Range partitioning by date in PostgreSQL - Time & Space Complexity

Time Complexity: Range partitioning by date

O(p + r)

Understanding Time Complexity

When using range partitioning by date in a database, we want to understand how query speed changes as data grows.

We ask: How does the time to find data change when the table gets bigger?

Scenario Under Consideration

Analyze the time complexity of this PostgreSQL range partitioning setup and query.

CREATE TABLE sales (
  id SERIAL PRIMARY KEY,
  sale_date DATE NOT NULL,
  amount NUMERIC
) PARTITION BY RANGE (sale_date);

CREATE TABLE sales_2023 PARTITION OF sales
  FOR VALUES FROM ('2023-01-01') TO ('2024-01-01');

SELECT * FROM sales WHERE sale_date = '2023-06-15';
    

This code creates a main sales table partitioned by date and queries data for one day.

Identify Repeating Operations

Look at what repeats when the query runs.

• Primary operation: Checking partitions to find the right date range.
• How many times: Once per partition, but only until the matching partition is found.

How Execution Grows With Input

As more partitions are added for new date ranges, the database checks fewer rows inside each partition.

Input Size (n)	Approx. Operations
10 partitions	Checks about 10 partitions, but scans fewer rows per partition
100 partitions	Checks about 100 partitions, still scans fewer rows per partition
1000 partitions	Checks about 1000 partitions, but each partition is small

Pattern observation: The number of partitions checked grows linearly, but each partition is smaller, so total work stays manageable.

Final Time Complexity

Time Complexity: O(λ + r)

This means the time grows with the number of partitions λ checked plus the rows r scanned in the matching partition.

Common Mistake

[X] Wrong: "Partitioning always makes queries run in constant time regardless of data size."

[OK] Correct: The database still needs to find the right partition, which takes time proportional to the number of partitions, and then scan rows inside it.

Interview Connect

Understanding how partitioning affects query time shows you can design databases that handle growing data smoothly.

Self-Check

"What if we changed range partitioning by date to list partitioning by region? How would the time complexity change?"

Practice

1. What is the main purpose of range partitioning by date in PostgreSQL?

easy

A. To create random partitions without any order

B. To split data into parts based on date ranges for better management

C. To encrypt date columns for security

D. To combine all data into a single large table

Solution

1. Step 1: Understand range partitioning concept

   Range partitioning divides data into segments based on continuous ranges, such as dates.

2. Step 2: Identify the purpose of date-based partitioning

   Using date ranges helps organize data by time periods, improving query speed and management.

3. Final Answer:

   To split data into parts based on date ranges for better management -> Option B

4. Quick Check:

   Range partitioning by date = split data by date ranges [OK]

Hint: Range partitioning splits data by continuous date intervals [OK]

Common Mistakes:

• Thinking partitioning combines data instead of splitting
• Confusing partitioning with encryption
• Assuming partitions are random, not range-based

2. Which of the following is the correct syntax to create a range partitioned table by a date column order_date in PostgreSQL?

easy

A. CREATE TABLE orders (id INT, order_date DATE) PARTITION BY RANGE (order_date);

B. CREATE TABLE orders PARTITION BY RANGE (order_date) (id INT, order_date DATE);

C. CREATE TABLE orders (id INT, order_date DATE) PARTITION BY LIST (order_date);

D. CREATE TABLE orders (id INT, order_date DATE) PARTITION BY HASH (order_date);

Solution

1. Step 1: Check correct partitioning clause placement

   In PostgreSQL, PARTITION BY RANGE (column) comes after table columns definition.

2. Step 2: Identify correct partition type for date ranges

   Range partitioning is used for continuous ranges like dates, so PARTITION BY RANGE is correct.

