A fairly useless toolkit for measuring gas costs in CosmWasm smart contracts. The true intent of this project is to provide a guided walkthrough of creating and deploying a non-standard cw contract, complete with automated deployment scripts, testing, and analysis. This goes a little outside the realm of the standard mintable/burnable token, or NFT collection deployment to encourage thinking outside of the box, with some portion of the project being completely off-chain.

The project measures gas consumption across different message sizes, formats, and character types. Results are stored in CSV format locally and on-chain.

Example results are available in this analysis report.

• Rust (beginner)
• JavaScript
• Cosmos SDK basics

graph TD
    A[Build Contract] --> B[Deploy Contract]
    B --> C[Run Gas Tests]
    C --> D[Store Results On-Chain]
    D --> E[Analyze Results]
    E --> F[Generate Visualizations]

• Gas testing for message sizes, formats, and character types
• On-chain storage of test results
• Statistical regression analysis
• Transaction hash storage as proof
• Automated deployment to testing workflow
• Multi-chain configuration options
• Interactive chart generation

cw-gas-test/
├── src/                 # Rust contract code
│   └── lib.rs           # Gas testing contract
├── artifacts/           # Compiled contract (after build)
├── scripts/             # JavaScript automation
│   ├── config.js        # Chain configuration
│   ├── deploy.js        # Contract deployment
│   ├── test-gas.js      # Gas testing
│   ├── analyze-results.js # Analysis + Report Generation
│   ├── package.json     # JS dependencies
│   ├── deployments.json # Tracks deployment addresses per chain
│   └── .env             # Mnemonic ONLY (create from template)
├── Cargo.toml           # Rust dependencies
├── rust-toolchain.toml  # Rust version spec
├── build.sh             # Compilation script (build contract only)
├── analyze_results.py   # [Optional] Python data visualization
└── cw-gas-test.sh       # Complete workflow (build + deploy + test + analyze)

1. Rust (1.74.0+)
2. Node.js (v18.0.0+)
3. Yarn (or npm)
4. Docker (for contract optimization)
5. Fee tokens on the target network
6. jq (for JSON parsing in scripts)
7. Python (optional, for extra visualization)

git clone https://github.com/Cordtus/cw-gas-test.git
cd cw-gas-test

# Make the scripts executable
chmod +x build.sh cw-gas-test.sh

// =============================
// CHAIN CONFIGURATION
// =============================
// Edit for your target chain
export const config = {
  // Network settings
  RPC_ENDPOINT: 'http://localhost:26657',
  REST_ENDPOINT: 'http://localhost:1317',
  CHAIN_ID: 'gaia-1',
  ADDRESS_PREFIX: 'gaia',

  // Token settings
  TOKEN_NAME: 'STAKE',      // Display name / ticker / symbol (e.g. 'ATOM', 'EVMOS')
  TOKEN_DENOM: 'ustake',    // Base denomination with prefix (e.g. 'uatom', 'aevmos')
  GAS_PRICE: '0.025ustake',
  GAS_ADJUSTMENT: 1.3,      // Buffer to avoid tx failure

  // Contract settings
  CONTRACT_LABEL: 'gas_test_contract',
  WASM_PATH: '../artifacts/cw_gas_test.wasm',
  
  // Optional: Existing contract address if you do NOT want to deploy a new one on target chain.
  // If your target chain is found in deployments.json, the contract address specified there will be used.
  // In any other case, this script will build & deploy a new contract.
  CONTRACT_ADDRESS: '',     // Optional: reuse existing contract [check ./deployments.json]

  // Test configuration
  TEST_MESSAGE_LENGTHS: [1, 10, 50, 100, 200, 500, 1000, 2000],
  OUTPUT_FILE: 'gas_results.csv',
  REQUEST_DELAY: 1000,      // Delay between requests in milliseconds
  TX_CONFIRMATION_TIMEOUT: 10000, // Max time to wait for tx confirmation (ms)
  TX_POLLING_INTERVAL: 3000,      // How often to check for tx confirmation (ms)
};

cd scripts
cp .env.template .env
# then edit .env to include your mnemonic

./build.sh

# From inside "scripts/"
yarn install         # or npm install
yarn deploy          # deploy.js
yarn test            # test-gas.js
yarn analyze         # analyze-results.js

