The SSIS Script Task is a very powerful component to use in a SSIS package, and most of the time you can achieve with it what you want.
Still I have faced a few situations where a Script Task was not the best solution. In those cases I made a .NET Console application and executed it from within a SSIS package.
The full article is posted on SQLShack.com, so you can read further there ..
With a Lookup in a dataflow in a SSIS Package you can achieve a similar result as with a T-SQL Join clause in a SELECT statement: you can check if a row exists in a different dataset, based on one or more (business / natural / surrogate) key columns. However there are a number of differences. One of these differences took me several hours of investigating some time ago, when I had this lookup in SSIS that did not find any row. I did not understand why, because I had a similar T-SQL statement with a JOIN that found all rows. What could be different (it appeared to be a CHAR column connected to a VARCHAR)?
So today I am going to compare the two.
Subjects of comparison:
Simple setup for two lookup transformations using the Adventureworks database
A Transact-SQL query with a join is more forgiving than a SSIS Lookup when different datatypes are joined.
In a SSIS Lookup this does not work because an error is shown during design:
In T-SQL, this join can be made and will work:
In SSIS the lookup can be developed, but will not work in practice, when the CHAR column value is not the maximum length (and thus has trailing spaces).
A “join” of a nvarchar to a nchar column ..
.. will not work !
In T-SQL, this join can be made and will work:
In a SSIS Lookup you can use different connection managers for source and Lookup transformation, so this is easy to do.
In T-SQL this would be more complex to implement, for instance you could use linked servers if your security policy allows this.
An equi-join is a join where the column values that are joined simply must be equal to each other.
By default a SSIS Lookup will do an equi-join. In the visual editor you connect one or more columns from the Available Input Columns to the Available Lookup Columns. Designed in this way it will be an equi-join.
There are two ways I know of to implement a different kind of join:
In T-SQL you can do more complex joins, for instance with a BETWEEN clause for an EffectiveStartDate and EffectiveEndDate.
In a SSIS Lookup you can use different connection managers for source and Lookup transformation, so this is easy to do.
In T-SQL this would be practicly impossible, or at least much more complex to accomplish (e.g. with OPENROWSET).
A SSIS Lookup is case-sensitive when Full cache is used, so differences in casing cause that lookup values are not found.
However when you switch to No Cache or Partial Cache , the lookup can become case-insensitive! Keith Mescha brought this to my attention (see below). Thank you, Keith.
So the safest thing to do is convert the column values for columns to connect from Available Input Columns to Available Lookup Columns to either upper- or lowercase on both sides. In T-SQL you can use the LOWER() and UPPER() functions for this, in SSIS you can add a Derived column to your dataflow, where you also can use the LOWER() and UPPER() functions. Depending on your needs, you can replace the existing column or add the upper or lower value as new column.
Example of a derived column that converts the value of a column of the Available Input Columns to uppercase.
In T-SQL the result of this join will depend on collation settings of the SQL Server Instance or column(s) affected in the join.
By default, SQL Server is installed using a case-insensitive collation (e.g. Latin1_General_CI_AS, where CI stands for Case Insensitive), and when the collation is not overruled on column-level, the join will work case-insensitively.
From my own experience I know SSIS Lookups can be faster when developed correctly. This means:
In this blog post I have compared the SSIS Lookup transformation with the T-SQL Join.
In general T-SQL is more flexible and more forgiving, but when you need to go outside the SQL Server Instance for source or lookup it is easier to use a SSIS Lookup, but beware of the peculiarities of the SSIS Lookup, as described in this article.
Performance-wise, a SSIS Lookup can perform better than a T-SQL join, but this of course depends on a lot of design factors in both your Lookup as well as SQL database.
(c) 2017 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
I have made a compilation of all my blog posts of the past year, and it was more work than you would think.
Especially because it also contains a non-published new article.
After filling in the form you can download the e-book for free.
I promise you the following:
(c) 2016 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
| Date | Changes |
|---|---|
| 27 August, 2017 | Alan brought to my attention that the sensitive value might get exposed in the log. And yes, he was right. I found out that the Raise Change Event (value True) caused this, so I put it to False. Then everything still worked and the sensitive value was not exposed via the log. So the changed advice is to leave Raise Change Event on False for the variable containing the plain value of the sensitive parameter. |
This time I want to tell something about the use of sensitive parameters. Passwords, connection strings, anything that is .. uhm .. sensitive.
SSIS Package Parameters and Environment variables can hold sensitive values, like passwords, connection strings and so on. You do not want these to be seen by people who are not authorized to have access to this information.
