datahaven/README.md
Facundo Farall d2bf185bcc
feat: 🚀 Add deploy command to CLI (#87)
### New Features
1. Add the `deploy` command to our CLI.
1. Conditionally deploys kurtosis eth network if we're in `stagenet`
environment.
    2. Deploys DH nodes.
3. Deploys contracts (all of them). In `mainnet` and `testnet` it
shouldn't deploy EL contracts, but for now that's not implemented.
4. Configures parameters, validators and relayers in the same way as
`launch`.
5. Currently, it only deploys `beefy` and `beacon` relayers, `execution`
and `solochain` relayers are pending for a subsequent PR.
2. Add `waitFor` utility function that receives a lambda.

### Refactors
1. Several common functionalities used both by the `launch` and `deploy`
command have been moved to the `cli/handlers/common` directory, from
where both commands use them. These include
    1. Checks for installed dependencies.
    2. Common constants.
    3. The `LaunchedNetwork` class has been moved to this directory.
    4. DataHaven nodes common functions.
    5. Kurtosis common functions.
    6. Relayer common functions.
7. Kubernetes functions (although only used by `deploy`, it seemed
fitting to have it here still).
8. Remove CLI questions and separator prints from `deploy-contracts.ts`
and `set-datahaven-parameters.ts` scripts. These things should be in the
`cli/launch` folder, which consumes the functions in these scripts.
9. Remove `setParametersFromCollection` from `utils` folder and put it
in `cli`.
10. Create base snowbridge relayer configs for `local` and `stagenet` as
two separate directories.

### Fixes
1. Sets the default time of the `deploy` command to 6s as Lodestar is
slower than Lighthouse.
2. In `runShellCommandWithLogger` only print `stderr` if the command
fails.

### Additional Minor Changes
1. K8s secret key names changed from `dh-beefy-relay-eth-key` to
`dh-beefy-relay-ethereum-key` and `dh-beacon-relay-sub-key` to
`dh-beacon-relay-substrate-key`, for simplicity in the deployment
script.
11. Update suggested configs for `.vscode` configs.

---------

Co-authored-by: Steve Degosserie <723552+stiiifff@users.noreply.github.com>
2025-06-12 10:24:03 +02:00

6.5 KiB

DataHaven 🫎

An EVM compatible Substrate chain, powered by StorageHub and secured by EigenLayer.

Repo Structure

datahaven/
├── .github/ # GitHub Actions workflows.
├── contracts/ # Implementation of the DataHaven AVS (Autonomous Verifiable Service) smart contracts to interact with EigenLayer.
├── operator/ # DataHaven node based on Substrate. The "Operator" in EigenLayer terms.
├── test/ # Integration tests for the AVS and Operator.
├── resources/ # Miscellaneous resources for the DataHaven project.
└── README.md

E2E CLI

This repo comes with a CLI for launching a local DataHaven network, packaged with:

  1. A full Ethereum network with:
    • 2 x Execution Layer clients (e.g., reth)
    • 2 x Consensus Layer clients (e.g., lodestar)
    • Blockscout Explorer services for EL (if enabled with --blockscout)
    • Dora Explorer service for CL
    • Contracts deployed and configured for the DataHaven network.
  2. A DataHaven solochain.
  3. Snowbridge relayers for cross-chain communication.

To launch the network, follow the instructions in the test README.

Docker

This repo publishes images to DockerHub.

Tip

If you cannot see this repo you must be added to the permission list for the private repo.

To aid with speed it employs the following:

  • sccache: De-facto caching tool to speed up rust builds.
  • cargo chef: A method of caching building the dependencies as a docker layer to cut down compilation times.
  • buildx cache mounts: Using buildx's new feature to mount an externally restored cache into a container.
  • cache dance: Weird workaround (endorsed by docker themselves) to inject caches into containers and return the result back to the CI.

