Helper Predicates And Multi-Column Filters

Rob Farley has an interesting post on optimizing a lookup when you have separate date and time columns:

Here we see a Seek Predicate that looks for OrderDate values between two values that have been worked out elsewhere in the plan, but creating a range in which the right values must exist. This isn’t >= 20110805 00:00 and < 20110806 00:00 (which is what I would’ve made it), it’s something else. The value for start of this range must be smaller than 20110805 00:00, because it’s >, not >=. All we can really say is that when someone within Microsoft implemented how the QO should respond to this kind of predicate, they gave it enough information to come up with what I call a “helper predicate.”

Now, I would love Microsoft to make more functions sargable, but that particular request was Closed long before they retired Connect.

But maybe what I mean is for them to make more helper predicates.

The problem with helper predicates is that they almost certainly read more rows than you want. But it’s still way better than looking through the whole index.

Read the whole thing.

Related Posts

Nested Loops And Implicit Reordering

Dmitry Piliugin shows how the SQL Server optimizer can end up reordering data in a nested loops join to improve performance: The purpose is to minimize random access impact. If we perform an Index Seek (with a partial scan, probably) we read the entries in the index order, in our case, in the order of […]

Read More

When Table Join Order Oughtn’t Matter…But It Sometimes Does

Bert Wagner looks at join order in SQL Server: SQL is a declarative language: you write code that specifies *what* data to get, not *how* to get it. Basically, the SQL Server query optimizer takes your SQL query and decides on its own how it thinks it should get the data. It does this by […]

Read More

Leave a Reply

Your email address will not be published. Required fields are marked *

Categories

April 2018
MTWTFSS
« Mar  
 1
2345678
9101112131415
16171819202122
23242526272829
30