All Forums UDA
jerome 3 posts Joined 04/08
30 Jun 2011
Join Index's PI data distribution

Hi ,

We have an Join index being used in our project (on a fact table).
As everybody know- we have an PI in the Join index and as usual based on this PI ,data gets distributed across AMPs.
We would use HASHAMP(HASHBUCKET(HASHROW(PI_Col))) to find the distribution of the records in a normal table.
In a similar manner - would it be possible to find the distribution of data in the Join index ?


mnylin 118 posts Joined 12/09
30 Jun 2011

If you're looking to estimate it before index creation, use the SELECT statement from the definition in a derived table and wrap the functions you mentioned around the DT definition. Or, after it is created, you can query the DBC tables and see the space used per amp by the index.

You must sign in to leave a comment.