First I want to give a *short* intro on how (not) to save sensitive parameters in SSIS.
There are several options which are explained enough on other internet sites:
My opinion on those options:
So the rest of this article is based on using the DontSaveSensitive property.
You have to set it in the SSIS package, but also in the project properties:
Setting the package property “ProtectionLevel”
Setting the Project Protection Level
I have set up a very simple demo. It consists of the following steps:
See the script for this in my previous blog post.
In the demo I use my existing DEV folder with the DEV environment inside that folder.
In my case I use Windows Integrated Security, therefore the connection string on itself is not sensitive. I’ll just use this example, because a connection string could also contain a username and password.
The environment variable “MyDBConnectionString”. Check the checkbox to make it sensitive
Do I need to say more?
Not so exiting either, except that you should set Raise Change Event to True for the variable, so that any other properties set by an expression using the variable will certainly change when the variable changes.
Before you can do that, you must make sure that the column is visible. The printscreens below explain how.
Clicking on this icon shows the Variable Grid Options
On the Variable Grid Options check “Raise event when variable value changes”
For the variable set “Raise Change Event” to True
| Don’t set “Raise Change Event” to True! |
|---|
| See change log at the top of the article for details. |
Here you can see that the Expression @[User::MyDBConnectionString] is used to set the ConnectionString property of the OLEDB Connection Manager OLE_SRC_MyDatabase.
The “problem” is that you cannot use the sensitive parameter in an expression that uses the value for a non-sensitive property, in my case the ConnectionString property of an OLEDB Connection Manager.
This error is shown:
This is why you have to copy the value of the sensitive parameter to a non-sensitive variable first using a Script Task.
I personally think it is a pity that Microsoft did not create a GETSENSITIVEVALUE(« SensitiveParameterOrVariable ») expression, so you could use an Expression Task, but they didn’t, so a script task is the way to go.
Add a Script Task to the control flow of your SSIS Package and adjust public void Main() as below:
The control flow of the demo package now looks as follows:
The final control flow of the demo package
To demonstrate how it works, the “design time” connection string, that is stored in the variable MyDBConnectionString, is different from the sensitive environment variable MyDBConnectionString.
I have added an Execute SQL Task SQL_FooBar that uses the OLEDB Connection Manager OLE_SRC_MyDatabase.
The connection string in the variable connects to a database named FF (which is Dutch SMS talk for “Just a moment”/temporary).
The connection string in the sensitive environment variable connects to tempdb.
To proof that the connection string from the sensitive environment variable (which passes it to the package parameter) is used during package execution, I take the FF database offline, after the OLEDB Connection Manager is configured.
Taking database FF offline ..
So when the connection manager would not be configured from the sensitive parameter during run time, the Execute SQL Task would certainly fail during execution.
Now it’s time to deploy my project with the testpackage to my DEV folder in SSISDB. I do not have screenshots of that, because I assume that you will know how to do this. If not, check my previous blog posts.
This time I do the configuration manually: in the project I add a reference to the DEV environment (not in the picture), then I connect the environment variable to the package parameter.
After it is configured, I execute the package.
Shortly after it has succeeded, proof for me that it was using the configured connectionstring with database tempdb, and not the design time value with database FF.
If you want to debug the SSIS package in Visual Studio you have a problem: due to DontSaveSensitive the package parameter will have no value, and you cannot debug the package.
My workaround for this is to put the debug value for the sensitive value in the SSIS variable. In the Script Task the variable will only get the value from the parameter when it is not empty. So during debug time, the variable value will be preserved.
If your value is too sensitive for this, either use a less sensitive value referring to a test database/server etc., or remove the value from the variable after debugging the package. In the latter case, also delete copies of the package in the bin folder.
Short answer: yes, I think pretty safe.
Why? If a DBA unticks the “Sensitive” checkbox, the value is wiped out and cannot be read.
Of course a DBA or hacker with sysadmin access to the SQL Instance and knowledge could deploy a SSIS Package, just to pull out the sensitive value. But there are little things a DBA cannot do..
In this article you could read some practical insights on using the ProtectionLevel DontSaveSensitive for a SSIS package containing a sensitive parameter configured from a sensitive environment variable during execution.
Steps to take are:
(c) 2016/2017 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
Hi there, today I want to share my scripts on package deployment. As manual deployment can be cumbersome and prone to errors, I have automated this boring repetitive task.
This article is based on SQL Server 2016, but I have included scripts to be used for SQL Server 2012 and 2014.
