fuel-core Documentation

Rust full node implementation of the Fuel v2 protocol.

56,915 stars Rust Rust 1 file ยท ~3,869 tokens #blockchain#fuel
RAW Doc

Docker & Self-Hosted Deployment Verified OCI

Deploy fuel-core on any server, homelab, or cloud container platform.

1-Click Terminal Command
bash (docker run)
docker run -d --name fuel-core -p 4000:4000 --restart always fuel-core
Configured Services & Ports
โš™๏ธ fuel-core
image: fuel-core
๐Ÿ”Œ Ports: 4000:4000
docker-compose.yml
yaml (docker-compose.yml)
version: "3.9"

services:
  fuel-core:
    build: 
      context: . 
      dockerfile: deployment/Dockerfile
    image: fuel-core
    ports:
      - "4000:4000"
    volumes:
      - db:/mnt/db
volumes:
  db:

File: README.md

Fuel Client






Fuel client implementation.

Versions currently used in networks

Network Version
Fuel Ignition 0.48.1
Testnet 0.48.1
Devnet 0.48.1

Contributing

If you are interested in contributing to Fuel, see our CONTRIBUTING.md guidelines for coding standards and review process.

Before pushing any changes or creating pull request please run source ci_checks.sh.

Building

If you plan to use already pre-compiled binaries you can directly go to Running a Ignition node.

System Requirements

There are several system requirements including clang.

MacOS

bash
brew update
brew install cmake

Debian

bash
apt update
apt install -y cmake pkg-config build-essential git clang libclang-dev

Arch

bash
pacman -Syu --needed --noconfirm cmake gcc pkgconf git clang

Rust setup

You'll need wasm32-unknown-unknown target installed.

bash
rustup target add wasm32-unknown-unknown

Running a Ignition node

If you want to participate in the Ignition network with your own node you can launch it following these simple commands.

From pre-compiled binaries

Follow : https://docs.fuel.network/docs/node-operator/fuel-ignition/mainnet-node/

From source

Clone the fuel-core repository :

text
git clone https://github.com/FuelLabs/fuel-core.git

Go to the latest release tag for ignition on the fuel-core repository :

text
git checkout v0.45.1

Build your node binary:

bash
make build

To run the node follow : https://docs.fuel.network/docs/node-operator/fuel-ignition/mainnet-node/

Running a Local network from source

Clone the fuel-core repository :

text
git clone https://github.com/FuelLabs/fuel-core.git

Build your node binary:

bash
make build

To run the node follow : https://docs.fuel.network/docs/node-operator/fuel-ignition/local-node/

Setup to make contributions

Compiling

We recommend using xtask to build fuel-core for development:

sh
cargo xtask build

This will run cargo build as well as any other custom build processes we have such as re-generating a GraphQL schema for the client.

Testing

The ci_checks.sh script file can be used to run all CI checks,
including the running of tests.

shell
source ci_checks.sh

The script requires pre-installed tools. For more information run:

shell
cat ci_checks.sh

Running

The service can be launched by executing fuel-core run. The list of options for running can be accessed via the help option:

console
$ ./target/debug/fuel-core run --help

USAGE:
    fuel-core run [OPTIONS]

OPTIONS:
        --snapshot <SNAPSHOT>
          Snapshot from which to do (re)genesis. Defaults to local testnet configuration

          [env: SNAPSHOT=]
        ...

For many development purposes it is useful to have a state that won't persist and the db-type option can be set to in-memory as in the following example.

Example

console
$ ./target/debug/fuel-core run --db-type in-memory
2023-06-13T12:45:22.860536Z  INFO fuel_core::cli::run: 230: Block production mode: Instant
2023-06-13T12:38:47.059783Z  INFO fuel_core::cli::run: 310: Fuel Core version v0.18.1
2023-06-13T12:38:47.078969Z  INFO new{name=fuel-core}:_commit_result{block_id=b1807ca9f2eec7e459b866ecf69b68679fc6b205a9a85c16bd4943d1bfc6fb2a height=0 tx_status=[]}: fuel_core_importer::importer: 231: Committed block
2023-06-13T12:38:47.097777Z  INFO new{name=fuel-core}: fuel_core::graphql_api::service: 208: Binding GraphQL provider to 127.0.0.1:4000

To disable block production on your local node, set --poa-instant=false

Example

console
$ ./target/debug/fuel-core run --poa-instant=false
2023-06-13T12:44:12.857763Z  INFO fuel_core::cli::run: 232: Block production disabled

Troubleshooting

Publishing

We use publish-crates action for automatic publishing of all crates.

If you have problems with publishing, you can troubleshoot it locally with act.

shell
act release -s GITHUB_TOKEN=<YOUR_GITHUB_TOKEN> -j publish-crates-check --container-architecture linux/amd64 --reuse

It requires GitHubToken to do request to the GitHub. You can create it
with this instruction.

