There is always a limit on how large tempdb can get. Even if there are no size limitations on the data files in tempdb, the limit is still the free space on disk drives. Here I’ll show how to calculate the maximum size of tempdb (data only, not log).
sp_MSforeachdb is an undocumented system stored procedure that lets you iterate through all databases on an instance and execute a T-SQL statement. This can be a very handy tool for the DBA. For instance, let us create a script that lists all tables and their size on a server instance.
Want to know the disk space available on your database server from T-SQL? This can be done in many ways, here I'll present two of them. The first way is…
When a database is created, the logged in user is set as the database owner as default. Often, this user account is a physical person and he or she may get more permissions in the database than intended. Normally, a system account should be used as database owner. For instance sa.
Note that when you change the database owner the old account may get less permissions than before, not being the owner (who can do anything) anymore. So check how and where the old account is used before changing.
Are you managing a lot of SQL Server instances? I was setting up a standardized set of SQL Server Alerts on a number of servers and soon realized this was something that could be automated by using a script.
In SQL Server Management Studio, it is possible to execute a multi server query against several servers at the same time. The description from TechNet says most of it:
This topic describes how to query multiple servers at the same time in SQL Server 2012, by creating a local server group, or a Central Management Server and one or more server groups, and one or more registered servers within the groups, and then querying the complete group. The results that are returned by the query can be combined into a single results pane, or can be returned in separate results panes. The results set can include additional columns for the server name and the login that is used by the query on each server. Central Management Servers and subordinate servers can be registered by using only Windows Authentication. Servers in local server groups can be registered by using Windows Authentication or SQL Server Authentication.
Here I’ll show how to create a multi server query from a group under registered servers.
When SQL Server evaluates if an index should be used to retrieve data in a query, selectivity is one of the most important factors. Of course, given that the index column matches the filter or join columns.
If a column only has one or very few unique values, index selectivity is low. True/false columns and gender is common examples.
If there are only unique values in a column, index selectivity is high. Primary keys meets these requirements as they are always unique.
SQL Server favors indexes that are built on columns with high selectivity, since this may return fewer rows and therefore at a lower cost.
If the index has multiple columns, it is preferable (if possible) that the columns are in the order of their selectivity (most selective first).
So how is selectivity computed?
When you submit a query to SQL Server – before the query is executed – SQL Server tries to optimize the way data is retrieved from the tables and indexes to create an execution plan that is as effective as possible. One of the decisions the optimizer in SQL Server has to make is which of the join methods (if any) to use.
There are three join methods: Nested Loops Join, Hash Join and Merge Join.
They all have their relative strengths and weaknesses, which will be covered in this blog post.
Windows environment variables are a great way to store server specific information that may be accessed from different systems on the server. Connection strings, folder paths and ip addresses are common examples of what may be suitable to store as environment variables on the server.
A XML document is hierarchical. A value gets meaning from its position in the document. Nodes can be nested so that a parent node can contain children nodes. This way the child nodes inherits all the values from the parent.
In a relational database the same thing can be achieved, but by using a different technology. Here we use key values in separate tables instead. Information from the parent or child can be retrieved by using relations between tables.
XML is commonly used in applications to store information about objects, and a relational database is often used to persist that information.
So how can a hierarchical XML document be inserted to relational tables? In this blog post I will show one possible technique, using only T-SQL and XQuery. In the example the tables will have their own internal primary keys defined which will add to the complexity, but it is a rather common scenario.