>Couple of suggestions from a section in a book I have concerning avoiding deadlocks.
>
>(1) Always access objects in same order- if two processes are updating in same order, deadlock can not occur because one table will always get the lock, locking the other out. Also, moving to stored procedures helps in this regard because it centralizes code.
>
>(2) Keep transactions as short as possible
>
>(3) Keep isolations levels as low as possible.
>
>Charlie

Good sugggestions. But why does a SELECT have anything at all to do with a deadlock?