In the last episode (May 26), Daniel Kasak said:
> I have a query:
>
> select RSLCOMID, AccountNo, Client
> from CallData
> group by RSLCOMID, AccountNo, Client
>
> An 'explain' command gives:
>
>
> +----------+------+---------------+------+---------+------+--------+---------------------------------+
> | table | type | possible_keys | key | key_len | ref | rows | Extra
> |
>
> +----------+------+---------------+------+---------+------+--------+---------------------------------+
> | CallData | ALL | NULL | NULL | NULL | NULL | 938904 | Using
> temporary; Using filesort |
>
> +----------+------+---------------+------+---------+------+--------+---------------------------------+
>
> I have indexes on all three fields in the select list. The query takes
> about 90 seconds to run (500Mhz / 256MB). Is there any way to speed the
> query up? Would an index across all three fields improve things?
That's the only thing that will. Mysql can only use one index per
table.
--
Dan Nelson
dnelson@stripped
