datahaven/test
Steve Degosserie 1f0cd6de27
feat: Enhanced Helm deployment with ingress management and solochain support (#115)
This PR enhances the Helm-based deployment system with several key
improvements organized into the following areas:

## New Features

### Ingress Management
- **Ingress per replica**: Added `ingress-per-replica` chart template
that automatically generates an ingress for each node replica, exposing
individual pod instances
- **Traefik integration**: Added chart values to deploy Traefik as the
ingress controller for local K8s clusters, enabling proper ingress
testing (requires adding names to `/etc/hosts`)

### Solochain Support
- **New relay chart**: Added dedicated Helm chart for Solochain relay
deployment
- **CLI integration**: Updated DataHaven CLI to deploy both Execution
and Solochain relayers

## Configuration Improvements

### Environment Structure
- **Modular configs**: Refactored environmental configurations from
single `values.yaml` files into separate component-specific overrides
for better organization

### Node Configuration
- **Base config updates**: Improved base configurations for bootnode and
validator nodes
- **Network protocol**: Reverted from litep2p back to libp2p to resolve
node communication issues on Stagenet
- **Archive node routing**: In Stagenet, relayers now connect to the
bootnode (configured as archive node) instead of validator nodes

## Storage & Deployment

### Persistent Storage
- **Relayer database**: Added support for persistent volumes to store
relayer databases instead of using ephemeral storage

### Deployment Scripts
- **Documentation cleanup**: Removed obsolete test deploy.sh script and
updated the deployment README with clearer instructions

These changes provide a more robust, scalable, and maintainable
deployment system for DataHaven infrastructure.

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added support for deploying new relayer types: Solochain and Execution
relayers, with dedicated configuration and secret management.
* Introduced per-replica ingress configuration for node deployments,
allowing each replica to have its own ingress resource and hostname.
* Added persistent storage options for relay data, configurable via
storage path, class, and size.
* Added new deployment configuration files for local and stagenet
environments, including Traefik ingress controller setup.
* Introduced a new relay category, "Solochain Relayers," for standalone
chain operations and cross-chain communication.

* **Improvements**
* Updated deployment configurations to use container-specific
environment YAML files for more granular control.
* Enhanced relay and node configurations with new flags and backend
options, including dynamic peer ID generation and automatic bootnode
discovery.
* Updated relayer endpoints to consistently use the bootnode for
connections.
* Refined relay configuration files for improved structure, clarity, and
endpoint management.

* **Bug Fixes**
* Corrected deployment logic to reference the correct
environment-specific configuration files during Helm deployments.

* **Documentation**
* Simplified and updated deployment documentation to focus on CLI-based
deployment, removing outdated manual instructions and adding a concise
overview of components and relayer types.

* **Chores**
* Removed deprecated deployment scripts and outdated configuration files
to streamline the deployment process.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Claude <noreply@anthropic.com>
2025-08-11 14:30:01 +03:00
..
.papi test: init benchmark infra (#110) 2025-08-06 13:44:42 +02:00
cli feat: Enhanced Helm deployment with ingress management and solochain support (#115) 2025-08-11 14:30:01 +03:00
configs Fix: 🏗️ Message encoding / decoding (#113) 2025-07-16 07:38:58 +00:00
contract-bindings Fix: 🏗️ Message encoding / decoding (#113) 2025-07-16 07:38:58 +00:00
docker perf(CLI): Add option to use local Docker build in CLI for faster iteration (#77) 2025-05-16 18:04:40 -03:00
docs test: wait for event utils (#121) 2025-08-01 20:56:46 +02:00
framework test: 🏗️ Setup e2e testing framework (#104) 2025-07-16 18:51:07 +02:00
launcher feat: Enhanced Helm deployment with ingress management and solochain support (#115) 2025-08-11 14:30:01 +03:00
resources fix: 🐛 Use lodestar instead of lighthouse CL client (#91) 2025-06-09 12:29:31 -03:00
scripts test: 🏗️ Setup e2e testing framework (#104) 2025-07-16 18:51:07 +02:00
suites test: 🏗️ Setup e2e testing framework (#104) 2025-07-16 18:51:07 +02:00
utils test: wait for event utils (#121) 2025-08-01 20:56:46 +02:00
.gitignore test: ⚙️ Add Kurtosis Launch Script (#34) 2025-04-08 15:42:45 +00:00
biome.json fix: 🚨 Add error in TS for missing awaits (#81) 2025-05-19 22:28:43 +00:00
bun.lock Bum Polkadot api version to 1.15.1 (#125) 2025-07-31 09:40:58 +02:00
bun.lockb feat: 🏗️ run execution relayer (#73) 2025-06-05 15:00:03 +00:00
bunfig.toml test: 🧙 Generate Type Bindings for Contracts (#58) 2025-05-01 11:14:19 +01:00
package.json Bum Polkadot api version to 1.15.1 (#125) 2025-07-31 09:40:58 +02:00
README.md test: 🏗️ Setup e2e testing framework (#104) 2025-07-16 18:51:07 +02:00
tsconfig.json test: 🏗️ Setup e2e testing framework (#104) 2025-07-16 18:51:07 +02:00
wagmi.config.ts test: 🧙 Generate Type Bindings for Contracts (#58) 2025-05-01 11:14:19 +01:00

DataHaven E2E Testing

Quick start guide for running DataHaven end-to-end tests. For comprehensive documentation, see E2E Testing Guide.

