Historical Data Replay

Replay historical blockchain data from TimescaleDB for analytics, backtesting, and data recovery.

Historical Transaction Replay

Use start_slot and end_slot parameters to replay transactions from a specific slot range.

Request

// Replay transactions from slot 100000 to 100100
const stream = client.SubscribeTransactions({
  start_slot: 100000,
  end_slot: 100100,
  filter_votes: true,
  commitment: 'FINALIZED'
}, metadata);

let count = 0;
stream.on('data', (tx) => {
  count++;
  console.log(`[${count}] Slot ${tx.slot}: ${tx.signature}`);
});

stream.on('end', () => {
  console.log(`Replay complete: ${count} transactions`);
});

# Replay transactions from specific slot range
request = geyser_pb2.SubscribeTransactionsRequest(
    start_slot=100000,
    end_slot=100100,
    filter_votes=True,
    commitment=geyser_pb2.FINALIZED
)

count = 0
for tx in stub.SubscribeTransactions(request, metadata=metadata):
    count += 1
    print(f"[{count}] Slot {tx.slot}: {tx.signature}")

print(f"Replay complete: {count} transactions")

Historical Block Replay

Replay blocks from a slot range:

const stream = client.SubscribeBlocks({
  start_slot: 100000,
  end_slot: 100010,
  commitment: 'FINALIZED'
}, metadata);

stream.on('data', (block) => {
  console.log(`Block ${block.slot}: ${block.transaction_count} transactions`);
});

Use Cases

Backfill Missing Data

async function backfillData(startSlot, endSlot) {
  return new Promise((resolve) => {
    const transactions = [];

    const stream = client.SubscribeTransactions({
      start_slot: startSlot,
      end_slot: endSlot,
      filter_votes: true
    }, metadata);

    stream.on('data', (tx) => {
      transactions.push(tx);
    });

    stream.on('end', () => {
      resolve(transactions);
    });
  });
}

const data = await backfillData(100000, 101000);
console.log(`Backfilled ${data.length} transactions`);

Analytics & Backtesting

def analyze_slot_range(start, end):
    """Analyze transactions in a slot range"""
    request = geyser_pb2.SubscribeTransactionsRequest(
        start_slot=start,
        end_slot=end,
        account_include=['JUP6LkbZbjS1jKKwapdHNy74zcZ3tLUZoi5QNyVTaV4'],
        filter_votes=True
    )

    swaps = []
    for tx in stub.SubscribeTransactions(request, metadata=metadata):
        swaps.append(tx)

    return analyze_swaps(swaps)

Performance Tips

Use batching for large slot ranges
Set appropriate filter_votes to reduce data volume
Use FINALIZED commitment for historical data
Historical queries are batched at 1000 transactions per batch

See gRPC Streaming Overview for more information.

Historical Data Replay

On this page