The following pattern is quite common in database programming: IF EXISTS(some query) BEGIN   DO SOMETHING; END When such code runs in high concurrency situations, it may not work as expected. I will provide a repro when such logic fails 40% of the time. The following script provides a test table and attempts to ...

If your data has a small percentage of duplicates, then IGNORE_DUP_KEY may speed up your inserts. For larger amounts of duplicates, IGNORE_DUP_KEY may slow them down significantly. I set up two tables, stripping down all the irrelevant details, as follows:   CREATE TABLE t1(n INT NOT NULL PRIMARY KEY) GO CREATE TABLE ...