> ## Documentation Index
> Fetch the complete documentation index at: https://www.helius.dev/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Transaction Monitoring with Yellowstone gRPC

> Stream Solana transactions in real-time with program filtering, execution details, and token balance change tracking.

<Note>
  This guide documents the open **Yellowstone gRPC** protocol. For the LaserStream-flavored version with the `helius-laserstream` SDK, see [LaserStream → Transaction Monitoring](/laserstream/guides/transaction-monitoring).
</Note>

## Overview

Transaction monitoring enables you to track transaction execution, success/failure status, program interactions, and token balance changes across Solana in real-time. This guide covers filtering strategies and practical implementations for different transaction monitoring use cases.

<Info>
  **Prerequisites:** This guide assumes you've completed the [Yellowstone gRPC Quickstart](/grpc/quickstart) and have a working stream setup.
</Info>

## Transaction Filtering Options

<Tabs>
  <Tab title="Program Filtering">
    **Monitor transactions involving specific programs**

    Track all transactions that interact with your programs:

    ```typescript theme={"system"}
    const subscribeRequest: SubscribeRequest = {
      transactions: {
        client: {
          accountInclude: [
            "TokenkegQfeZyiNwAJbNbGKPFXCWuBvf9Ss623VQ5DA", // Token Program
            "11111111111111111111111111111111",              // System Program
            "675kPX9MHTjS2zt1qfr1NYHuzeLXfQM9H24wFSUt1Mp8"  // Your program
          ],
          accountExclude: [],
          accountRequired: [],
          vote: false,
          failed: false
        }
      },
      commitment: CommitmentLevel.CONFIRMED
    };
    ```

    <Note>
      **Best for:** Program-specific monitoring, DeFi protocol tracking, smart contract interactions
    </Note>
  </Tab>

  <Tab title="Account-Specific">
    **Monitor transactions affecting specific accounts**

    Track transactions that modify specific accounts:

    ```typescript theme={"system"}
    const subscribeRequest: SubscribeRequest = {
      transactions: {
        client: {
          accountInclude: [
            "EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v", // USDC mint
            "YourWalletAddress"                                // Your wallet
          ],
          vote: false,
          failed: true // Include failures to track errors
        }
      },
      commitment: CommitmentLevel.CONFIRMED
    };
    ```

    <Tip>
      **Use case:** Wallet monitoring, token mint tracking, specific account activity
    </Tip>
  </Tab>

  <Tab title="Advanced Filtering">
    **Combine multiple filter criteria**

    Use required accounts and exclusions for precise filtering:

    ```typescript theme={"system"}
    const subscribeRequest: SubscribeRequest = {
      transactions: {
        client: {
          accountInclude: ["TokenkegQfeZyiNwAJbNbGKPFXCWuBvf9Ss623VQ5DA"],
          accountRequired: ["YourProgramId"], // Must include this program
          accountExclude: ["VoteProgram"],     // Exclude vote-related txs
          vote: false,
          failed: false
        }
      },
      commitment: CommitmentLevel.CONFIRMED
    };
    ```

    <Warning>
      **Filter logic:** accountInclude (OR) AND accountRequired (AND) AND NOT accountExclude
    </Warning>
  </Tab>
</Tabs>

## Practical Examples

### Example 1: Monitor DEX Transactions

Track all transactions involving popular DEX programs:

```typescript theme={"system"}
import { StreamManager } from './stream-manager'; // From quickstart guide

async function monitorDEXTransactions() {
  const streamManager = new StreamManager(
    "your-grpc-endpoint",
    "YOUR_API_KEY",
    handleDEXTransaction
  );

  const subscribeRequest: SubscribeRequest = {
    transactions: {
      client: {
        accountInclude: [
          "9WzDXwBbmkg8ZTbNMqUxvQRAyrZzDsGYdLVL9zYtAWWM", // Raydium V4
          "675kPX9MHTjS2zt1qfr1NYHuzeLXfQM9H24wFSUt1Mp8", // Raydium V5
          "CAMMCzo5YL8w4VFF8KVHrK22GGUQpMpTFb6xRmpLFGNnSm", // Raydium CLMM
          "JUP6LkbZbjS1jKKwapdHNy74zcZ3tLUZoi5QNyVTaV4"   // Jupiter
        ],
        vote: false,
        failed: false
      }
    },
    commitment: CommitmentLevel.CONFIRMED
  };

  await streamManager.connect(subscribeRequest);
}

function handleDEXTransaction(data: any): void {
  if (data.transaction?.transaction) {
    const tx = data.transaction.transaction;
    console.log(`\n🔄 DEX Transaction:`);
    console.log(`  Signature: ${tx.signature}`);
    console.log(`  Slot: ${data.transaction.slot}`);
    console.log(`  Status: ${tx.meta?.err ? 'Failed' : 'Success'}`);
    console.log(`  Fee: ${tx.meta?.fee || 0} lamports`);
    console.log(`  Compute Units: ${tx.meta?.computeUnitsConsumed || 0}`);
    
    // Show token balance changes
    if (tx.meta?.preTokenBalances?.length > 0) {
      console.log(`  Token Balance Changes:`);
      tx.meta.preTokenBalances.forEach((preBalance: any, index: number) => {
        const postBalance = tx.meta.postTokenBalances[index];
        if (preBalance && postBalance) {
          const change = postBalance.uiTokenAmount.uiAmount - preBalance.uiTokenAmount.uiAmount;
          if (change !== 0) {
            console.log(`    ${preBalance.mint}: ${change > 0 ? '+' : ''}${change}`);
          }
        }
      });
    }
  }
}
```

