Question) Is it a a good idea to identify skew factor of input tables of a query before executing the same, especially for large data?
Question) If yes what is threshold beyond which skew factor is not good so that indexes be re-arranged to have lower skew factor?
Hope only the skewfactor of target tables needs to be considered, bcoz it is where the data distribution is going to happen. cardinality of the source data for target table primary index column / columns should be unique or contains less duplicate. If it is going to have more non-unique value or duplicate value then skewness will get increased as most of the data will rush into single amp