Human-readable ranges of integers or dates

This is a real-world problem that I came across the other day. In a reporting scenario, I wanted to output a number of values in an easy, human-readable way for a report. But just making a long, comma-separated string of numbers doesn’t really make it very readable. This is particularly true when there are hundreds of values.

So here’s a powerful pattern to solve that task.

Continue reading

Comparing nullable columns

Do you ever compare the values of a lot of columns in two tables? Sure you do. Like, for instance, in a cross update, when you need to figure out which rows you should actually update. But it gets worse if the columns are nullable. The fact that any value could potentially be NULL vastly complicates the comparison and might wreak havoc not only on your code but also on your query performance.

But there’s hope.

Continue reading

Key Lookup without an output column?

Performance tuning the other day, I was stumped by a query plan I was looking at. Even though I had constructed a covering index, I was still getting a Key Lookup operator in my query plan. What I usually do when that happens is to check the operator’s properties to see what its output columns are, so I can include those columns in my covering index.

Here’s the interesting thing: there weren’t any output columns. What happened?

Continue reading

Selectively disable “Include actual execution plan”

The “include actual execution plan” feature in SQL Server Management Studio is an invaluable tool for performance tuning. It returns the actual execution plan used for each statement, including actual row counts, tempdb spills and a lot of other information you need to do performance tuning.

But sometimes you want to run a series of statements or procedures where you only want the execution plan for some of the statements. Here’s how:

Continue reading

Windowed DISTINCT aggregates

You may have discovered that the use of DISTINCT is not supported in windowed functions. A query that uses a distinct aggregate in a windowed function,

FROM somewhere;

will generate the following error message:

Msg 10759, Level 15, State 1, Line 1
Use of DISTINCT is not allowed with the OVER clause.

There are, however, a few relatively simple workarounds that are suprisingly efficient.

Continue reading

Last row per group

A very common challenge in T-SQL development is filtering a result so it only shows the last row in each group (partition, in this context). Typically, you’ll see these types of queries for SCD 2 dimension tables, where you only want the most recent version for each dimension member. With the introduction of windowed functions in SQL Server, there are a number of ways to do this, and you’ll see that performance can vary considerably.

Continue reading

Catching circular references in parent-child structures

A popular form of organizing dimensions is in parent-child structures, also known as “unbalanced” or “ragged” dimensions, because any branch can have an arbitrary number of child levels. There are many advantages to this type of representation, but their recursive nature also brings some challenges. In this post, we’re going to look at circular references, and how you can trap them before they run out of control.

Continue reading

Different query plans for “OR” type queries

The SQL Server query optimizer can find interesting ways to tackle seemingly simple operations that can be hard to optimize. Consider the following query on a table with two indexes, one on (a), the other on (b):

FROM #data
WHERE a<=10 OR b<=10000;

The basic problem is that we would really want to use both indexes in a single query.

In this post, we’re going to take a look at a few examples of how this type of query would be optimized, as well as how statistics can affect the query plan, and finally, we’ll take a look at a slightly rare plan operator called “Merge Join (Concatenation)”.

Continue reading