Guide
Required setup instructions
The local setup is only intended for testing or development purposes.
This is intended to show how you can run a validator on your local machine, which is only for testing and development purposes. This also means that signatures from local validators are only able to be accessed by locally ran relayers.
When running a validator locally, your validator will write its signatures to a local directory. This directory can be named whatever you like, just remember to use this directory when setting your Configuration.
# Pick an informative name specific to the chain you're validating
MY_VALIDATOR_SIGNATURES_DIRECTORY=/tmp/hyperlane-validator-signatures-ethereum
# Create the directory
mkdir $MY_VALIDATOR_SIGNATURES_DIRECTORY
For running a validator in a production environment, see the AWS setup guide.
Take a look at the Agent configuration page and the Configuration reference for a full list of configuration possibilities. The list below is not complete, however it should be enough to get started.
Your validator takes arguments and environment variables as configuration. See below for both the general configurations and those that are specific to the setup instructions you followed.
Argument | Description |
|---|---|
--reorgPeriod | The number of block confirmations a validator should wait before signing the Mailbox merkle root. Note that signing a root that is later invalidated (i.e. due to a re-org) is considered fraudulent behavior and will eventually be slashable.
See Latencies |
--originChainName | The name of the chain being validated (e.g. ethereum) |
--chains.[origin chain name].connection.url | The RPC URL of the node for the chain you are validating. Note Polygon mainnet requires an RPC URL of an archive node (see An RPC nodefor details).
e.g. --chains.ethereum.connection.url or --chains.polygon.connection.url |
--db | The path to where the validator should write persistent data to disk. Ensure this path to be persistent when using cloud setups. When using Docker, make sure to mount the persistent path/volume into the container.
See Configuration referencefor more info |
Environment variable | Description |
|---|---|
CONFIG_FILES | If you want to add additional configuration files you can add additional paths here as a comma separated list. These files must be accessible within the filesystem your agent has access to. If you're running in Docker, see Config files with Docker for tips on mounting your config files into your Docker container. |
Local setup
Production Setup (AWS)
Argument | Description |
|---|---|
--validator.key | Your validator's private key, which is used to sign merkle roots. |
--chains.${localchainname}.signer.key | Your validator's private key, which will be used to submit a transaction on chain that publicly announce your validator's checkpoint syncer. |
--checkpointSyncer.type | Set to localStorage. |
--checkpointSyncer.path | The path to your local directory where validator signatures will be written. This should be the value of $MY_VALIDATOR_SIGNATURES_DIRECTORY from the Local Setup.Example: --checkpointSyncer.path='/tmp/hyperlane-validator-signatures-ethereum' |
Note that relayers must be configured with
--allowLocalCheckpointSyncers
to be able to read signatures from this validator.Argument | Description |
|---|---|
--validator.type | Set to the aws literal. |
--validator.id | The alias of your validator's AWS KMS key, prefixed with alias/.
Example: alias/hyperlane-validator-signer-${chain_name} |
--chains.${localchainname}.signer.type | Set to the aws literal. |
--chains.${localchainname}.signer.id | The alias of your validator's AWS KMS key, prefixed with alias/.
Example: alias/hyperlane-validator-signer-${chain_name} |
--validator.region | The region of your AWS KMS key.
Example: us-east-1. |
--checkpointSyncer.type | Set to s3. |
--checkpointSyncer.bucket | The AWS S3 bucket name. |
--checkpointSyncer.region | The region of your AWS S3 bucket.
Example: us-east-1. |
Environment variable | Description |
|---|---|
AWS_ACCESS_KEY_ID | The access key ID of your validator's AWS IAM user. |
AWS_SECRET_ACCESS_KEY | The secret access key of your validator's AWS IAM user. |
The recommended installation method for a production environment is using a Docker image.
Docker
Source
We encourage folks to validate on as many chains as they are interested in supporting. We recommend that resources are not shared between validator instances.
validator.env
# These are example values to roughly illustrate
# what a .env file should look like
HYP_BASE_ORIGINCHAINNAME=ethereum
HYP_BASE_REORGPERIOD=20
# ...
# ...
To run the validator binary with the environment variables specified in
validator.env:Using Docker
Without Docker
Using the
--env-file flag, we can pass in the environment variables to the validator:docker run -it --env-file validator.env $DOCKER_IMAGE ./validator
If you're specifying a path to your own config file in the
CONFIG_FILES environment variable, check out the Config files with Docker section.If you're running with a Local Setup validator on the same machine, which requires a locally ran relayer to be able to access these validator signatures, be sure to mount your local validator's signature directory on your host machine into your Docker container.
For example, if your local validator is writing signatures to
/tmp/hyperlane-validator-signatures-ethereum, you should mount a directory for the Docker container. This is the same directory set in the $HYP_BASE_CHECKPOINTSYNCER_PATH environment variable. If the command below fails with
docker: invalid reference format, the whitespaces may have been malformed and you should remove them from the command.docker run \
-it \
--env-file validator.env \
--mount type=bind,source="$(pwd)"/hyperlane-validator-signatures-ethereum,target=/tmp/hyperlane-validator-signatures-ethereum \
# you can pass multiple `--mount` flags to mount several directories
$DOCKER_IMAGE \
./validator
Using
env and xargs, we can run the built binary from within the hyperlane-monorepo/rust directory with the environment variables found in validator.env:env $(cat validator.env | grep -v "#" | xargs) ./target/release/validator
If everything is configured correctly, you should see json files being written to your S3 bucket (if you followed the AWS setup) or to your local signatures directory (if you followed the Local Setup). New json files get written every time a new outbound message is inserted into the Mailbox.
Relayers need to know where to find your validator's signatures. Your validator will automatically attempt to announce itself by writing to the
ValidatorAnnounce contract on the chain that you're validating. To do this, your validator must have a small amount of tokens to pay for the gas for this transaction.
If your validator has not yet announced itself, and does not have enough tokens to pay for gas, it will log a message specifying how many tokens are needed.
Last modified 17d ago