3. Final Answer:

   CREATE TABLE orders (id INT, order_date DATE) PARTITION BY RANGE (order_date); -> Option A

4. Quick Check:

   Syntax: columns then PARTITION BY RANGE [OK]

Hint: Define columns first, then PARTITION BY RANGE (date_column) [OK]

Common Mistakes:

• Placing PARTITION BY before columns
• Using LIST or HASH instead of RANGE for dates
• Incorrect syntax order causing errors

3. Given the following partitioned table and partitions:

CREATE TABLE sales (id INT, sale_date DATE) PARTITION BY RANGE (sale_date);
CREATE TABLE sales_2023 PARTITION OF sales FOR VALUES FROM ('2023-01-01') TO ('2024-01-01');
CREATE TABLE sales_2024 PARTITION OF sales FOR VALUES FROM ('2024-01-01') TO ('2025-01-01');

What will be the result of this query?

SELECT tableoid::regclass, * FROM sales WHERE sale_date = '2023-06-15';

medium

A. Returns rows from sales_2024 partition with sale_date '2023-06-15'

B. Returns no rows because '2023-06-15' is not in any partition

C. Returns rows from both partitions

D. Returns rows from sales_2023 partition with sale_date '2023-06-15'

Solution

1. Step 1: Identify which partition contains '2023-06-15'

   The date '2023-06-15' falls between '2023-01-01' and '2024-01-01', so it belongs to sales_2023 partition.

2. Step 2: Understand query behavior on partitioned tables

   Query on partitioned table routes to matching partition(s) based on WHERE clause; here, only sales_2023 matches.

3. Final Answer:

   Returns rows from sales_2023 partition with sale_date '2023-06-15' -> Option D

4. Quick Check:

   Date in sales_2023 range = rows from sales_2023 [OK]

Hint: Check date range to find correct partition for query [OK]

Common Mistakes:

• Choosing wrong partition based on date
• Assuming query scans all partitions
• Ignoring partition boundaries

4. You try to create a partition for a range partitioned table by date with this command:

CREATE TABLE sales_2025 PARTITION OF sales FOR VALUES FROM ('2025-01-01') TO ('2024-12-31');

What is the problem with this statement?

medium

A. The TO date is earlier than the FROM date, causing a range error

B. Partition names cannot contain numbers

C. You must specify LIST partitioning, not RANGE

D. The sales table must be dropped before adding partitions

Solution

1. Step 1: Check the FROM and TO values in partition definition

   The TO value '2024-12-31' is before the FROM value '2025-01-01', which is invalid for range partitions.

2. Step 2: Understand partition range rules

   Range partitions require FROM value to be less than TO value to define a valid range.

3. Final Answer:

   The TO date is earlier than the FROM date, causing a range error -> Option A

4. Quick Check:

   FROM must be less than TO in range partitions [OK]

Hint: FROM date must be before TO date in range partitions [OK]

Common Mistakes:

• Swapping FROM and TO dates
• Thinking partition names cannot have numbers
• Confusing range with list partitioning

5. You have a large sales table partitioned by month using range partitioning on sale_date. You want to add a new partition for March 2024. Which of the following commands correctly adds this partition?

hard

A. CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-03-01') TO ('2024-03-31');

B. CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-02-28') TO ('2024-03-31');

C. CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-03-01') TO ('2024-04-01');

D. CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-03-01') TO ('2024-03-30');

Solution

1. Step 1: Understand range partition boundaries for months

   Range partitions use inclusive FROM and exclusive TO, so March 2024 is from '2024-03-01' up to but not including '2024-04-01'.

2. Step 2: Check each option's date range correctness

   CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-03-01') TO ('2024-04-01'); correctly uses FROM '2024-03-01' TO '2024-04-01'. Options B, C, and D have incorrect boundaries that either overlap or exclude days.

3. Final Answer:

   CREATE TABLE sales_2024_03 PARTITION OF sales FOR VALUES FROM ('2024-03-01') TO ('2024-04-01'); -> Option C

4. Quick Check:

   Range partitions: FROM inclusive, TO exclusive [OK]

Hint: Use TO date as first day of next month for monthly partitions [OK]

Common Mistakes:

• Using TO date as last day of month (should be exclusive)
• Overlapping partition ranges
• Using incorrect FROM dates