CTE materialization behavior in PostgreSQL - Cheat Sheet & Quick Revision

Recall & Review

beginner

What does CTE stand for in SQL?

CTE stands for Common Table Expression. It is a temporary named result set that you can reference within a SELECT, INSERT, UPDATE, or DELETE statement.

intermediate

What is materialization in the context of CTEs in PostgreSQL?

Materialization means the CTE is executed once and its result is stored temporarily. Later, the stored result is reused wherever the CTE is referenced in the query.

intermediate

How does PostgreSQL treat CTEs by default before version 12?

Before PostgreSQL 12, CTEs are always materialized by default. This means the CTE query runs once and the result is saved for reuse.

advanced

What change did PostgreSQL 12 introduce regarding CTE materialization?

PostgreSQL 12 introduced the ability to inline CTEs (not materialize) by default when possible, improving performance by allowing the planner to optimize the query better.

advanced

How can you force materialization or inlining of a CTE in PostgreSQL 12 and later?

You can use the keywords MATERIALIZED or NOT MATERIALIZED after the CTE name to force materialization or inlining respectively. Example: WITH cte_name AS MATERIALIZED (...) or WITH cte_name AS NOT MATERIALIZED (...).

What happens when a CTE is materialized in PostgreSQL?

A The CTE query runs multiple times for each reference.

B The CTE query runs once and the result is stored temporarily.

C The CTE is ignored by the query planner.

D The CTE is converted into a permanent table.

Materialization means the CTE query runs once and its result is stored temporarily for reuse.

Before PostgreSQL 12, how were CTEs treated by default?

A Converted to temporary tables.

B Always inlined into the main query.

C Ignored if not used.

D Always materialized.

Before version 12, PostgreSQL always materialized CTEs by default.

Which keyword forces a CTE to be inlined in PostgreSQL 12 and later?

A NOT MATERIALIZED

B MATERIALIZED

C INLINE

D NO_MATERIALIZE

The NOT MATERIALIZED keyword forces the CTE to be inlined instead of materialized.

Why might inlining a CTE improve query performance?

A Because it runs the CTE multiple times.

B Because it stores the result permanently.

C Because the planner can optimize the whole query better.

D Because it disables query optimization.

Inlining allows the query planner to optimize the entire query together, often improving performance.

Which PostgreSQL version introduced the option to inline CTEs by default?

A PostgreSQL 12

B PostgreSQL 11

C PostgreSQL 13

D PostgreSQL 10

PostgreSQL 12 introduced the ability to inline CTEs by default when possible.

Explain what CTE materialization means and how it affects query execution in PostgreSQL.

Think about how many times the CTE query runs and what happens to its result.

Describe the difference between MATERIALIZED and NOT MATERIALIZED keywords in PostgreSQL CTEs.

Consider how these keywords influence whether the CTE is stored or merged into the main query.

Practice

1. What does CTE materialization control in PostgreSQL?

easy

A. The data type of columns in the CTE

B. The order of execution of SQL statements

C. The type of join used in the query

D. Whether the CTE results are stored temporarily or inlined into the main query

Solution

1. Step 1: Understand CTE basics

   A Common Table Expression (CTE) is a temporary result set used within a query.

2. Step 2: Identify materialization role

   Materialization controls if PostgreSQL stores the CTE result temporarily or inlines it into the main query for optimization.

3. Final Answer:

   Whether the CTE results are stored temporarily or inlined into the main query -> Option D

4. Quick Check:

   CTE materialization = store or inline [OK]

Hint: Materialization means store or inline CTE results [OK]

Common Mistakes:

• Confusing materialization with join types
• Thinking it controls execution order
• Assuming it changes data types

2. Which of the following is the correct syntax to force PostgreSQL to store a CTE result instead of inlining it?

easy

A. WITH cte_name AS (SELECT * FROM table_name) MATERIALIZED

B. WITH cte_name AS NOT MATERIALIZED (SELECT * FROM table_name)

C. WITH cte_name AS MATERIALIZED (SELECT * FROM table_name)

D. WITH cte_name AS MATERIALIZED SELECT * FROM table_name

Solution

1. Step 1: Recall correct CTE syntax with materialization

   PostgreSQL syntax to force materialization is WITH cte_name AS MATERIALIZED (query).