Pre-requisites

  • Kurtosis: For launching test networks
  • Bun v1.2 or higher: TypeScript runtime and package manager
  • Docker: For container management
  • Foundry: To deploy contracts
  • Helm: The Kubernetes Package Manager
MacOS

Important

If you are running this on a Mac, zig is a pre-requisite for crossbuilding the node. Instructions for installation can be found here.

Quick Start

# Install dependencies
bun i

# Interactive CLI to launch a full local DataHaven network
bun cli launch

# Run all the e2e tests
bun test:e2e

# Run all the e2e tests with limited concurrency
bun test:e2e:parallel

# Run a specific test suite
bun test suites/some-test.test.ts

For more information on the E2E testing framework, see the E2E Testing Framework Overview.

Other Common Commands

Command Description
bun cli stop Stop all local DataHaven networks (interactive, will ask for confirmation on each component of the network)
bun cli deploy Deploy the DataHaven network to a remote Kubernetes cluster
bun generate:wagmi Generate contract TypeScript bindings for the contracts in the contracts directory
bun generate:types Generate Polkadot API types
bun generate:types:fast Generate Polkadot API types with the --fast-runtime feature enabled

Local Network Deployment

Follow these steps to set up and interact with your local network:

  1. Deploy a minimal test environment

    bun cli launch
    

    This script will:

    1. Check for required dependencies.
    2. Launch a DataHaven solochain.
    3. Start a Kurtosis network which includes:
      • 2 Ethereum Execution Layer clients (reth)
      • 2 Ethereum Consensus Layer clients (lodestar)
      • Blockscout Explorer services for EL (if enabled with --blockscout)
      • Dora Explorer service for CL
    4. Deploy DataHaven smart contracts to the Ethereum network. This can optionally include verification on Blockscout if the --verified flag is used (requires Blockscout to be enabled).
    5. Perform validator setup and funding operations.
    6. Set parameters in the DataHaven chain.
    7. Launch Snowbridge relayers.
    8. Perform validator set update.

    Note

    If you want to also have the contracts verified on Blockscout, you can pass the --verified flag to the bun cli launch command, along with the --blockscout flag. This will do all the previous, but also verify the contracts on Blockscout. However, note that this takes some time to complete.

  2. Explore the network

    • Block Explorer: http://127.0.0.1:3000.
    • Kurtosis Dashboard: Run kurtosis web to access. From it you can see all the services running in the network, as well as their ports, status and logs.

Troubleshooting

E2E Network Launch doesn't work

Script halts unexpectedly

When running bun cli launch the script appears to halt after the following:

## Setting up 1 EVM.

==========================

Chain 3151908

Estimated gas price: 2.75 gwei

Estimated total gas used for script: 71556274

Estimated amount required: 0.1967797535 ETH

==========================

This is due to how forge streams output to stdout, but is infact still deploying contracts to the chain. You should be able to see in blockscout the deploy script is indeed still working.

Errors with deploying forge scripts on kurtosis network

Try running forge clean to clear any spurious build artefacts, and running forge build again. Also try deploying manually to the still running kurtosis network.

Blockscout is empty

If you look at the browser console, if you see the following:

Content-Security-Policy: The page's settings blocked the loading of a resource (connect-src) at http://127.0.0.1:3000/node-api/proxy/api/v2/stats because it violates the following directive: "connect-src ' ...

this is a result of CORS and CSP errors due to running this as a local docker network.

Make sure you are connected directly to http://127.0.0.1:3000 (not localhost).

Alternatively, you can try installing a browser addon such as anti-CORS / anti-CSP to circumvent this problem.

Weird forge Errors

In the /contracts directory, you can try to run forge clean and forge build to see if it fixes the issue.

Linux: See if disabling ipV6 helps

I have found that ipV6 on Arch Linux does not play very nicely with Kurtosis networks. Disabling it completely fixed the issue for me.

macOS: Verify Docker networking settings

Docker Network Settings

If using Docker Desktop, make sure settings have permissive networking enabled.

Polkadot-API types don't match expected runtime types

If you've made changes to the runtime types, you need to re-generate the TS types for the Polkadot-API. Don't worry, this is fully automated.

From the ./test directory run the following command:

bun generate:types

This script will:

  1. Compile the runtime using cargo build --release in the ../operator directory.
  2. Re-generate the Polkadot-API types using the newly built WASM binary.

Note

The script uses the --release flag by default, meaning it uses the WASM binary from ./operator/target/release. If you need to use a different build target, you may need to adjust the script or run the steps manually.

Further Information

  • Kurtosis: Used for launching a full Ethereum network
  • Zombienet: Used for launching a Polkadot-SDK based network
  • Bun: TypeScript runtime and ecosystem tooling