Guide | Operation Client Diversity: Migrate Prysm to Teku

Step by step guide on how to switch for the solo home or cloud staker.

PreviousEthereum Merge Upgrade Checklist for Home Stakers and Validators NextGuide: How to buy ETH

Last updated 1 year ago

🎊 Support us on Gitcoin Grants: We improve this guide with your support!🙏

⏩ Complete step by step guide

The following steps align with our mainnet guide. You may need to adjust file names and directory locations where appropriate. The core concepts remain the same.

As per best practices, always try everything on a testnet before doing it for real on mainnet.

🔥 Problem: Why the commotion?

If improving the stability of the beacon chain is not a good enough reason for you to switch from Prysm to either Teku or Nimbus, you also need to consider that due to the design of the beacon chain you will be subject to severe financial penalties if Prysm ever has an issue. ~Lamboshi on Twitter

🚀 Solution: Increase client diversity by migrating to Teku

🚧 How to Migrate from Prysm to Teku

PegaSys Teku (formerly known as Artemis) is a Java-based Ethereum 2.0 client designed & built to meet institutional needs and security requirements. PegaSys is an arm of ConsenSys dedicated to building enterprise-ready clients and tools for interacting with the core Ethereum platform. Teku is Apache 2 licensed and written in Java, a language notable for its materity & ubiquity.

⛓️ 1. Setup Teku CL

Install git.

sudo apt-get install git -y

Install Java 17 LTS

sudo apt update
sudo apt install openjdk-17-jdk -y

Verify Java 17+ is installed.

java --version

Install and build Teku.

mkdir ~/git
cd ~/git
git clone https://github.com/ConsenSys/teku.git
cd teku
./gradlew distTar installDist

This build process may take a few minutes.

Verify Teku was installed properly by displaying the version.

cd $HOME/git/teku/build/install/teku/bin
./teku --version

Copy the Teku binary file to /usr/bin/teku

sudo cp -r $HOME/git/teku/build/install/teku /usr/bin/teku

Teku combines both the beacon chain and validator into one process.

Setup a directory structure for Teku.

sudo mkdir -p /var/lib/teku
sudo mkdir -p /etc/teku
sudo chown $USER:$USER /var/lib/teku

Create your teku.yaml configuration file.

sudo nano /etc/teku/teku.yaml

Paste the following configuration into the file.

# network
network: "mainnet"
initial-state: "https://beaconstate.info/eth/v2/debug/beacon/states/finalized"

# validators
validators-graffiti: "<MY_GRAFFITI>"

# execution engine
ee-endpoint: http://localhost:8551 
ee-jwt-secret-file: "/secrets/jwtsecret" 

# fee recipient
validators-proposer-default-fee-recipient: "<0x_CHANGE_THIS_TO_MY_ETH_FEE_RECIPIENT_ADDRESS>"

# metrics
metrics-enabled: true
metrics-port: 8008

# database
data-path: "/var/lib/teku"
data-storage-mode: "prune"

Replace**<0x_CHANGE_THIS_TO_MY_ETH_FEE_RECIPIENT_ADDRESS>** with your own Ethereum address that you control. Tips are sent to this address and are immediately spendable, unlike the validator's attestation and block proposal rewards.
Replace <MY_GRAFFITI> with your own graffiti message. However for privacy and opsec reasons, avoid personal information. Optionally, leave it blank by deleting the flag option.

🛑 2. Stop and disable Prysm

Stop and disable the Prysm services. Choose your guide.

sudo systemctl stop validator beacon-chain
sudo systemctl disable validator beacon-chain

sudo systemctl stop prysmvalidator
sudo systemctl stop prysmbeacon
sudo systemctl disable prysmvalidator
sudo systemctl disable prysmbeacon

Confirm the Prysm validator is stopped by checking the service status.

service validator status

service prysmvalidator status

Delete existing Prysm validators keys so that there's no accidental starting of Prysm's validator.

sudo rm -rf ~/.eth2validators

sudo rm -rf /var/lib/prysm/validator

As a double check, verify that Prysm validator can't find it's keys by starting the validator again.

sudo systemctl start validator

 sudo systemctl start prysmvalidator

Observe the logs and check for errors about missing validator keys.

journalctl -fu validator

journalctl -fu prysmvalidator

Finally, stop Prysm validator.

sudo systemctl stop validator

 sudo systemctl stop prysmvalidator

Before continuing - Required Waiting Period !!!

Wait until your validator's last attestation is in a finalized epoch - usually about 15 minutes.

By waiting for a finalized epoch, there's no need to migrate the slashing database.

Confirm that your validator has stopped attesting with block explorer beaconcha.in or beaconscan.com

🧱 3. Update firewall / port forwarding.

