I am developing a component where I am using Entity Framework 4.0.
In a particular scenario I need to make use of a TransactionScope.
try
{
    using (TransactionScope t = new TransactionScope())
    {
        myEntity.MyObjectsOne.Add(new MyObjectOne());
        myEntity.MyObjectsTwo.Add(new MyObjectTwo());
        myEntity.SaveChanges();
        throw new Exception();
        t.Complete();
    }
}
catch (Exception e)
{
    // What should I do?
}
When the transaction fails and thus it is not completed, myEntity will still reflect the changes made i.e. the adding of MyObjectOne and MyObjectTwo.
In such case what is the best practice, in order to avoid inconsistency?
 
     
     
    