: Re: performance on single column index with few distinct values
Hi Daniel,
Here's some more data which might help. This query does not appear to be
using the composite index although there is one on the master_id and
ticket_id. The below query takes about 3 seconds. Anyway I think I wil
Hi,
remember that using index can be slower that FTS. if there is a low index column
selectivity, it's normal that your query doesn't use it.
try using some hints to force index usage (like in oracle) or put your table as
an index organiztion table.
look also at myisam merge tables.
mathias
S
Hi Daniel,
Here's some more data which might help. This query does not appear to be
using the composite index although there is one on the master_id and
ticket_id. The below query takes about 3 seconds. Anyway I think I will
still move to seperate tables for each master id. Thanks for the repl
On 5/28/05, Terence wrote:
>
> Master ID is used to distinguish multiple helpdesks. In this table there
> are 100k records, but only 10 distinct master_id's.
>
> ticket_id master_id
> 1 1
> 2 1
> 3 2
> 4 2
> 5 3
> ...
Is there a composite index on (master_id, ticket_id)? Since your queries
are selecting on a particular master_id, and ordering by ticket_id, along
with the limit I think MySQL would be able to use such an index in an
optimization.
-Daniel
-Original Message-
From: Terence [mailto:[EMAIL PR