Allow Teku ports:

sudo ufw allow 9000/tcp
sudo ufw allow 9000/udp

Delete Prysm firewall rules:

sudo ufw delete allow 13000/tcp
sudo ufw delete allow 12000/udp

Verify that your firewall configuration is correct.

sudo ufw status numbered

Example output of firewall configuration:

     To                         Action      From
     --                         ------      ----
[ 1] 22/tcp                     ALLOW IN    Anywhere       # SSH
[ 2] 9000/tcp                   ALLOW IN    Anywhere       # eth2 p2p traffic
[ 3] 9000/udp                   ALLOW IN    Anywhere       # eth2 p2p traffic
[ 4] 30303/tcp                  ALLOW IN    Anywhere       # eth1
[ 5] 22/tcp (v6)                ALLOW IN    Anywhere (v6)  # SSH
[ 6] 9000/tcp (v6)              ALLOW IN    Anywhere (v6)  # eth2 p2p traffic
[ 7] 9000/udp (v6)              ALLOW IN    Anywhere (v6)  # eth2 p2p traffic
[ 8] 30303/tcp (v6)             ALLOW IN    Anywhere (v6)  # eth1

Your router's port forwarding setup or cloud provider settings will need to be updated to ensure your validator's firewall ports are open and reachable.

You'll need to add new port forwarding rules for Teku and remove the existing Prysm port forwarding rules.

Optional - Update your server and reboot for best practice.

sudo apt update && sudo apt upgrade
sudo apt dist-upgrade && sudo apt autoremove
sudo reboot

🗝️ 4. Import Validator Keys

Copy your validator_keys directory to the data directory we created above and remove the extra deposit_data file. If you no longer have the validator keys on your node, you will need to restore from file backup or restore from secret recovery phrase.

cp -r $HOME/eth2deposit-cli/validator_keys /var/lib/teku
rm /var/lib/teku/validator_keys/deposit_data*

🛑 FINAL WARNING REMINDER !!! Do not start the Teku validator client until you have stopped the Prysm one, or you will get slashed (penalized and exited from the system).

Wait until your validator's last attestation is in a finalized epoch - usually about 15 minutes.

Confirm that your validator has stopped attesting with block explorer beaconcha.in or beaconscan.com

Storing your keystore password in a text file is required so that Teku can decrypt and load your validators automatically.

Replace <my_keystore_password_goes_here> with your keystore password between the single quotation marks and then run the command to save it to validators-password.txt

echo '<my_keystore_password_goes_here>' > $HOME/validators-password.txt

Confirm that your keystore password is correct.

cat $HOME/validators-password.txt

Move the password file and make it read-only.

sudo mv $HOME/validators-password.txt /etc/teku/validators-password.txt
sudo chmod 600 /etc/teku/validators-password.txt

Clear the bash history in order to remove traces of keystore password.

shred -u ~/.bash_history && touch ~/.bash_history

When specifying directories for your validator-keys, Teku expects to find identically named keystore and password files. For example keystore-m_12221_3600_1_0_0-11222333.json and keystore-m_12221_3600_1_0_0-11222333.txt

Create a corresponding password file for every one of your validators.

for f in /var/lib/teku/validator_keys/keystore*.json; do cp /etc/teku/validators-password.txt /var/lib/teku/validator_keys/$(basename $f .json).txt; done

Verify that your validator's keystore and validator's passwords are present by checking the following directory.

ll /var/lib/teku/validator_keys

Add validator-keys configuration to teku.yaml

cat >> /etc/teku/teku.yaml << EOF
validator-keys: "/var/lib/teku/validator_keys:/var/lib/teku/validator_keys"
EOF

🚀 5. Setup and start the Teku service

Run the following to create a unit file to define yourbeacon-chain.service configuration.

cat > $HOME/beacon-chain.service << EOF
# The eth beacon chain service (part of systemd)
# file: /etc/systemd/system/beacon-chain.service 

[Unit]
Description     = eth beacon chain service
Wants           = network-online.target
After           = network-online.target 

[Service]
User            = $USER
ExecStart       = /usr/bin/teku/bin/teku -c /etc/teku/teku.yaml
Restart         = on-failure
Environment     = JAVA_OPTS=-Xmx5g

[Install]
WantedBy	= multi-user.target
EOF

Move the unit file to /etc/systemd/system

sudo mv $HOME/beacon-chain.service /etc/systemd/system/beacon-chain.service

Update file permissions.

sudo chmod 644 /etc/systemd/system/beacon-chain.service

Run the following to enable auto-start at boot time and then start your beacon node service.

sudo systemctl daemon-reload
sudo systemctl enable beacon-chain
sudo systemctl start beacon-chain

Syncing the beacon node might take up to 36 hours depending on your hardware. Keep validating using your current Prysm setup until it completes. However, thanks to Teku's Checkpoint sync, you'll complete this step in a few minutes.

