Execution client
Pre-release software
The Nimbus execution client is currently available as a proof of concept - all aspects of it, including resource requirements, command line interface and in particular the database format will change!
Have fun with it and let us know how it goes while keeping the above in mind.
If you're looking for information about setting up an execution client for validator duties or any other production usage, see the execution clients guide.
The Nimbus execution client is a light-weight implementation of the Ethereum execution protocol. Paired with a beacon node or light client, it provides access to Ethereum blockchain for dapps and users alike via the standard Web3 API.
Building from source
The Nimbus execution client is currently only provided as a source code distribution.
Clone the nimbus-eth1 repository
Run the build process
To build the Nimbus execution client and its dependencies, make sure you have all prerequisites and then run:
This may take a few minutes.
When the process finishes, the nimbus_execution_client executables can be found in the build subdirectory.
Syncing using era files
Syncing Nimbus requires a set of era1 and era files. These can be generated from a geth and nimbus consensus client respectively or downloaded from a third-party repository.
In addition to the era files themselves, you will need at least 200GB of free space on a fast SSD in your data directory, as set by the --data-dir command line option.
era file downloading
era and era1 files for testing purposes could at the time of writing be found here - these sources may or may not be available:
- https://mainnet.era.nimbus.team/
- https://mainnet.era1.nimbus.team/
- https://hoodi.era.nimbus.team/
The Hoodi network does not have era1 files since it never operated as a proof-of-work chain
- https://holesky.era.nimbus.team/
The Holesky network does not have era1 files since it never operated as a proof-of-work chain
- https://sepolia.era.nimbus.team/
- https://sepolia.era1.nimbus.team/
A wider community maintained list of era and era1 files can be found eth-clients github history-endpoints
Downloading these files can take a long time, specially if you are downloading sequentially.
For easier and fast download, please use the era_downloader.sh script provided in the nimbus-eth1 repository.
You'll need:
aria2installed:- macOS:
brew install aria2 - Ubuntu/Debian:
sudo apt install aria2
- macOS:
- Standard Unix tools:
bash,awk,find,grep,curl
It is recommended that you place the era files in the data directory under era1 and era respectively. Era files can be shared between multiple nodes and can reside on a slow drive - use the --era1-dir and --era-dir options if they are located outside of the data directory.
See the era file guide for more information.
Future versions of Nimbus will support other methods of syncing, such as snap sync.
Performing a full sync of mainnet from era files takes several days - its speed varies greatly depending on hardware. Use one of the testnets to get started more quickly!
Launch the client
In order for the execution client to operate, you need to connect a consensus node. This can be the Nimbus beacon node, a supported consensus client or a light client.
The consensus node connects to the execution client via the Engine API which is enabled using --engine-api and by default runs on port 8551.
During startup, a jwt.hex file will be placed in the data directory containing authentication information that the consensus node uses to connect - make sure to use the same jwt.hex file on both consensus and execution node.
Optionally quickstart with a pre-synced database
Unverified pre-synced database
The pre-synced database is provided by the Nimbus team which contained the state, but using this database is trusting the team to have provided a valid database. This gives you a headstart on syncing, but if you don't trust the provider, you should do a full sync instead, either from era files or from the p2p network. The pre-synced database is not available for all networks, and is only available for mainnet
If you want to skip the era file import and start with a pre-synced database, you can download a pre-synced database from the Nimbus team. This database is for now only available for the mainnet.
# Download the pre-synced database
wget https://eth1-db.nimbus.team/mainnet-static-vid-keyed.tar.gz
# Extract the database into the data directory
tar -xzf mainnet-static-vid-keyed.tar.gz
This will extract the pre-synced database into the current directory, which you can then use as your data directory.
Using the consensus node to sync
While era files cover the majority of chain history. In most cases, Nimbus will automatically sync recent blocks via peer-to-peer networking. However, if your node is stuck, has no peers, or you're on a weak network connection, you can optionally use nrpc to sync recent blocks directly from a connected consensus node using the Engine API.
This method of syncing loads blocks from the consensus node and passes them to the execution client via the Engine API.
Future versions of Nimbus will support snap sync.