient=0x0000000000000000000000000000000000000000 - Lodestar validator:
--suggestedFeeRecipient 0x0000000000000000000000000000000000000000 - Teku:
--validators-proposer-default-fee-recipient=0x0000000000000000000000000000000000000000 - Lighthouse:
--suggested-fee-recipient 0x0000000000000000000000000000000000000000
As usual, when changing the configuration for your consensus client, you will need to reload this configuration and probably restart the client.
There are some privacy implications in using a fee recipient address. That address will be forever linked with your validator when it proposes a block after the Merge. If you want to maximize your privacy, you should use your validator deposit address as your fee recipient address if you still have control over it since it is already linked with your validator.
Here are the detailed configuration options for the fee recipient for each client and their documentation.
Maximal extractable value (MEV) refers to the maximum value that can be extracted from block production in excess of the standard block reward and gas fees by including, excluding, and changing the order of transactions in a block. After the Merge, there will be an opportunity for Ethereum validators to get that value. If you want to get that value, you will need to install some additional software and configure your consensus client to communicate with that software when the time comes to propose a block.
The use of mev-boost and obtaining MEV by a validator is entirely optional. A validator might want to avoid obtaining MEV for any reason. There are plenty of good reasons to avoid MEV. It is a decision left to each staker. However, if you want to maximize your profits, you should consider obtaining that additional value.
A preview of MEV and mev-boost can be seen on https://youtu.be/sZYJiLxp9ow
Create a user account for the service to run under. This account will not be able to log into the machine. It will only be used to run the service.
$ sudo useradd --no-create-home --shell /bin/false mevboostDownload the latest stable version of mev-boost from https://github.com/flashbots/mev-boost/releases (avoid any pre-release version). As of this date, the latest stable release version is 1.4.0 . Adjust the following instructions accordingly if there is a newer stable release version with a different archive name. The file name should likely end with linux_amd64.tar.gz (for linux and AMD64 instructions set). Use a different archive if you are on a different architecture or a different operating system.
$ cd ~
$ wget https://github.com/flashbots/mev-boost/releases/download/v1.4.0/mev-boost_1.4.0_linux_amd64.tar.gzVerify that the SHA256 Checksum as shown in the checksums.txt file is the same as the file we just downloaded.
$ sha256sum mev-boost_1.4.0_linux_amd64.tar.gzExtract the archive. Install mev-boost globally and remove the download leftovers.
$ tar xvf mev-boost_1.4.0_linux_amd64.tar.gz
$ sudo cp mev-boost /usr/local/bin
$ rm mev-boost LICENSE README.md mev-boost_1.4.0_linux_amd64.tar.gz
$ sudo chown mevboost:mevboost /usr/local/bin/mev-boostCreate a systemd service file to store the service config which tells systemd to run mev-boost as the mevboost user.
$ sudo nano /etc/systemd/system/mevboost.servicePaste the following into the file to run mev-boost on Mainnet. You must replace https://example.com in this configuration with one or many existing relays. We have a list of relays you can explore. Exit and save once done (Ctrl + X, Y, Enter).
[Unit]
Description=mev-boost (Mainnet)
Wants=network-online.target
After=network-online.target
[Service]
Type=simple
User=mevboost
Group=mevboost
Restart=always
RestartSec=5
ExecStart=mev-boost \
-mainnet \
-min-bid 0.05 \
-relay-check \
-relays https://example.com
[Install]
WantedBy=multi-user.targetYou can add multiple relays comma-separated to the -relays flag, like this: -relays https://relay1,https://relay2.
Selecting your relays can be an important decision for some stakers. Some relay might sensor transactions. Some relay might be more ethical in terms of which strategies they accept or not. You should do your own diligence when selecting which relay you want to use.
Reload systemd to reflect the changes. If you edit that file again to add or remove relays, you need will to reload systemd again to reflect the changes before restarting your mev-boost service.
$ sudo systemctl daemon-reloadAnd then start the service with the following command and check the status to make sure it’s running correctly.
$ sudo systemctl start mevboost
$ sudo systemctl status mevboostIf mev-boost crashes with "SIGILL: illegal instruction" then you need to use a portable build.
If you did everything right, it should say active (running) in green. If not then go back and repeat the steps to fix the problem. Press Q to quit. Lastly, enable mev-boost to start on boot.
$ sudo systemctl enable mevboostAdd this flag to the ExecStart of your consensus client and, if needed, validator client service. Note that client flags can be in flux
through August 2022. When in doubt consult each client's --help.
- Prysm consensus:
--http-mev-relay=http://127.0.0.1:18550 - Prysm validator:
--enable-builder - Nimbus combined:
--payload-builder=true --payload-builder-url=http://127.0.0.1:18550 - Lodestar consensus:
--builder --builder.urls http://127.0.0.1:18550 - Lodestar validator:
--builder - Teku combined:
--validators-builder-registration-default-enabled=true --builder-endpoint=http://127.0.0.1:18550 - Lighthouse consensus:
--builder http://127.0.0.1:18550 - Lighthouse validator:
--builder-proposals
Tell systemd you made the changes: sudo systemctl daemon-reload
And restart the service(s) you changed: sudo systemctl restart SERVICENAME
When a new version is released, you can update mev-boost. Find the latest stable version of mev-boost from https://github.com/flashbots/mev-boost/releases. Download the archive. The file name should likely end with linux_amd64.tar.gz (for linux and AMD64 instructions set). Use a different archive if you are on a different architecture or a different operating system. Verify the archive checksum, extract the archive, stop the service, install mev-boost globally, remove the download leftovers and restart the service.
$ cd ~
$ wget https://github.com/flashbots/mev-boost/releases/download/v1.4.0/mev-boost_1.4.0_linux_amd64.tar.gz
$ sha256sum mev-boost_1.4.0_linux_amd64.tar.gz
$ tar xvf mev-boost_1.4.0_linux_amd64.tar.gz
$ sudo systemctl stop mevboost
$ sudo cp mev-boost /usr/local/bin
$ rm mev-boost LICENSE README.md mev-boost_1.4.0_linux_amd64.tar.gz
$ sudo chown mevboost:mevboost /usr/local/bin/mev-boost
$ sudo systemctl start mevboostWhen proposing a block with an MEV relay, your validator will blindly sign a block and that block will be published by the relay itself or a builder behind it. The configured fee recipient address for your validator will not directly receive the associated rewards for using that relay. It will not be used for the actual fee recipient field for that block. Instead and in most cases, a separate transaction will be included in that block that will send your profit to the configured fee recipient address. That should be the case even if you have multiple different configured fee recipient addresses when running multiple validators. Make sure to look at the included transactions in your block to find a transaction with your MEV rewards. Make sure to contact the MEV operator support if you feel like you did not receive your share of the profit. Make sure to get in touch with EthStaker if you feel like your did not receive your share of the profit.
If you have any question or if you need additional support, make sure to get in touch with the EthStaker community on:
- Discord: discord.io/ethstaker
- Reddit: reddit.com/r/ethstaker
Based on CoinCashew's Ethereum Merge Upgrade Checklist for Home Stakers and Validators.