Start Teku
You can run Teku as a beacon node and validator in a single process, or as separate processes.
We recommend you run the beacon node and validator as a single process if they are to run on the same machine.
By default, Teku connects to mainnet
. Use the --network
command line option to specify an alternate network.
Prerequisites
Start the clients in a single process
Start the beacon node and validator as a single process by specifying the validator options using the teku
command.
teku \
--ee-endpoint=http://localhost:8551 \
--ee-jwt-secret-file=jwtsecret.hex \
--metrics-enabled=true \
--rest-api-enabled=true \
--checkpoint-sync-url=https://beaconstate.ethstaker.cc \
--validators-proposer-default-fee-recipient=0xFE3B557E8Fb62b89F4916B721be55cEb828dBd73 \
--validator-keys=validator/keys/validator_888eef.json:validator/passwords/validator_888eef.txt
Use the --validator-keys
option to specify the directories or files to load the encrypted keystore file(s) and associated password file(s) from.
Run the clients separately
Validators must connect to a beacon node to publish attestations or propose blocks. The beacon node requires internet access, but the connected validators can run on machines without internet access.
Start the beacon node
Run Teku as a beacon node.
teku \
--ee-endpoint=http://localhost:8551 \
--ee-jwt-secret-file=jwtsecret.hex \
--metrics-enabled=true \
--rest-api-enabled=true \
--checkpoint-sync-url=https://beaconstate.ethstaker.cc \
Specify --rest-api-enabled
to allow validators to connect to the beacon node.
Don't pass the validator keys as a command line option to both the beacon node and validator client. This can cause a slashable offense.
By default, validator clients can connect to the beacon node at http://127.0.0.1:5051
.
Use the --rest-api-interface
and --rest-api-port
options to update the address.
You can specify --rest-api-host-allowlist
to allow access to the REST API from specific hostnames.
Start the validator
To run a validator, connect to a running beacon node.
Use the validator-client
or
vc
subcommand to run a Teku as a validator.
teku validator-client \
--beacon-node-api-endpoint=http://192.10.10.101:5051,http://192.140.110.44:5051 \
--validator-keys=validator/keys:validator/passwords
Ensure that the validator keys are only provided to the validator. Don't pass the validator keys as command line options to both the beacon node and validator client. This can cause a slashable offense.
Specify one or more beacon nodes using the --beacon-node-api-endpoint
option.
You can supply multiple beacon node endpoints to the validator. The first one is used as the primary node, and others as failovers.
Confirm Teku is running
Use the /liveness
endpoint to check whether the node is up.
The endpoint returns the status 200 OK
if the node is up or syncing.
- curl HTTP request
- Result
curl -I -X GET "http://192.10.10.101:5051/teku/v1/admin/liveness"
HTTP/1.1 200 OK
Date: Fri, 05 Feb 2021 03:58:30 GMT
Server: Javalin
Content-Type: application/json
Cache-Control: max-age=0
Content-Length: 0
Exit codes
Exit code 1
This code indicates a scenario where Teku has exited with a fatal error; however, restarting Teku without changes is a logical step in correcting it. This code represents the error is related to something external to Teku.
Example: If the beacon chain controller has issues starting P2P services on a particular port, the system will exit with a fatal error code of 1. This could correct itself with a restart.
Example 2: If the migrate database command does not successfully migrate the database, Teku will exit with a code 1 fatal error. Restarting the system could correct this.
Exit code 2
This code indicates a scenario where Teku has exited with a fatal error. Restarting Teku will not correct this.
Note that most user configuration errors fall into this category.
Example: If you specify an invalid database version in the migrate database command, Teku will exit with a code 2 fatal error. To correct this, make the appropriate changes to the setup and then restart Teku.
Usages
The exit codes for Teku are important to understand so that you know how to approach a restart and mitigate the issue.
The service unit configuration file is an example of how status codes can be used.
RestartPreventExitStatus=
and RestartForceExitStatus=
can be used to automatically restart Teku or stop it in case of failure.
The following is an example of a systemd.service
file:
RestartForceExitStatus=1
RestartPreventExitStatus=2