Question

I have a list of users in a database table called Users (SQL Server 2008 R2). In addition to the user’s UserName, there are two fields that classify the user – for simplicity we’ll say Department and JobTitle.

|  UserName  |  Department  |  JobTitle  |
------------------------------------------
|    Joe     |     IT       |  SysAdmin  |
|    Jim     |     IT       |    DBA     |
|    Jeff    |    Sales     |  SalesMgr  |
|    Mack    |    Sales     |    Rep     |

I also have a table, ActiveJobs, that lists certain combinations of Department and JobTitle that I actually care about.

|  Department  |  JobTitle  |
-----------------------------
|     IT       |  SysAdmin  |
|    Sales     |  SalesMgr  |
|    Sales     |    Rep     |

I want to select each of the records from Users that matches the combination of Department / JobTitle in ActiveJobs. I thought this query would do it:

SELECT Users.*
FROM Users
INNER JOIN ActiveJobs DEP
ON Users.Department = DEP.Department
INNER JOIN ActiveJobs JOB
ON Users.JobTitle = JOB.JobTitle

But that returns the same User record more than once in many cases (which I think is caused by the duplicates in the Department column – but I don’t really understand why). For the example above, I’m getting (Joe, Joe, Jim, Mack) even though I was hoping to just get (Joe, Jim, Mack).

What query would get the subset of User records that has a matching combination of Department and JobTitle in Active Jobs?

Answer 1

Put an “AND” in your join clause instead of joining twice.

SELECT Users.*
FROM Users
INNER JOIN ActiveJobs DEP
ON Users.Department = DEP.Department AND Users.JobTitle = DEP.JobTitle