Running an Amaru Node on Preprod (Fast-Forward)

This is the fast path to a synced Amaru node: install the binary, bootstrap from a snapshot, and start it on preprod.

1. Hardware requirements

Network	CPU Cores	Free RAM	Free storage
Mainnet	2	4GB	30GB
Preprod	2	2GB	10GB

Note

The Amaru process should not exceed 1 GB of RSS memory under standard conditions. Allocating 4GB of RAM provides sufficient headroom for the process to handle the load associated with mainnet stress.

2. Installing Amaru

version reference

Amaru releases follow a calendar-based scheme (e.g. 10.11.20260827). Always check the releases page for the latest version before installing. Amaru is still in an exploratory phase, so expect frequent releases and occasional breaking changes.

Pick whichever method fits your environment:

Release Binaries — pre-compiled executables

Statically-linked executables for Linux, macOS, and Windows, with shell completion scripts, are attached to every release and nightly build. Download the archive for your platform and put the amaru binary on your $PATH.

Debian / RPM package — create amaru user and systemd service

Also installs a dedicated system user and a systemd service, it’s the recommended path for a production deployment; see Installing Amaru in the mainnet guide.

Debian/Ubuntu:

VERSION=10.11.20260827 ARCH=x86_64  # check releases page for latest; also available for aarch64
curl -fsSL -o amaru-$VERSION-linux-$ARCH.deb "https://github.com/pragma-org/amaru/releases/download/v$VERSION/amaru-$VERSION-linux-$ARCH.deb"
sudo apt install ./amaru-$VERSION-linux-$ARCH.deb

Fedora/RHEL/CentOS:

VERSION=10.11.20260827 ARCH=x86_64  # check releases page for latest; also available for aarch64
curl -fsSL -o amaru-$VERSION-linux-$ARCH.rpm "https://github.com/pragma-org/amaru/releases/download/v$VERSION/amaru-$VERSION-linux-$ARCH.rpm"
sudo dnf install ./amaru-$VERSION-linux-$ARCH.rpm

Docker image

docker pull ghcr.io/pragma-org/amaru:latest

Multi-arch (amd64/arm64), published on every change to main (:latest, a nightly build) and for tagged releases. See the package page for all tags.

Homebrew (macOS & Linux)

brew tap pragma-org/amaru https://github.com/pragma-org/amaru && brew trust --tap pragma-org/amaru
brew install amaru

Nix (macOS & Linux)

nix profile install --no-write-lock-file github:pragma-org/amaru#amaru

Build from source

Install a Rust toolchain (edition 2024, Rust 1.97 or newer, see rustup), then from a clone of the amaru repository:

cargo build --release

The binary is produced at target/release/amaru. Amaru also provides a Docker-based build and run path, see docker/README.md in the repository.

3. Bootstrapping the node

Create a directory for your chosen network :

mkdir -p $HOME/amaru/preprod/
cd $HOME/amaru/preprod

Amaru bootstraps itself with Mithril-derived snapshots :

amaru node bootstrap --network preprod

This downloads a window of pre-generated Mithril-derived ledger snapshots from Amaru’s snapshot bucket and imports them directly into the chain and ledger databases. amaru node run then only needs to sync the blocks produced since then.

$HOME/amaru/preprod/
├── chain.preprod.db/
├── ledger.preprod.db/
└── snapshots/

Note

Bootstrap fails if the target directories (chain or ledger) already contain data, remove them first before re-bootstrapping: amaru node rm --wipe-all-dbs --network preprod or rm -rf chain.preprod.db/ ledger.preprod.db/

4. Running the node

Note

Amaru ships with the genesis files and protocol parameters for mainnet, preprod, and preview compiled into the binary, there is nothing to download. Just pass --network when starting the node.

For a custom network (testnet_<magic>), pass a JSON era history file with --era-history and override the global parameters (security parameter, epoch length, etc.) with the corresponding flags, see Configuring the node in the mainnet guide, or run amaru node run --help-global-parameters for the full list.

amaru node run --network preprod

Flag	Description
--network	Network to run on: mainnet, preprod, preview, or a custom testnet_<magic>
--db-chain	Directory for the chain database (defaults to ./chain.<network>.db)
--db-ledger	Directory for the ledger database (defaults to ./ledger.<network>.db)
--peers-listen-on	Address to accept incoming peer connections on (default 0.0.0.0:3000)
--peer	Upstream peer(s) to sync from; repeatable or comma-separated
--peers-max-upstream	Maximum number of upstream (outbound) peers Amaru maintains at once (default 3)
--submit-api-listen-on	If set, exposes an HTTP POST /api/submit/tx endpoint for transaction submission

For the full list of options run amaru node run --help.

Unless --tui-off is passed (or stdout isn’t a terminal), amaru node run opens an embedded terminal dashboard showing sync progress, peers, and logs live, that’s the quickest way to confirm the node is catching up to the tip. See The Amaru TUI for a full tour of what it shows.

Caution

As with cardano-node, do not treat the node as caught up, or submit transactions, until the tip shown matches the current chain tip on a public explorer.

For OpenTelemetry metrics/logs/traces, Monitoring dashboards, submitting transactions, upgrading, and other operational details, see Advanced Installation in the mainnet guide.