Well, if you are going to read a large fraction of the total rows, it
imay be faster to skip the index.
Veniamin Konoplev wrote:
> Hello!
>
> I have a big indexed table (by columns "date,...").
> If a range of date in the SELECT statement inreases,
> the index stops to work.
>
> mysql> explain select count(*),SUM(In_byts) from ip_summary_d where date >
> '2001-10-21' and date < '2001-11-01';
>
> +--------------+-------+---------------+---------+---------+------+---------+------------+
> | table | type | possible_keys | key | key_len | ref | rows | Extra
> |
>
> +--------------+-------+---------------+---------+---------+------+---------+------------+
> | ip_summary_d | range | PRIMARY | PRIMARY | 3 | NULL | 2394968 | where
> used |
>
> +--------------+-------+---------------+---------+---------+------+---------+------------+
> 1 row in set (0.00 sec)
>
> mysql> explain select count(*),SUM(In_byts) from ip_summary_d where date >
> '2001-10-20' and date < '2001-11-01';
>
> +--------------+------+---------------+------+---------+------+----------+------------+
> | table | type | possible_keys | key | key_len | ref | rows | Extra
> |
>
> +--------------+------+---------------+------+---------+------+----------+------------+
> | ip_summary_d | ALL | PRIMARY | NULL | NULL | NULL | 13719096 | where used
> |
>
> +--------------+------+---------------+------+---------+------+----------+------------+
> 1 row in set (0.00 sec)
>
>
> Help me please, what could be wrong ?!
>
>