### Example 2: Monitor Failed Transactions

Track failed transactions to identify issues:

```typescript theme={"system"}
async function monitorFailedTransactions() {
  const streamManager = new StreamManager(
    "your-grpc-endpoint",
    "YOUR_API_KEY",
    handleFailedTransaction
  );

  const subscribeRequest: SubscribeRequest = {
    transactions: {
      client: {
        accountInclude: ["YourProgramId"], // Your program
        vote: false,
        failed: true // Only failed transactions
      }
    },
    commitment: CommitmentLevel.CONFIRMED
  };

  await streamManager.connect(subscribeRequest);
}

function handleFailedTransaction(data: any): void {
  if (data.transaction?.transaction?.meta?.err) {
    const tx = data.transaction.transaction;
    console.log(`\n❌ Failed Transaction:`);
    console.log(`  Signature: ${tx.signature}`);
    console.log(`  Slot: ${data.transaction.slot}`);
    console.log(`  Error: ${JSON.stringify(tx.meta.err)}`);
    console.log(`  Fee: ${tx.meta.fee} lamports`);
    console.log(`  Compute Units: ${tx.meta.computeUnitsConsumed || 0}`);
    
    // Log instruction details for debugging
    if (tx.transaction?.message?.instructions) {
      console.log(`  Instructions:`);
      tx.transaction.message.instructions.forEach((inst: any, i: number) => {
        console.log(`    ${i}: Program ${inst.programIdIndex}, Data: ${inst.data}`);
      });
    }
  }
}
```

### Example 3: Monitor High-Value Transactions

Track transactions with significant SOL transfers:

```typescript theme={"system"}
async function monitorHighValueTransactions() {
  const streamManager = new StreamManager(
    "your-grpc-endpoint",
    "YOUR_API_KEY",
    handleHighValueTransaction
  );

  const subscribeRequest: SubscribeRequest = {
    transactions: {
      client: {
        accountInclude: ["11111111111111111111111111111111"], // System Program
        vote: false,
        failed: false
      }
    },
    commitment: CommitmentLevel.CONFIRMED
  };

  await streamManager.connect(subscribeRequest);
}

function handleHighValueTransaction(data: any): void {
  if (data.transaction?.transaction?.meta) {
    const tx = data.transaction.transaction;
    const preBalances = tx.meta.preBalances || [];
    const postBalances = tx.meta.postBalances || [];
    
    // Calculate largest balance change
    let maxChange = 0;
    preBalances.forEach((preBalance: number, index: number) => {
      const postBalance = postBalances[index] || 0;
      const change = Math.abs(postBalance - preBalance);
      maxChange = Math.max(maxChange, change);
    });
    
    // Only log transactions with > 10 SOL moved
    const changeInSOL = maxChange / 1e9;
    if (changeInSOL > 10) {
      console.log(`\n💰 High-Value Transaction:`);
      console.log(`  Signature: ${tx.signature}`);
      console.log(`  Slot: ${data.transaction.slot}`);
      console.log(`  Max SOL Transfer: ${changeInSOL.toFixed(2)} SOL`);
      console.log(`  Fee: ${tx.meta.fee / 1e9} SOL`);
      console.log(`  Accounts: ${tx.transaction?.message?.accountKeys?.length || 0}`);
    }
  }
}
```

## Transaction Data Structure

Understanding the transaction data structure helps extract relevant information:

<Accordion title="Transaction Message Structure">
  **Core transaction data:**

  ```typescript theme={"system"}
  {
    signature: string;
    isVote: boolean;
    transaction: {
      message: {
        accountKeys: string[];        // All accounts involved
        instructions: Instruction[];  // Program instructions
        recentBlockhash: string;     // Recent blockhash used
      };
      signatures: string[];          // Transaction signatures
    };
    meta: {
      err: any;                      // Error details if failed
      fee: number;                   // Transaction fee in lamports
      computeUnitsConsumed: number;  // Compute units used
      preBalances: number[];         // Account balances before
      postBalances: number[];        // Account balances after
      preTokenBalances: TokenBalance[];
      postTokenBalances: TokenBalance[];
      logMessages: string[];         // Program log messages
    };
  }
  ```
</Accordion>

