Re: Low cardinality in OLTP with high concurrency
From: Phil H <phil_herring_at_yahoo.com.au>
Date: Wed, 25 Mar 2009 14:18:51 -0700 (PDT)
Message-ID: <ab315f81-c3be-436a-9c0a-c555c05c5b49_at_y34g2000prb.googlegroups.com>
If the column you're selecting against doesn't change often, you can try partitioning the table on that value or its hash. That will mean that your queries will scan 1/16th of the table (on average) and won't need an index. Of course, this will only help if the values aren't highly skewed.
Date: Wed, 25 Mar 2009 14:18:51 -0700 (PDT)
Message-ID: <ab315f81-c3be-436a-9c0a-c555c05c5b49_at_y34g2000prb.googlegroups.com>
If the column you're selecting against doesn't change often, you can try partitioning the table on that value or its hash. That will mean that your queries will scan 1/16th of the table (on average) and won't need an index. Of course, this will only help if the values aren't highly skewed.
There may be some other things you can try. It would help if you provided more information about the table in question.
- Phil