I have a table with 75 rows. The table is defined as MULTISET table with UNIQUE PRIMARY INDEX.
When I calculated the skew factor, it came close to 77.
If the table is created with UPI, this must ensure that the data is distributed evenly accross AMPs. What could be the possible reason for this high skew factor?
Any help is appreciated.
The rows always are distributed on the AMPs by the hash of the columns selected for the Primary Index.
This is so regardless the table is SET or MULTISET.