blink

Fast contract bytecode indexer for Ethereum. Extracts contract creation data to Parquet files.

Installation

cargo install --path .

Or build from source:

cargo build --release
./target/release/blink --help

Usage

# Basic usage (uses ETH_RPC_URL environment variable)
blink contracts --start-block 0 --end-block 1000000

# With explicit RPC URL
blink contracts --rpc https://eth.llamarpc.com --start-block 19000000

# Custom concurrency settings for faster extraction
blink contracts --start-block 0 \
    --batch-size 100 \
    --max-concurrent-requests 64 \
    --max-concurrent-chunks 8

# Alchemy-friendly throttled settings
blink contracts --start-block 16800000 --end-block 17799999 \
    --batch-size 5 \
    --max-concurrent-requests 4 \
    --max-concurrent-chunks 1 \
    --max-retries 20 \
    --initial-backoff-ms 2000 \
    --max-backoff-ms 10000

# Resume from where you left off (skips existing chunks)
blink contracts --start-block 0 --end-block 20000000

# Force overwrite existing chunks
blink contracts --start-block 0 --overwrite

Options

blink contracts [OPTIONS] --start-block <START_BLOCK>

Options:
      --rpc <RPC>                          RPC URL (fallback: ETH_RPC_URL)
      --start-block <START_BLOCK>          Start block (inclusive)
      --end-block <END_BLOCK>              End block or "latest" [default: latest]
      --chunk-size <CHUNK_SIZE>            Blocks per output file [default: 100000]
      --batch-size <BATCH_SIZE>            Blocks per JSON-RPC request [default: 50]
      --max-concurrent-requests <N>        Concurrent HTTP requests [default: 32]
      --max-concurrent-chunks <N>          Concurrent chunk processing [default: 4]
      --output-dir <OUTPUT_DIR>            Output directory [default: ./data/blink]
      --overwrite                          Overwrite existing chunk files
      --max-retries <MAX_RETRIES>          Max retries per batch [default: 5]
      --initial-backoff-ms <MS>            Initial retry backoff [default: 1000]
      --max-backoff-ms <MS>                Max retry backoff [default: 30000]
      --fast                               Use aggressive defaults

Output Schema

Each parquet file contains contract creations with this schema:

Column	Type	Description
block_number	uint32	Block where contract was created
block_hash	binary	Block hash
create_index	uint32	Index of creation within block
transaction_hash	binary	Transaction hash (nullable for genesis)
contract_address	binary	Deployed contract address
deployer	binary	EOA that initiated the transaction
factory	binary	Address that called CREATE/CREATE2
init_code	binary	Constructor bytecode
code	binary	Deployed runtime bytecode
init_code_hash	binary	Keccak256 of init_code
n_init_code_bytes	uint32	Length of init_code
n_code_bytes	uint32	Length of deployed code
code_hash	binary	Keccak256 of deployed code
chain_id	uint64	Chain ID

Performance

blink is designed for speed:

• Batch requests: Groups multiple trace_block calls into single HTTP requests
• Parallel chunks: Processes multiple 100k block ranges simultaneously
• Semaphore-controlled concurrency: Prevents RPC overload while maximizing throughput

Typical performance with a good RPC endpoint:

• ~500-2000 blocks/sec depending on RPC and block density

On rate-limited RPCs (e.g., Alchemy), lower concurrency and batch sizes to avoid 429s. Backoff is exponential with a cap (see --max-backoff-ms).

Reports

Each run writes a report JSON to ./data/blink/.blink/reports/ with status, error (if any), and per-chunk metadata including output file paths and sizes.

Requirements

• Ethereum node with trace_block support (Erigon, Reth, or tracing-enabled Geth)
• Rust 1.70+