sql-optimization-patterns — detailed patterns and worked examples

Optimization Patterns

Pattern 1: Eliminate N+1 Queries

Problem: N+1 Query Anti-Pattern

# Bad: Executes N+1 queries
users = db.query("SELECT * FROM users LIMIT 10")
for user in users:
    orders = db.query("SELECT * FROM orders WHERE user_id = ?", user.id)
    # Process orders

Solution: Use JOINs or Batch Loading

-- Solution 1: JOIN
SELECT
    u.id, u.name,
    o.id as order_id, o.total
FROM users u
LEFT JOIN orders o ON u.id = o.user_id
WHERE u.id IN (1, 2, 3, 4, 5);

-- Solution 2: Batch query
SELECT * FROM orders
WHERE user_id IN (1, 2, 3, 4, 5);

# Good: Single query with JOIN or batch load
# Using JOIN
results = db.query("""
    SELECT u.id, u.name, o.id as order_id, o.total
    FROM users u
    LEFT JOIN orders o ON u.id = o.user_id
    WHERE u.id IN (1, 2, 3, 4, 5)
""")

# Or batch load
users = db.query("SELECT * FROM users LIMIT 10")
user_ids = [u.id for u in users]
orders = db.query(
    "SELECT * FROM orders WHERE user_id IN (?)",
    user_ids
)
# Group orders by user_id
orders_by_user = {}
for order in orders:
    orders_by_user.setdefault(order.user_id, []).append(order)

Pattern 2: Optimize Pagination

Bad: OFFSET on Large Tables

-- Slow for large offsets
SELECT * FROM users
ORDER BY created_at DESC
LIMIT 20 OFFSET 100000;  -- Very slow!

Good: Cursor-Based Pagination

-- Much faster: Use cursor (last seen ID)
SELECT * FROM users
WHERE created_at < '2024-01-15 10:30:00'  -- Last cursor
ORDER BY created_at DESC
LIMIT 20;

-- With composite sorting
SELECT * FROM users
WHERE (created_at, id) < ('2024-01-15 10:30:00', 12345)
ORDER BY created_at DESC, id DESC
LIMIT 20;

-- Requires index
CREATE INDEX idx_users_cursor ON users(created_at DESC, id DESC);

Pattern 3: Aggregate Efficiently

Optimize COUNT Queries:

-- Bad: Counts all rows
SELECT COUNT(*) FROM orders;  -- Slow on large tables

-- Good: Use estimates for approximate counts
SELECT reltuples::bigint AS estimate
FROM pg_class
WHERE relname = 'orders';

-- Good: Filter before counting
SELECT COUNT(*) FROM orders
WHERE created_at > NOW() - INTERVAL '7 days';

-- Better: Use index-only scan
CREATE INDEX idx_orders_created ON orders(created_at);
SELECT COUNT(*) FROM orders
WHERE created_at > NOW() - INTERVAL '7 days';

Optimize GROUP BY:

-- Bad: Group by then filter
SELECT user_id, COUNT(*) as order_count
FROM orders
GROUP BY user_id
HAVING COUNT(*) > 10;

-- Better: Filter first, then group (if possible)
SELECT user_id, COUNT(*) as order_count
FROM orders
WHERE status = 'completed'
GROUP BY user_id
HAVING COUNT(*) > 10;

-- Best: Use covering index
CREATE INDEX idx_orders_user_status ON orders(user_id, status);

Pattern 4: Subquery Optimization

Transform Correlated Subqueries:

-- Bad: Correlated subquery (runs for each row)
SELECT u.name, u.email,
    (SELECT COUNT(*) FROM orders o WHERE o.user_id = u.id) as order_count
FROM users u;

-- Good: JOIN with aggregation
SELECT u.name, u.email, COUNT(o.id) as order_count
FROM users u
LEFT JOIN orders o ON o.user_id = u.id
GROUP BY u.id, u.name, u.email;

-- Better: Use window functions
SELECT DISTINCT ON (u.id)
    u.name, u.email,
    COUNT(o.id) OVER (PARTITION BY u.id) as order_count
FROM users u
LEFT JOIN orders o ON o.user_id = u.id;