To run a docker image locally (moonsonglabs/datahaven:local), from the /test folder run:

bun build:docker:operator

Working with IDEs

VS Code (and its forks)

IDE configurations are ignored from this repo's version control, to allow for personalisation. However, there are a few key configurations that we suggest for a better experience. Here are the key suggested configurations to add to your .vscode/settings.json file:

Rust

{
  "rust-analyzer.linkedProjects": ["./operator/Cargo.toml"],
  "rust-analyzer.cargo.allTargets": true,
  "rust-analyzer.procMacro.enable": false,
  "rust-analyzer.server.extraEnv": {
    "CARGO_TARGET_DIR": "target/.rust-analyzer",
    "SKIP_WASM_BUILD": 1
  },
  "rust-analyzer.diagnostics.disabled": ["unresolved-macro-call"],
  "rust-analyzer.cargo.buildScripts.enable": false
}

These settings optimise Rust Analyzer for the DataHaven codebase:

  • Marks the operator/ folder as a linked project for analysis. The root of this repo is a workspace, and this is the rust project that should be analysed by rust-analyzer.
  • Disables proc macros and build scripts to improve performance. Otherwise, Substrate's proc macros will make iterative checks from rust-analyzer unbearably slow.
  • Sets a dedicated target directory for Rust Analyzer to avoid conflicts with other build targets like release builds.
  • Disables WASM builds during analysis for faster feedback.

Solidity

For Juan Blanco's Solidity Extension, add the following to your .vscode/settings.json file:

{
  "solidity.formatter": "forge",
  "solidity.compileUsingRemoteVersion": "v0.8.28+commit.7893614a",
  "[solidity]": {
    "editor.defaultFormatter": "JuanBlanco.solidity"
  }
}

These settings configure Solidity support:

  • Uses Forge as the formatter for consistency with the project's tooling.
  • Sets a specific Solidity version for compilation. This one should match the version used in foundry.toml.
  • Sets the Solidity extension as the default formatter.

Typescript

This repo uses Biome for TypeScript linting and formatting. To make the extension work nicely with this repo, add the following to your .vscode/settings.json file:

{
  "biome.lsp.bin": "test/node_modules/.bin/biome",
  "[typescript]": {
    "editor.defaultFormatter": "biomejs.biome",
    "editor.codeActionsOnSave": {
      "source.organizeImports.biome": "always"
    }
  }
}
  • Sets the Biome binary to the one in the test/ folder.
  • Sets Biome as the default formatter for TypeScript.
  • Sets Biome to always organise imports on save.

CI

Using the act binary, you can run GitHub Actions locally.

For example, to run the entire e2e workflow:

act -W .github/workflows/e2e.yml -s GITHUB_TOKEN="$(gh auth token)"

Which results in:

INFO[0000] Using docker host 'unix:///var/run/docker.sock', and daemon socket 'unix:///var/run/docker.sock'
INFO[0000] Start server on http://192.168.1.97:34567
[E2E - Kurtosis Deploy and Verify/kurtosis] ⭐ Run Set up job
[E2E - Kurtosis Deploy and Verify/kurtosis] 🚀  Start image=catthehacker/ubuntu:rust-24.04
[E2E - Kurtosis Deploy and Verify/kurtosis]   🐳  docker pull image=catthehacker/ubuntu:rust-24.04 platform= username= forcePull=true
[E2E - Kurtosis Deploy and Verify/kurtosis] using DockerAuthConfig authentication for docker pull
[E2E - Kurtosis Deploy and Verify/kurtosis]   🐳  docker create image=catthehacker/ubuntu:rust-24.04 platform= entrypoint=["tail" "-f" "/dev/null"] cmd=[] network="host"
[E2E - Kurtosis Deploy and Verify/kurtosis]   🐳  docker run image=catthehacker/ubuntu:rust-24.04 platform= entrypoint=["tail" "-f" "/dev/null"] cmd=[] network="host"
[E2E - Kurtosis Deploy and Verify/kurtosis]   🐳  docker exec cmd=[node --no-warnings -e console.log(process.execPath)] user= workdir=
[E2E - Kurtosis Deploy and Verify/kurtosis]   ✅  Success - Set up job
[E2E - Kurtosis Deploy and Verify/kurtosis]   ☁  git clone 'https://github.com/oven-sh/setup-bun' # ref=v2
...
[E2E - Kurtosis Deploy and Verify/kurtosis]   ✅  Success - Post Install Foundry [212.864597ms]
[E2E - Kurtosis Deploy and Verify/kurtosis] ⭐ Run Complete job
[E2E - Kurtosis Deploy and Verify/kurtosis] Cleaning up container for job kurtosis
[E2E - Kurtosis Deploy and Verify/kurtosis]   ✅  Success - Complete job
[E2E - Kurtosis Deploy and Verify/kurtosis] 🏁  Job succeeded