Effective Deadlock Handling in SQL Transactions

Handling Deadlocks Gracefully

Write a SQL query to handle deadlocks using error handling techniques.

Solution:

BEGIN TRY
    BEGIN TRANSACTION;

    -- Update conflicting rows.
    UPDATE TableA SET Column1 = 'Value1' WHERE ID = 1;
    WAITFOR DELAY '00:00:05';
    UPDATE TableB SET Column2 = 'Value2' WHERE ID = 2;

    COMMIT TRANSACTION;
END TRY
BEGIN CATCH
    -- Rollback and log the deadlock error.
    ROLLBACK TRANSACTION;
    PRINT 'Deadlock detected: ' + ERROR_MESSAGE();
END CATCH;

Explanation:

• Purpose of the Query :
• The goal is to handle deadlocks gracefully using TRY...CATCH.
• Key Components :
• ROLLBACK TRANSACTION: Ensures that the transaction is rolled back in case of a deadlock.
• ERROR_MESSAGE(): Captures the deadlock error message.
• Why Handle Deadlocks? :
• Deadlocks can cause transactions to fail, so handling them improves system robustness.
• Logging deadlocks helps identify and resolve contention issues.
• Real-World Application :
• In financial systems, deadlock handling ensures consistent account updates.

Additional Notes:

• Use SET DEADLOCK_PRIORITY to control which transaction is chosen as the deadlock victim.
• Analyze deadlock graphs to understand the root cause.
• Important Considerations:
• Avoid long-running transactions to reduce deadlock likelihood.

For more Practice: Solve these Related Problems:

• Write a SQL query to simulate a deadlock scenario between two transactions and handle it using TRY...CATCH.
• Write a SQL query to prioritize a specific transaction as the deadlock victim using SET DEADLOCK_PRIORITY.
• Write a SQL query to log deadlock details, including the resources involved and the affected queries.
• Write a SQL query to analyze deadlock graphs and identify the root cause of contention.

Go to:

• SQL Error Handling and Debugging Exercises Home ↩
• SQL Exercises Home ↩

PREV : Identifying Bottlenecks with Query Profiling.
NEXT : Debugging Recursive Queries.

Have another way to solve this solution? Contribute your code (and comments) through Disqus.