Is current perm of a table is inversely proportional to Skew factor [SF] ?
I had a table [WFLOW_RUN] with 93% skew factor. Hence have changed its PI, and copied data to have better distribution.. The SF has been reduced, but the current perm has been increased 2 times... but both have same rows..
What's the number of rows and the number of AMPs in your system?
Nobody cares about skew for a small table like this...
Check if the 2nd table is defined with FALLBACK