I have a situation where I have to join the same table twice but with a different joining column. If I will declare one column as Primary index and one column as secondary index, will that be a wise implementation in performance point of view? Or is there any other way to do this. Please suggest.
Chosing the Join columns of one of the joins as the Primary Index makes definitely sense. Regarding the secondary index: It really depends if the index can be used. This is influenced by selectivity, join type (nested join possible?), etc.
I don't think there is a general advise which can be given. More details would be required.