One Star

How to restart a job from the failed point data.

I would like to know  how to handle the below scenario:
 I * I am loading a data from a oracle DB to target Teradata DB .
· L*  Let say there are 1000 records and while loading the data the job get failed .
·    *  Let say it has read the data till 500 records  and its failed after that .
·     *  So I would like to restart the job  to load the data from 501 records not from the start . 
Is there any way to achieve this . I have just taken an example,  in real time the data might can be in millions .