Syncing is complete when your beacon node's slot matches that of a block explorer's slot number (i.e. https://beaconcha.in/)

Check the beacon node syncing progress with the following:

journalctl -fu beacon-chain

Check the logs to verify the services are working properly and ensure there are no errors.

#view and follow the log
journalctl -fu beacon-chain

Confirm that your new Teku validator has started attesting with block explorer beaconcha.in or beaconscan.com

🛠 Some helpful systemd commands

🗄 Viewing and filtering logs

#view and follow the log
journalctl -fu beacon-chain

#view log since yesterday
journalctl --unit=beacon-chain --since=yesterday

#view log since today
journalctl --unit=beacon-chain --since=today

#view log between a date
journalctl --unit=beacon-chain --since='2020-12-01 00:00:00' --until='2020-12-02 12:00:00'

🔎 View the status of the beacon chain

sudo systemctl status beacon-chain

🔁 Restart the beacon chain

sudo systemctl restart beacon-chain

🛑 Stop the beacon chain

sudo systemctl stop beacon-chain

📡 6. Update Prometheus and Grafana monitoring

Select your Ethereum execution engine and then re-create your prometheus.yml configuration file to match Teku's metric's settings.

cat > $HOME/prometheus.yml << EOF
global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'nodes'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'geth'
     scrape_interval: 15s
     scrape_timeout: 10s
     metrics_path: /debug/metrics/prometheus
     scheme: http
     static_configs:
     - targets: ['localhost:6060']
EOF

cat > $HOME/prometheus.yml << EOF
global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'nodes'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'besu'
     scrape_interval: 15s
     scrape_timeout: 10s
     metrics_path: /metrics
     scheme: http
     static_configs:
     - targets:
       - localhost:9545
EOF

cat > $HOME/prometheus.yml << EOF
global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'nodes'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'nethermind'
     scrape_interval: 15s
     scrape_timeout: 10s
     honor_labels: true
     static_configs:
       - targets: ['localhost:9091']
EOF

Nethermind monitoring requires Prometheus Pushgateway. Install with the following command.

sudo apt-get install -y prometheus-pushgateway

Pushgateway listens for data from Nethermind on port 9091.

cat > $HOME/prometheus.yml << EOF
global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'nodes'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'openethereum'
     scrape_interval: 15s
     scrape_timeout: 10s
     metrics_path: /metrics
     scheme: http
     static_configs:
     - targets: ['localhost:6060']
EOF

cat > $HOME/prometheus.yml << EOF
global:
  scrape_interval:     15s # By default, scrape targets every 15 seconds.

  # Attach these labels to any time series or alerts when communicating with
  # external systems (federation, remote storage, Alertmanager).
  external_labels:
    monitor: 'codelab-monitor'

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
   - job_name: 'node_exporter'
     static_configs:
       - targets: ['localhost:9100']
   - job_name: 'nodes'
     metrics_path: /metrics    
     static_configs:
       - targets: ['localhost:8008']
   - job_name: 'erigon'
     scrape_interval: 10s
     scrape_timeout: 3s
     metrics_path: /debug/metrics/prometheus
     scheme: http
     static_configs:
       - targets: ['localhost:6060']
EOF

Move it to /etc/prometheus/prometheus.yml

sudo mv $HOME/prometheus.yml /etc/prometheus/prometheus.yml

Update file permissions.

sudo chmod 644 /etc/prometheus/prometheus.yml

Finally, restart the services.

sudo systemctl restart grafana-server prometheus prometheus-node-exporter

Import your new Teku dashboard into Grafana.

Open http://localhost:3000 or http://<your validator's ip address>:3000 in your local browser.
Login with your credentials
Download and save Teku's Dashboard json file.
Click Create + icon > Import
Add the ETH2 client dashboard via Upload JSON file

Credits: https://grafana.com/grafana/dashboards/13457

🌊 7. Clean up Prysm Storage

After a period of stable attestations on Teku, you can safely dispose of the former Prysm files and reclaim disk space.

# executables
rm -rf ~/prysm 

# Validator Keys
rm -rf ~/.eth2validators

# Beacon Chain Data
rm -rf ~/.eth2

# executables
sudo rm /usr/local/bin/validator
sudo rm /usr/local/bin/beacon-chain

# datadir
sudo rm -rf /var/lib/prysm

# users
sudo deluser prysmbeacon
sudo deluser prysmvalidator

Well done on successfully switching! Cheers to client diversity and a healthy beacon chain.

🙏 Alternative Community Migration Guides

Why You Should Switch to LighthouseLighthouse

Migrate from another client - The Nimbus Beacon Chain Book

https://www.reddit.com/r/ethstaker/comments/pu30fa/short_guide_to_migrate_from_prysm_to_teku_or/