./cw-gas-test.sh

Results in gas_results.csv:

Regression analysis output:

## Regression Analysis
- Base gas cost: 129256.99 gas units
- Marginal cost per byte: 41.37 gas units
- R-squared: 1.0000

## Formula
Total Gas = 129256.99 + 41.37 × Message Size (bytes)
Total Cost = Total Gas × 0.02 ustake/gas unit

Key metrics:

1. Base Gas Cost: Fixed transaction cost (y-intercept)
2. Marginal Gas Cost: Cost per additional byte (slope)
3. Format Efficiency: Comparison between data formats

The formula Gas = 129256.99 + 41.37 × MessageSize indicates a base cost of ~129,257 gas plus ~41.37 gas per byte.

The smart contract includes:

1. Message Storage:

   • StoreMessage: Store any message with its actual length
   • StoreFixedLength: Store a message padded/truncated to a specific length

2. Test Run Data:

   • RecordTestRun: Save aggregated test data with transaction proofs
   • ClearData: Remove old test data (admin only)

3. Queries:

   • GetConfig: Contract configuration
   • GetMessage: Retrieve stored message by ID
   • ListMessages: List stored messages (paginated)
   • GetTestRuns: Retrieve test run statistics (paginated)
   • GetGasSummary: Get gas usage analysis summary

4. Error Handling:

   • Custom error types for better error handling
   • Input validation with meaningful error messages
   • Proper authorization checks

You can edit parameters in scripts/config.js (like TEST_MESSAGE_LENGTHS, network endpoints) to suit your environment:

// New parameter options
MAX_PARALLEL_REQUESTS: 3,          // Control parallel execution
GENERATE_VISUALIZATION: true,      // Toggle visualization generation
SMALL_MESSAGE_THRESHOLD: 200,      // Threshold for small/large message
RETRY_ATTEMPTS: 2,                 // Number of retries on failure
MAX_MESSAGE_SIZE: 10000,           // Max msg size

The test suite examines:

1. Message Length: Test how gas scales with increasing message size
2. Message Format: Compare JSON, Base64, and Hex encoded data
3. Character Type: Test different characters (ASCII, Unicode, Emoji)

To modify test parameters or add new test types:

1. Edit test-gas.js to add new test functions
2. Update the corresponding contract message handlers in lib.rs
3. Modify analyze-results.js to include new metrics in the analysis

If you already have a contract deployed:

1. Put its address into config.js under CONTRACT_ADDRESS, or

2. Place the address in scripts/deployments.json for the target chain: Example Formatting:

   {"deployments": [{"chainId":"contractAddress"}]}

3. Please submit a pull request to include your contribution!

If your target chain exists in deployments.json, the scripts will skip re-deployment and run tests directly.

Potential enhancements:

• Batch transaction benchmarking
• Complex storage pattern testing
• Web-based visualization dashboard
• Cross-chain comparison framework
• CosmWasm memory operation benchmarks

This tool can be adapted for other CosmWasm-enabled chains by updating config.js.

For different CosmWasm versions, update Cargo.toml dependencies to match the target chain and recompile with ./build.sh.

Directory of CosmWasm enabled networks, courtesy of Confio. Refer to the Cosmos Chain Registry for chain parameters and other information.

• CosmWasm: Smart contracting platform for Cosmos ecosystem
• Gas: Computational cost metric in blockchains
• WASM: WebAssembly, compiled format for CosmWasm contracts
• Mnemonic: Seed phrase for cryptographic key generation
• RPC Endpoint: Interface for blockchain node interaction
• Base Gas: Fixed transaction cost component
• Marginal Gas: Variable cost per data unit
• Regression Analysis: Statistical relationship modeling method

MIT License – see LICENSE file

• CosmWasm Docs
• Cosmos SDK Docs
• Rust Book
• CosmWasm Examples