Use CTEs for Clarity:

-- Using Common Table Expressions
WITH recent_users AS (
    SELECT id, name, email
    FROM users
    WHERE created_at > NOW() - INTERVAL '30 days'
),
user_order_counts AS (
    SELECT user_id, COUNT(*) as order_count
    FROM orders
    WHERE created_at > NOW() - INTERVAL '30 days'
    GROUP BY user_id
)
SELECT ru.name, ru.email, COALESCE(uoc.order_count, 0) as orders
FROM recent_users ru
LEFT JOIN user_order_counts uoc ON ru.id = uoc.user_id;

Pattern 5: Batch Operations

Batch INSERT:

-- Bad: Multiple individual inserts
INSERT INTO users (name, email) VALUES ('Alice', 'alice@example.com');
INSERT INTO users (name, email) VALUES ('Bob', 'bob@example.com');
INSERT INTO users (name, email) VALUES ('Carol', 'carol@example.com');

-- Good: Batch insert
INSERT INTO users (name, email) VALUES
    ('Alice', 'alice@example.com'),
    ('Bob', 'bob@example.com'),
    ('Carol', 'carol@example.com');

-- Better: Use COPY for bulk inserts (PostgreSQL)
COPY users (name, email) FROM '/tmp/users.csv' CSV HEADER;

Batch UPDATE:

-- Bad: Update in loop
UPDATE users SET status = 'active' WHERE id = 1;
UPDATE users SET status = 'active' WHERE id = 2;
-- ... repeat for many IDs

-- Good: Single UPDATE with IN clause
UPDATE users
SET status = 'active'
WHERE id IN (1, 2, 3, 4, 5, ...);

-- Better: Use temporary table for large batches
CREATE TEMP TABLE temp_user_updates (id INT, new_status VARCHAR);
INSERT INTO temp_user_updates VALUES (1, 'active'), (2, 'active'), ...;

UPDATE users u
SET status = t.new_status
FROM temp_user_updates t
WHERE u.id = t.id;

Advanced Techniques

Materialized Views

Pre-compute expensive queries.

-- Create materialized view
CREATE MATERIALIZED VIEW user_order_summary AS
SELECT
    u.id,
    u.name,
    COUNT(o.id) as total_orders,
    SUM(o.total) as total_spent,
    MAX(o.created_at) as last_order_date
FROM users u
LEFT JOIN orders o ON u.id = o.user_id
GROUP BY u.id, u.name;

-- Add index to materialized view
CREATE INDEX idx_user_summary_spent ON user_order_summary(total_spent DESC);

-- Refresh materialized view
REFRESH MATERIALIZED VIEW user_order_summary;

-- Concurrent refresh (PostgreSQL)
REFRESH MATERIALIZED VIEW CONCURRENTLY user_order_summary;

-- Query materialized view (very fast)
SELECT * FROM user_order_summary
WHERE total_spent > 1000
ORDER BY total_spent DESC;

Partitioning

Split large tables for better performance.

-- Range partitioning by date (PostgreSQL)
CREATE TABLE orders (
    id SERIAL,
    user_id INT,
    total DECIMAL,
    created_at TIMESTAMP
) PARTITION BY RANGE (created_at);

-- Create partitions
CREATE TABLE orders_2024_q1 PARTITION OF orders
    FOR VALUES FROM ('2024-01-01') TO ('2024-04-01');

CREATE TABLE orders_2024_q2 PARTITION OF orders
    FOR VALUES FROM ('2024-04-01') TO ('2024-07-01');

-- Queries automatically use appropriate partition
SELECT * FROM orders
WHERE created_at BETWEEN '2024-02-01' AND '2024-02-28';
-- Only scans orders_2024_q1 partition

Query Hints and Optimization

-- Force index usage (MySQL)
SELECT * FROM users
USE INDEX (idx_users_email)
WHERE email = 'user@example.com';

-- Parallel query (PostgreSQL)
SET max_parallel_workers_per_gather = 4;
SELECT * FROM large_table WHERE condition;

-- Join hints (PostgreSQL)
SET enable_nestloop = OFF;  -- Force hash or merge join

6.7 KiB Raw Blame History