unique key based on 2 columns in SQl Server 2008?

Question

is it posible to have SQL Server to have a unique key based on 2 columns?

I can have duplicates in both columns but not at the same time:

MfgID  :   CustNum   
1      :     Cust01   
1      :     Cust02  
2      :     Cust02  
1      :     Cust03  
3      :     Cust03  
3      :     Cust04  
1      :     Cust02

In the example above all of these would be fine EXCEPT the last row. I'd like SQL Server to kck out an error on this for me.

Answer 1

CREATE UNIQUE NONCLUSTERED INDEX IX_TableName_MfgID_CustNum ON TableName
(
        MfgID  ,
        Column2
) WITH( IGNORE_DUP_KEY = OFF)

Answer 2

CREATE TABLE table1(
  MfgID INTEGER NOT NULL,
  CustNum VARCHAR(255) NOT NULL,
  CONSTRAINT unique_1 UNIQUE (MfgID, CustNum)
)

OR

ALTER TABLE table1 ADD CONSTRAINT unique_1 UNIQUE(fgID, CustNum)