I’ve made several posts about joins but one option that I haven’t talked about as yet, are what are called non-equi joins.
These are less common, and they often don’t optimize all that well, but it’s worth noting that joins don’t always have to be on equality, or what are called equi-joins.
Non-equi join sounds complex but all we’re saying is that the things that we’re joining on are not equal signs.
2025-02-06
This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.
Section: Development Level: Medium
Question:
You see the following statement in a query that you are reviewing.
What is the purpose of the NULLIF function in this statement?
SELECT @a / NULLIF(@b, 0);
What was the author trying to achieve?
Answer:
The variable @a is being divided by the variable @b, but the variable @b might be zero. In that case, a divide by zero error would be returned.
2025-02-05
One thing you might not have considered is that a table can also be joined to itself.
Imagine that I want a list of employees, but I also want their manager’s details, the employee ID and the name for the manager. Now the problem is they’re both in the same table, but they’re different rows in that same table because a manager is also an employee.
The way I did this in the example is to say that I’m reading from Employees and then joining to Employees i.e. the same table.
2025-02-05
This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.
Section: Development Level: Intro
Question:
I execute the following query:
SELECT st.SalesTerritoryID AS [ID],
st.SalesTerritoryName AS [Name],
st.DateCreated
FROM dbo.SalesTerritories AS st
UNION
SELECT p.ProductID AS [Product],
p.[Description],
p.DateCreated AS [StartDate]
FROM dbo.Products AS p;
What alias will be returned for the each column in the returned results?
2025-02-04
Of course, you might also need to join more than two tables. The example above shows how to do that.
I started with the Orders table, then joined to the OrderLines table, and finally, I’ve also joined to Products.
It’s another example of why I prefer the modern join syntax. Instead of just listing three tables and a bunch of WHERE clause predicates, for each table involved, I specify how it’s joined.
2025-02-04
Our free SDU Tools for developers and DBAs, now includes a very large number of tools, with procedures, functions, and views. A while back, we added the ability to retrieve the version of the toolset programmatically, but an interesting request that we had, asked for the ability to query the toolset for the tools themselves. The Tools view now does that.
The view returns the following columns:
ToolName ToolTypeCode ToolType Category Description VideoURL
2025-02-03
After discussing LEFT OUTER JOIN, you might wonder if there is also a RIGHT OUTER JOIN. There is, and it’s the reverse of the left one.
What a RIGHT OUTER JOIN from Products to ProductGroups says is that I want at least one row for every row in ProductGroups. You can see that in the query example above. The product group Hot Food currently has no products in it. The RIGHT OUTER JOIN causes it to be returned, as similar to what happens with a LEFT OUTER JOIN, the columns from the non-matching table are returned as NULL. In this example, the product description for that group is returned as NULL.
2025-02-03
The next type of join that I want to discuss in these posts is a left outer join. We use this mostly, when we know that not all rows are matching.
For example, in the query shown above, I wanted to add up the ordered quantity values for all products. The problem is that if I used an INNER JOIN, I would only get products that match at least one order line.
2025-02-02
There are two ways you can write common join syntax. There’s an older way to write the query and a new way to write the query.
In the image above, you can see the new way written first, and then the old way written below. The newer syntax was introduced as part of ANSI SQL, back in 1992. It has been recommended for over 30 years now.
When you look at the older syntax, you’ll notice that I’m doing two things in the WHERE clause. I’m both determining how the tables are joined, and I’m also filtering the rows that I want.
2025-02-01
The most common form of join between two tables is what’s called an inner join.
In the example shown above, I’ve selected some columns from different tables. I’ve got description and product ID from the product table and I’ve got the product group name from the product group. What I wanted was a list of products but I want to know what group they’re part of. So to do that I had to use the products table, but I also had to join to the product groups table so I could get the name of the product group.
2025-01-31