handling skewness issues

[rajat kumar]

Hi All,

How to overcome skewness issues in spark ?

I read that we can add some randomness to key column before join and remove
that random part after join.

is there any better way ? Above method seems to be a workaround.

thanks
rajat