<Accordion title="Token Balance Changes">
  **Token balance structure:**

  ```typescript theme={"system"}
  {
    accountIndex: number;
    mint: string;                   // Token mint address
    owner: string;                  // Account owner
    uiTokenAmount: {
      amount: string;               // Raw token amount
      decimals: number;             // Token decimals
      uiAmount: number;             // Human-readable amount
      uiAmountString: string;       // String representation
    };
  }
  ```
</Accordion>

<Accordion title="Instruction Details">
  **Program instruction structure:**

  ```typescript theme={"system"}
  {
    programIdIndex: number;         // Index in accountKeys array
    accounts: number[];             // Account indices involved
    data: string;                   // Instruction data (base58)
  }
  ```
</Accordion>

## Filter Logic Reference

<CardGroup cols={2}>
  <Card title="Include Logic (OR)" icon="plus">
    **accountInclude:** Transaction must involve ANY of these accounts

    **Example:** `["A", "B"]` matches transactions involving account A OR account B
  </Card>

  <Card title="Required Logic (AND)" icon="check">
    **accountRequired:** Transaction must involve ALL of these accounts

    **Example:** `["A", "B"]` matches transactions involving account A AND account B
  </Card>

  <Card title="Exclude Logic (NOT)" icon="minus">
    **accountExclude:** Transaction must NOT involve any of these accounts

    **Example:** `["A", "B"]` excludes transactions involving account A or account B
  </Card>

  <Card title="Combined Logic" icon="code">
    **Final filter:** `(accountInclude OR empty) AND (accountRequired AND all) AND NOT (accountExclude OR any)`
  </Card>
</CardGroup>

## Performance Considerations

<Tabs>
  <Tab title="Volume Management">
    **Transaction streams can be high-volume**

    * Start with specific program filters
    * Use commitment levels appropriately
    * Monitor your processing capacity
    * Implement backpressure handling

    ```typescript theme={"system"}
    // Rate limiting example
    let transactionCount = 0;
    const startTime = Date.now();

    function handleTransaction(data: any): void {
      transactionCount++;
      
      if (transactionCount % 100 === 0) {
        const elapsed = (Date.now() - startTime) / 1000;
        const rate = transactionCount / elapsed;
        console.log(`Processing ${rate.toFixed(1)} tx/sec`);
      }
      
      // Your transaction processing logic
    }
    ```
  </Tab>

  <Tab title="Data Processing">
    **Optimize data extraction**

    * Process data asynchronously
    * Extract only needed fields
    * Use efficient data structures
    * Consider batching updates

    ```typescript theme={"system"}
    // Efficient data extraction
    function extractTransactionData(tx: any) {
      return {
        signature: tx.signature,
        slot: tx.slot,
        success: !tx.meta?.err,
        fee: tx.meta?.fee || 0,
        computeUnits: tx.meta?.computeUnitsConsumed || 0,
        tokenChanges: extractTokenChanges(tx.meta)
      };
    }
    ```
  </Tab>
</Tabs>

## Error Handling

Common transaction monitoring errors and solutions:

<Accordion title="Too Many Transactions">
  **Error:** Overwhelming transaction volume

  **Solutions:**

  * Add more specific filters (accountRequired, accountExclude)
  * Use higher commitment levels to reduce volume
  * Implement sampling or rate limiting
  * Process transactions asynchronously
</Accordion>

<Accordion title="Missing Transactions">
  **Error:** Expected transactions not appearing

  **Solutions:**

  * Verify program addresses are correct
  * Check if transactions actually occur
  * Try PROCESSED commitment for faster updates
  * Ensure filters aren't too restrictive
</Accordion>

<Accordion title="Parse Errors">
  **Error:** Cannot parse transaction data

  **Solutions:**

  * Handle missing fields gracefully
  * Validate data structure before processing
  * Log problematic transactions for debugging
  * Use try-catch blocks around parsing logic
</Accordion>

## Next Steps

<CardGroup cols={2}>
  <Card title="Slot & Block Monitoring" icon="cube" href="/grpc/slot-and-block-monitoring">
    Learn to monitor network consensus and block production
  </Card>

  <Card title="Advanced Patterns" icon="chart-line" href="/grpc/stream-pump-amm-data">
    Real-world example: monitoring Pump AMM data
  </Card>
</CardGroup>