Question

I have a table message_message with 3000000 records.

when I make a count(*) query, It’s very slow…:

mysql> select count(*) from message_message;

+———-+

| count(*) |

+———-+

| 2819416 |

+———-+

1 row in set (2 min 35.35 sec)

explain it:
mysql> explain select count(*) from message_message;

| id | select_type| table | type | possible_keys | key | key_len | ref | rows |Extra |

| 1 | SIMPLE | message_message | index | NULL | PRIMARY | 4 | NULL | 2939870 | Using index |

1 row in set (0.02 sec)

what happen?

Answer 1

Have a look at This Post in InnoDB you need to do a full table scan, where as in MyISAM its a index read.

If you use a where clause though it changes the execution pattern to use indexes, so in general InnoDB will be slower than MyISAM on full unrestricted counts, where as the performance matches up on restricted counts.