CodeBlanks Logo Icon CodeBlanks
Login Register

Cost model

Learn cost-based planning.

Cost model

The cost model computes the CPU and IO cost of each plan step as a number. The plan with the lowest total cost wins.

-- EXPLAIN çıktısı:
-- Seq Scan on orders (cost=0.00..34.00 rows=1000)

Cost is the sum of CPU and IO.

Cost model — details

Cost is an estimated resource use, not real time. EXPLAIN ANALYZE shows both estimate and actual time; it reveals the gap between plan and reality.

Practice Cost model

Run EXPLAIN ANALYZE on all rows of orders to compare estimate and actual time.

Starter code

query.sql
-- cost and time
EXPLAIN

Sample solution

query.sql
EXPLAIN ANALYZE SELECT * FROM orders;

Expected output

Seq Scan on orders (actual time=0.012..0.018 rows=5)

Questions and answers

Review the lesson questions together with their answers. To test yourself, solve the lesson interactively.

1 Cost model: match the terms

Term Meaning
cost cost
CPU CPU
IO IO
plan plan

2 See actual time

Question 
EXPLAIN ____ SELECT * FROM orders;
Answer 
EXPLAIN ANALYZE SELECT * FROM orders;

Explanation: ANALYZE is the right answer.

3 What does cost measure?

  • CPU ve IO Correct answer
  • An hour
  • Money
  • The number of rows

Explanation: Cost measures CPU and IO usage.

All lessons in Expert

1 Database design

  1. ER model
  2. Cardinality
  3. Surrogate keys
  4. Natural keys
  5. Denormalization
  6. Summary - Design

2 Index tuning

  1. B-tree structure
  2. Covering index
  3. Partial index
  4. Index selectivity
  5. Index maintenance
  6. Summary - Index tuning

3 Query optimizer

  1. What is an optimizer?
  2. Statistics
  3. Cost model
  4. Join strategies
  5. Plan cache
  6. Summary - Optimizer

4 Project: Performance tuning

  1. Scenario design
  2. Reading plans
  3. Adding indexes
  4. Rewriting queries
  5. Comparing results

5 Replication and sharding

  1. What is replication?
  2. Master-slave
  3. Sync vs async
  4. What is sharding?
  5. Shard key
  6. Summary - Replication and sharding

6 Distributed and big data

  1. Distributed queries
  2. MPP
  3. Columnar storage
  4. SQL on Hadoop
  5. SQL vs NoSQL
  6. Summary - Distributed and big data

7 Security and backup

  1. SQL injection
  2. Parameterised queries
  3. Authorisation
  4. Backup
  5. Recovery
  6. Summary - Security and backup

8 Expert final

  1. Design challenge
  2. Optimization challenge
  3. Security challenge
  4. Backup challenge