Skip to content

Setting up a Celestia full storage Node

This tutorial will guide you through setting up a Celestia full storage node, which is a celestia-node that doesn't connect to celestia-app (hence not a full consensus node), but stores all the data.

Overview of full storage nodes

Full storage nodes are Celestia nodes that store all the data. Full storage nodes send block shares, headers, and fraud proofs to light nodes. The light nodes gossip headers, fraud proofs, and sometimes block shares, between one another.

Full storage node

Hardware requirements

The following hardware minimum requirements are recommended for running the full storage node:

  • Memory: 32 GB RAM
  • CPU: 6 cores
  • Disk: 10 TB SSD Storage
  • Bandwidth: 1 Gbps for Download/1 Gbps for Upload

Setting up your full storage node

The following tutorial is done on an Ubuntu Linux 20.04 (LTS) x64 instance machine.

Setup the dependencies

You can follow the tutorial for setting up your dependencies

Install celestia-node

Note: Make sure that you have at least 250+ Gb of free space for Celestia full storage node

You can follow the tutorial for installing celestia-node

Run the full storage node

Initialize the full storage node

Run the following command:

sh
celestia full init --p2p.network arabica
celestia full init --p2p.network arabica
sh
celestia full init --p2p.network mocha
celestia full init --p2p.network mocha

Start the full storage node

Start the full storage node with a connection to a validator node's gRPC endpoint (which is usually exposed on port 9090):

In order for access to the ability to get/submit state-related information, such as the ability to submit PayForBlob transactions, or query for the node's account balance, a gRPC endpoint of a validator (core) node must be passed as directed below.

Refer to the ports section of the celestia-node troubleshooting page for information on which ports are required to be open on your machine.

sh
celestia full start --core.ip <ip-address>:<port>
celestia full start --core.ip <ip-address>:<port>

Using an RPC of your own, or one from the list on the Mocha testnet page or list on the Arabica devnet page, start your node.

You can create your key for your node by following the cel-key instructions

Once you start the full storage node, a wallet key will be generated for you. You will need to fund that address with testnet tokens to pay for PayForBlob transactions. You can find the address by running the following command:

sh
./cel-key list --node.type full --keyring-backend test --p2p.network <network>
./cel-key list --node.type full --keyring-backend test --p2p.network <network>

You can get testnet tokens from:

NOTE: If you are running a full-storage node for your sovereign rollup, it is highly recommended to request Arabica devnet tokens as Arabica has the latest changes that can be used to test for developing your sovereign rollup. You can still use Mocha testnet as well, it is just mostly used for validator operations.

Optional: run the full storage node with a custom key

In order to run a full storage node using a custom key:

  1. The custom key must exist inside the celestia full storage node directory at the correct path (default: ~/.celestia-full/keys/keyring-test)
  2. The name of the custom key must be passed upon start, like so:
sh
celestia full start --core.ip <ip-address> \
  --keyring.accname <name-of-custom-key> \
  --p2p.network arabica
celestia full start --core.ip <ip-address> \
  --keyring.accname <name-of-custom-key> \
  --p2p.network arabica
sh
celestia full start --core.ip <ip-address> \
  --keyring.accname <name-of-custom-key> \
  --p2p.network mocha
celestia full start --core.ip <ip-address> \
  --keyring.accname <name-of-custom-key> \
  --p2p.network mocha

Optional: start the full storage node with SystemD

If you would like to run the full storage node as a background process, follow the SystemD tutorial.

With that, you are now running a Celestia full storage node.

Stop the full storage node

In order to gracefully stop the full storage node, use Control + C in the terminal window where the node is running. Be sure to only do this once as the shutdown will not be instantaneous.