Without intention locks, if one transaction acquires a row-level exclusive lock, and another transaction wants to acquire a table-level shared lock, the DBMS wouldn’t know whether that table-level lock would conflict — unless it scanned every row. This is inefficient.

🔧 Intention Locks Solve This:

They allow hierarchical locking while preserving compatibility and efficiency.

🧠 Example: How Intention Locks Work

Let’s say we have a table Employees, and a transaction wants to:

🔹 Transaction T1:

• Acquire exclusive lock on a row (e.g., row 5).

It will first:

• Place an IX lock on the table Employees.
• Then place an X lock on row 5.

🔹 Transaction T2:

• Wants a shared lock on the entire table (to read all rows safely).

It requests:

• An S lock on the table.

❌ Conflict Detected:

• T2’s S lock is not compatible with T1’s existing IX lock, so T2 must wait.

This avoids:

• T2 thinking it's safe to read the table while T1 is modifying a row.
• Dirty reads or inconsistent views of the data.