Outdated database

If you encounter an error such as

console
thread 'main' panicked at 'unable to open database: DatabaseError(Error { message: "Invalid argument: Column families not opened: column-11, column-10, column-9, column-8, column-7, column-6, column-5, column-4, column-3, column-2, column-1, column-0" })', fuel-core/src/main.rs:23:66

Clear your local database using: rm -rf ~/.fuel/db

File descriptor limits

On some macOS versions the default file descriptor limit is quite low, which can lead to IO errors with messages like Too many open files or even fatal runtime error: Rust cannot catch foreign exceptions when RocksDB encounters these issues. Use the following command to increase the open file limit. Note that this only affects the current shell session, so consider adding it to ~/.zshrc.

bash
ulimit -n 10240

Log level

The service relies on the environment variable RUST_LOG. For more information, check the EnvFilter examples crate.

Human logging can be disabled with the environment variable HUMAN_LOGGING=false

Debugging

See the guide on debugging for an overview on running a debug build of a local node.

Docker & Kubernetes

sh
# Create Docker Image
docker build -t fuel-core . -f deployment/Dockerfile

# Delete Docker Image
docker image rm fuel-core

# Create Kubernetes Volume, Deployment & Service
kubectl create -f deployment/fuel-core.yml

# Delete Kubernetes Volume, Deployment & Service
kubectl delete -f deployment/fuel-core.yml

GraphQL service

The client functionality is available through a service endpoint that expect GraphQL queries.

Transaction executor

The transaction executor currently performs instant block production. Changes are persisted to RocksDB by default.

  • Service endpoint: /v1/graphql
  • Schema (available after building): crates/client/assets/schema.sdl

The service expects a mutation defined as submit that receives a Transaction in hex encoded binary format, as specified here.


File: benches/README.md

Fuel Core Benches

This directory contains a set of benchmarks for Fuel Core. The benchmarks are
used for calculating gas metering costs for the Fuel Core VM.

Running the benchmarks

To run the benchmarks, you can use cargo bench -p fuel-core-benches command available.
Alternatively you can use cargo criterion -p fuel-core-benches if you have it installed.
For more information on using criterion see the guide.

Profiling a benchmark

Sometimes it is useful to produce a flamegraph from a benchmark to verify
you are measuring the correct things.
To do this you can use the cargo flamegraph command.

Benchmarking a specific function

First build the benchmark with debug symbols:
CARGO_PROFILE_RELEASE_DEBUG=true cargo-criterion -p fuel-core-benches --no-run
Then you need to find the actual test binary that was built:
ls target/release/deps/vm* -lat | head -1
Should give you something like:
target/release/deps/vm-a17190f2ca5e7169
Then you can run the benchmark with the flamegraph command (runs the bench for 10 seconds):
flamegraph target/release/deps/vm-a17190f2ca5e7169 --bench bench_name --profile-time 10
For example if you want to profile the swwq op you would do (it's regex):
flamegraph target/release/deps/vm-a17190f2ca5e7169 --bench '^swwq/swwq$' --profile-time 10
Or for a dependent bench you would do:
flamegraph target/release/deps/vm-a17190f2ca5e7169 --bench '^srwq/100$' --profile-time 10

Using `collect` to generate bench data

Firstly you should run the benchmarks and collect the output:
cargo criterion -p fuel-core-benches --message-format json --bench vm > bench.json
You could just pipe the output into the collect binary but benches are slow to run
so I suggest saving the output like above.

Then you can run the collect binary to generate the data you want:
cargo run -p fuel-core-benches --bin collect --release -- --input bench.json
This will generate a gas-costs.yaml file in the current directory like:

yaml
burn: 35
call:
  base: 311
  dep_per_unit: 14

There are multiple output options. See collect --help for more information.

Generating the `fuel-vm` `default-gas-costs.rs`

Do the same as above except run the following collect command.
cargo run -p fuel-core-benches --bin collect --release -- --input bench.json -f rust --output default-gas-costs.rs
This will generate a default-gas-costs.rs file in the current directory like:

rust

This file can then replace the one in fuel-vm/src/gas/default-gas-costs.rs.


File: bin/fuel-core/chainspec/local-testnet/README.md

The configuration of the local network

Chain config

  • The ChainId is 0.
  • The initial privileged_address which can perform the network upgrade:
    shell
    {"address":"9f0e19d6c2a6283a3222426ab2630d35516b1799b503f37b02105bebe1b8a3e9","secret":"d80a243ef91956f626d1dad2f23bdfeb73fd0b363282b1eb2227ac5964144afb","type":"block_production"}
  • The public address of the authority node that produces blocks:
    shell
    {"address":"e0a9fcde1b73f545252e01b30b50819eb9547d07531fa3df0385c5695736634d","secret":"4dd0cdca64ef56a01fc81891f9beb6d898f19a22b2e287bce91d807fdf46589a","type":"block_production"}
  • The block gas limit is 30000000.

