4 ms·
Count(*) vs. COUNT(id) in MySQL
- lun4r 4y agomysql> EXPLAIN SELECT COUNT(*) FROM users \G \*\*\*\*\*\*\*\*\*\*\*\*\** 1. row \*\*\*\*\*\*\*\*\*\*\*\*\*\* id: 1 select_type: SIMPLE table: users partitions: NULL type: index possible_keys: NULL key: needs_review key_len: 1 ref: NULL rows: 5290406 filtered: 100.00 Extra: Using index 1 row in set, 1 warning (0.01 sec) mysql> EXPLAIN SELECT COUNT(id) FROM users \G \*\*\*\*\*\*\*\*\*\*\*\*\*\* 1. row \*\*\*\*\*\*\*\*\*\*\*\*\*\* id: 1 select_type: SIMPLE table: users partitions: NULL type: index possible_keys: NULL key: needs_review key_len: 1 ref: NULL rows: 5290406 filtered: 100.00 Extra: Using index 1 row in set, 1 warning (0.00 sec) Both (*) and (id) appear to be using the same index. According to the author, the second query would use PRIMARY instead.