2. Step 2: Check options for syntax correctness

   WITH cte_name AS MATERIALIZED (SELECT * FROM table_name) matches the correct syntax; others have misplaced keywords or missing parentheses.

3. Final Answer:

   WITH cte_name AS MATERIALIZED (SELECT * FROM table_name) -> Option C

4. Quick Check:

   Materialized keyword goes right after AS [OK]

Hint: Use AS MATERIALIZED right before the query in parentheses [OK]

Common Mistakes:

• Placing MATERIALIZED outside parentheses
• Using NOT MATERIALIZED to force storage
• Omitting parentheses around the query

3. Given the query:

WITH cte AS MATERIALIZED (SELECT id FROM users WHERE active = true) SELECT * FROM cte WHERE id < 10;

What is the effect of MATERIALIZED here?

medium

A. The CTE result is stored temporarily before filtering by id < 10

B. The CTE is inlined and filtered directly without storing

C. The query will produce a syntax error

D. The CTE is ignored and the main query runs alone

Solution

1. Step 1: Understand MATERIALIZED effect on CTE

   Using MATERIALIZED forces PostgreSQL to store the CTE result before the outer query uses it.

2. Step 2: Analyze query filtering

   The outer query filters the stored CTE result by id < 10, so filtering happens after storage.

3. Final Answer:

   The CTE result is stored temporarily before filtering by id < 10 -> Option A

4. Quick Check:

   MATERIALIZED = store then filter [OK]

Hint: MATERIALIZED stores CTE before outer query filters [OK]

Common Mistakes:

• Thinking filtering happens before storing
• Assuming MATERIALIZED causes syntax error
• Believing CTE is ignored

4. Consider this query:

WITH cte AS NOT MATERIALIZED (SELECT * FROM orders WHERE amount > 100) SELECT * FROM cte WHERE amount < 500;

Why might this query fail or behave unexpectedly?

medium

A. NOT MATERIALIZED is not supported in this PostgreSQL version causing syntax error

B. The CTE is stored and filtering is delayed causing slow performance

C. The query ignores the WHERE clause inside the CTE

D. The CTE is materialized twice causing duplicate rows

Solution

1. Step 1: Check PostgreSQL support for NOT MATERIALIZED

   NOT MATERIALIZED was introduced in PostgreSQL 12; older versions do not support it and cause syntax errors.

2. Step 2: Analyze query behavior if unsupported

   If the version is older, the query fails with syntax error due to unrecognized keyword.

3. Final Answer:

   NOT MATERIALIZED is not supported in this PostgreSQL version causing syntax error -> Option A

4. Quick Check:

   NOT MATERIALIZED requires PostgreSQL 12+ [OK]

Hint: Check PostgreSQL version for NOT MATERIALIZED support [OK]

Common Mistakes:

• Assuming NOT MATERIALIZED always works
• Thinking filtering is ignored
• Believing CTE duplicates rows

5. You have a complex query with a CTE that is used multiple times in the main query. You want to optimize performance by avoiding repeated execution of the CTE. Which materialization behavior should you choose and why?

hard

A. Use NOT MATERIALIZED to inline the CTE and let PostgreSQL optimize repeated execution

B. Use MATERIALIZED to store the CTE result once and reuse it multiple times

C. Avoid using CTEs and write subqueries instead for better performance

D. Use MATERIALIZED to force PostgreSQL to re-execute the CTE each time

Solution

1. Step 1: Understand repeated CTE usage

   If a CTE is referenced multiple times, storing its result once avoids repeated computation.

2. Step 2: Choose materialization behavior

   Using MATERIALIZED forces storing the CTE result, so it is computed once and reused, improving performance.

3. Final Answer:

   Use MATERIALIZED to store the CTE result once and reuse it multiple times -> Option B

4. Quick Check:

   MATERIALIZED = store once, reuse many [OK]

Hint: Store CTE once with MATERIALIZED to reuse efficiently [OK]

Common Mistakes:

• Thinking NOT MATERIALIZED avoids repeated execution
• Believing MATERIALIZED causes repeated execution
• Avoiding CTEs unnecessarily