Hi All,
I have read about deadlocks here on Google and I was surprised to read
that an update and a select on the same table could get into a
deadlock because of the table's index. The update and the select
access the index in opposite orders, thereby causing the deadlock.
This sounds to me as a bug in SQL Server!
My question is: Could you avoid this by reading the table with a
'select * from X(updlock)' before updating it? I mean: Would this
result in the update transaction setting a lock on the index rows
before accessing the data rows?
Merry Christmas!
/Fredrik Möller