IBM Mainframe Forum

Hi Friends,

I have a question on requirement w.r.t need for a new secondary index. The requirement is as below:

We have a root segment- Order segment - having key as "Order No". The other fields on the root segment are "customer-code", "order-date", and "ship-indicator" alongwith some other fields.

We need to display all the orders for a given "customer-code" (customer code is input by the user on the inquiry screen) on to the screen in sequence by "order-date" - order with the oldest date first. Only those orders should be displayed for which "order-date" < current-date AND for which "ship-indicator" field is blank.

The database already has secondary index on the root segment based on customer code.

However, since the requirement is to display the orders in sequence by "order-date", I am thinking of defining an index on "order-date" that should point to the root segment. Please let me know if I am correct in this approach.

Should I define:

(1) a single index on "Order-date". OR
(2) a single index on "Order-date" AND "ship-indicator".OR
(3) a single index on "Order-date" AND "ship-indicator AND cusotmer-code".

The order for fields in the index is as given above.

Please let me know your views/thoughts.

Thanks,
Ajay

This is a double post - I was going to lock it but I see Enrico has done that to the other one.