RangePartition indexData based on index column

[ashokblend]

This question is related to covering index.

We have around 100 columns in table and we are creating covering index with 6 index columns.
Below is some stats about no of task launched when hyperspace is enabled/disabled.
Hyperspace enabled: 127
Hyperspace disabled: 784

But query performance time which i get is almost same in both case.
Our source data is range partitioned to get dataskipping benefit during query.

As i understand, Hyperspace uses bucketing to organize data based on index column and also data is sorted by index column inside bucket... hence we should get better query performance when hyperspace is enabled ?

As i think of may be bucketed data is spread across file and sort happens within file and not across file. This might be reason for not getting better query performance.

So another quersion is can we rangepartitoin indexdata instead of bucketing ?

[ashokblend]

i think ##495 answers this question.