Tag Archives: Development

Articles on Microsoft SQL Server development

#0415 – SQL Server – Performance Tuning – Use STRING_AGG to generate comma separated strings


With more and more data being exchanged over APIs, generating comma-separated strings are becoming a much more common requirement.

A few years ago, I wrote about two different ways to generate comma-separated strings. The most common one I find to be in use when generating comma-separated values from a table is the intermediate conversion of XML. This however, is a very costly mechanism and can potentially take minutes for the query to run depending upon the amount of data involved.

SQL Server 2017 brings a new aggregate function that can be used to generate comma-separated values extremely fast. The function is STRING_AGG().

Here’s a sample of it’s usage:


 --WARNING: THIS SCRIPT IS PROVIDED AS-IS AND WITHOUT
-- WARRANTY.
-- FOR DEMONSTRATION PURPOSES ONLY
--Step 01: Generate Temp table to store source data
DECLARE @NamesTable TABLE ([Id] INT,
[Name] NVARCHAR(50)
);
--Step 02: Generate test data
INSERT INTO @NamesTable
VALUES (1, 'A'),
(2, 'D'),
(2, 'C'),
(3, 'E'),
(3, 'H'),
(3, 'G');
--Step 03: Using STRING_AGG to generate comma-separated strings
SELECT STRING_AGG(tbl.Name, ',') AS [CommaSeparatedString]
FROM @NamesTable AS tbl;
GO
/RESULTS**
CommaSeparatedString
A,D,C,E,H,G
*/

Advantages of STRING_AGG:

  • Can be used just like any other aggregate function in a query
  • Can work with any user supplied separator – doesn’t necessarily have to be a comma
  • No manual step required – Separators are not added at the end of the concatenated string
  • STRING_AGG() is significantly faster than using XML based methods
  • Can be used with any compatibility level as long as the version is SQL Server 2017 (or higher) and Azure SQL database

Here’s an example of how STRING_AGG can be used with any separator:

 --WARNING: THIS SCRIPT IS PROVIDED AS-IS AND WITHOUT
-- WARRANTY.
-- FOR DEMONSTRATION PURPOSES ONLY
--Step 01: Generate Temp table to store source data
DECLARE @NamesTable TABLE ([Id] INT,
[Name] NVARCHAR(50)
);
--Step 02: Generate test data
INSERT INTO @NamesTable
VALUES (1, 'A'),
(2, 'D'),
(2, 'C'),
(3, 'E'),
(3, 'H'),
(3, 'G');
--Step 03: Using STRING_AGG to generate comma-separated strings
SELECT STRING_AGG(tbl.Name, '-*-') AS [CustomSeparatorString]
FROM @NamesTable AS tbl;
GO
/RESULTS**
CustomSeparatorString
A--D--C--E--H--G /

A minor challenge

As with every new feature, there may be a small usability challenge with STRING_AGG. One cannot use keywords like DISTINCT to ensure that only distinct values are used for generating the comma-separated string. There is however a Azure feedback item open where you can exercise your vote if you feel this feature is useful.

Further Reading

  • Different ways to generate a comma-separated string from a table [Blog Link]
  • STRING_AGG() Aggregate Function [MSDN BOL]

Until we meet next time,

Be courteous. Drive responsibly.

Advertisements

#0412 – SQL Server – SSIS – Error – The value type (__ComObject) can only be converted to variables of type Object. Variables may not change type during execution.


Recently, we were manipulating a string in an “Execute SQL” task inside a SSIS package, when we ran into the following sequence of errors.

[Execute SQL Task] Error: The value type (__ComObject) can only be converted to variables of type Object.
[Execute SQL Task] Error: An error occurred while assigning a value to variable "MyStringVariable": "The type of the value (DBNull) being assigned to variable "User::MyStringVariable" differs from the current variable type (String). Variables may not change type during execution. Variable types are strict, except for variables of type Object.".
Error: The type of the value (DBNull) being assigned to variable "User::MyStringVariable" differs from the current variable type (String). Variables may not change type during execution. Variable types are strict, except for variables of type Object.

The Execute SQL was similar to something that we had done hundreds of times before, and therefore we were stumped by the error. I found the root cause interesting and hence wanted to write about it right away.

The Test Setup

Before we go ahead, allow me to walk through the sample SSIS package which we used to reproduce the issue. As I mentioned, it is a simple SSIS package with a single “Execute SQL Task”.

0412_01_SSISExecuteSQLTask

The Execute SQL task in the sample SSIS package

The “Execute SQL” task simply executes a T-SQL statement that returns a single-row result set and sets a package variable of type “string“.

DECLARE @myVariable VARCHAR(MAX);

SET @myVariable = 'SQLTwins';

SELECT @myVariable AS myVariable;

0412_02_SSISVariable

User Variable of type “string” in the test package

0412_03_SSISExecuteSQLDetails

Execute SQL task details showing sample T-SQL script

0412_04_SSISResultSetVariableMapping

Variable Mapping in the Execute SQL Task

When we execute this SSIS package, it fails with the error referenced above.

0412_05_ExecuteSQLFailure

Failed Execute SQL Task

0412_06_ExecuteSQLFailureDetails

Execute SQL Task Failure Details

The Solution

The solution was right there in our faces, but we failed to notice it for a while. If we read the error message carefully, we can isolate the following points:

  • The data-type of the variable from the Result Set output of the Execute SQL task is different from the data-type of the target user variable
  • SSIS detects this as an attempt to change the data-type, which is not allowed because variables types are strict unless defined as an “object”

Based on this, we set about looking at differences between the single-row result set and the SSIS user variable of type “string”. We soon realized that the result set was returning a VARCHAR(MAX).

It appears that the (MAX) was causing problems in the SSIS engine. As soon as we changed it to a fixed-length variable the package worked as expected.

DECLARE @myVariable VARCHAR(8000);

SET @myVariable = 'SQLTwins';

SELECT @myVariable AS myVariable;

0412_07_ExecuteSQLSuccess

Successful execution of Execute SQL after changing to a fixed-length data-type

Hope this little tip helps in your development efforts someday.

Until we meet next time,

Be courteous. Drive responsibly.