Hi i have few questions on Partition,

We have a fact table which has 320 million records, we join the fact table using the time dimension to display monetary values on reports.

We decided to partition the fact table based on time_key_id column in fact table.
When we directly limit the time_key_id column in fact table, it return data in few seconds. But joining the table with time dimension, it takes some minutes to return results.

Is this the scenairo in db2, why it takes some time to return result while joining with time dimension.

The time_key_id has index on both time dimension and also in fact table.

Thanks,
Sathish

For joining the table with time dimension, i think db2 will parse data in different storages that *may* impact the performance.

The rule is : even the best indexed join will be less performant than just limiting data in one table.

Have you had a look at the explain plan?