Gas costs

The gas costs was created from the benchmarks_fuel_core_0_35_0.json benchmark results.
The new_storage_per_byte is manually set to be 63.
The gtf is manually set to be 13.

State config

  • The coinbase address hard coded in the genesis contract:
    shell
    {"address":"7b4b30b2437b0073e5ba5a9324cf55831d180a89f66332b541827e12e647b751","secret":"9e24cfa071f6c1c4984a17ecf18061a8d0c9c304e7dd7703788bd122bd578650","type":"block_production"}
  • Contains many wallets with fake ETHs:
    • Wallet 1:
      shell
      {"address":"6b63804cfbf9856e68e5b6e7aef238dc8311ec55bec04df774003a2c96e0418e","secret":"de97d8624a438121b86a1956544bd72ed68cd69f2c99555b08b1e8c51ffd511c","type":"block_production"}
  • The base asset contract source code is taken from here.
    • The ContractId of the contract is 0x7e2becd64cd598da59b4d1064b711661898656c6b1f4918a787156b8965dc83c.
    • The derived(SubId is 0000000000000000000000000000000000000000000000000000000000000000) base AssetId from this contract is 0xf8f8b6283d7fa5b672b530cbb84fcccb4ff8dc40f8176ef4544ddb1f1952ad07.

File: bin/e2e-test-client/README.md

Fuel Core E2E Test Client

The test client provides a suite of idempotent tests intended to validate remote environments.

To customize the suite parameters for an existing node deployment, set the FUEL_CORE_E2E_CONFIG environment variable with a path to a valid configuration toml.

When FUEL_CORE_E2E_CONFIG is unset a default configuration is used which is suitable for local environment testing. The default configuration is:

toml
endpoint = "http://localhost:4000"
wallet_sync_timeout = "10s"
full_test = false
coinbase_contract_id = "7777777777777777777777777777777777777777777777777777777777777777"


[wallet_a]
secret = "de97d8624a438121b86a1956544bd72ed68cd69f2c99555b08b1e8c51ffd511c"

[wallet_b]
secret = "37fa81c84ccd547c30c176b118d5cb892bdb113e8e80141f266519422ef9eefd"
shell
FUEL_CORE_E2E_CONFIG="/Users/your_user_name/fuel/e2e.toml" cargo run -p fuel-core-e2e-client -- --nocapture

File: ci/README.md

CI Tooling

Cross-platform compilation is handled using cross.
This is because cross ensures binaries will run on as many platforms as possible.
It also works-around various bugs
in rust bind-gen that prevent us from simply using --target with cargo to cross-compile.

We use custom extensions of the docker images provided by cross
in order to support building rocksdb with clang. Each target requires its own dockerfile,
as they all have slightly different distros, package managers and compiler toolchains.

To test cross compilation of fuel core locally, build the provided dockerfiles
with the image tags expected by Cross.toml using the helper script:

shell
./build-images.sh

To cross-compile fuel core for a supported target, you can build it as you
would with cargo but using cross instead:

shell
cargo install cross
cross build --profile=release --target ${target} -p fuel-core

File: crates/trace/README.md

Fuel Core Trace

This crate provides a simple way to enable tracing for tests across the fuel repo.

Usage

To enable tracing for a test, add the following to the lib.rs file of the crate you want to trace:

rust
#[cfg(test)]
fuel_core_trace::enable_tracing!();

If you also want tracing in your integration tests, you need to add the above to each integration tests file.

Tracing is still disabled by default so to enable it set the following environment variable:

bash
export FUEL_TRACE=1

Or if you just want it for a single test run:

bash
FUEL_TRACE=1 cargo test

Now you will have error level tracing like:

text
2023-01-25T02:27:14.362856Z ERROR works: tracing: I'm visible if FUEL_TRACE=1 is set

You can use the RUST_LOG environment variable to control the level of tracing you want to see. For example:

bash
FUEL_TRACE=1 RUST_LOG=trace cargo test

Additional Subscribers

You can set a few different types of subscribers:

Compact output:

bash
FUEL_TRACE=compact cargo test

Pretty output:

bash
FUEL_TRACE=pretty cargo test

Log to file:

bash
FUEL_TRACE=log-file cargo test

Log to file and output to console:

bash
FUEL_TRACE=log-show cargo test

You can also set the log file path:

bash
FUEL_TRACE_PATH=/some/path FUEL_TRACE=log-file cargo test

If you don't set the path, it will default to CARGO_MANIFEST_DIR/logs/logfile.


File: crates/services/consensus_module/poa/README.md

Fuel Core PoA Manager

Coordinates block production in a PoA network. This is used as a stopgap until a PoS network is implemented, but also useful for testing.

For the PoS implementation, see fuel-core-bft,