How to automate SSIS Package deployment.
As you could read in one of my previous posts Using the project deployment model with environments I usually create a one-on-one relation between folders and environments.
So a folder = an environment, and within the folder several SSIS projects can exist.
What I have seen others do is creating a “Generic”environment. That is also a possibility, this is a matter of personal choice. I prefer having the environment having an explicit name identical to the foldername (e.g. DEV), just to decrease the risk that any changes are made in the wrong Generic environment variable values.
Our main goal for this step is to have a way to automate setting up a folder and environment, including all (scripted) environment variables. This script should be rerunable to be able to add or update environment variables when needed.
One thing you should know is that the database SSISDB has a number of stored procecures in the catalog schema, that I use to achieve this.
Well, here’s the script. Don’t bother copy and paste while you can download all scripts here.
--\ -------------------------------------------------------------------
---) Script to create folder, environment and settings.
--/ -------------------------------------------------------------------
/*
(c) Copyright 2016 - hansmichiels.com
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see http://www.gnu.org/licenses/.
*/
USE [SSISDB]
GO
DECLARE @environment_variables_table TABLE(
[id] int identity(1, 1),
--)> Just an autonumber to be able to go through the table without cursor.
[variable_name] nvarchar(128),
--) Name of the variable
[data_type] nvarchar(128),
--) Variable datatype e.g. [ Boolean | Byte | DateTime | Decimal | Double | Int16 | Int32 | Int64 | SByte | Single | String | UInt32 | UInt64 ]
--) Check all possible values using this query: SELECT DISTINCT [ssis_data_type] FROM [SSISDB].[internal].[data_type_mapping]
[sensitive] bit,
--) Indication of the environment variable is sensitive (e.g. a password).
[value] nvarchar(4000),
--) The variable value.
[description] nvarchar(1024),
--) Extra description for the variable.
[create_only] bit
--) Indication that a variable should only be created when it does not exist yet, but should not be replaced when it already exists.
--) (so the value in the script does not overwrite the existing value).
--) In this way you can prevent later configuration changes in an environment being undone by the values in the script.
)
DECLARE @id int
DECLARE @max_id int
DECLARE @folder_id int
DECLARE @folder_name sysname = N'unknown'
DECLARE @environment_name nvarchar(128) = N'unknown'
DECLARE @environment_description nvarchar(128) = N'unknown'
DECLARE @variable_name nvarchar(128)
DECLARE @data_type nvarchar(128)
DECLARE @sensitive bit
DECLARE @value sql_variant
DECLARE @description nvarchar(1024)
DECLARE @create_only bit --
DECLARE @exists bit
DECLARE @nsql nvarchar(max)
DECLARE @message nvarchar(255)
DECLARE @false BIT = 0, @true BIT = 1
--\
---) Environment settings.
---) This is a generic script for all environments, using the @@SERVERNAME (SQL Server Instance name)
---) to set values right for each environment.
--/
IF @@SERVERNAME LIKE 'VWS63-SQL1%1'
BEGIN
SELECT
@folder_name = N'DEV',
@environment_name = N'DEV',
@environment_description = N'My development environment'
END
IF @@SERVERNAME = '<SQL INSTANCE NAME>'
BEGIN
SELECT
@folder_name = N'TST',
@environment_name = N'TST',
@environment_description = N'My test environment'
END
IF @@SERVERNAME = '<SQL INSTANCE NAME>'
BEGIN
SELECT
@folder_name = N'PRD',
@environment_name = N'PRD',
@environment_description = N'My production environment'
END
IF @environment_name = N'unknown'
BEGIN
RAISERROR('Unknown environment', 16, 1)
-- Make sure that the script does not continue when the environment settings could not be set due to an unknown @@SERVERNAME.
GOTO Finally
END
----\
-----) Create folder, when needed.
----/
IF NOT EXISTS (
SELECT 1
FROM [SSISDB].[internal].[folders] f
WHERE f.name = @folder_name
)
BEGIN
SET @message = 'Folder “' + @folder_name + '” is being created.'
RAISERROR(@message , 0, 1) WITH NOWAIT;
EXEC [SSISDB].[catalog].[create_folder] @folder_name, @folder_id OUTPUT;
END
--\
---) Create environment, when needed.
--/
IF NOT EXISTS (
SELECT 1
FROM [SSISDB].[internal].[environments] e
JOIN [SSISDB].[internal].[folders] f
ON e.folder_id = f.folder_id
WHERE e.[environment_name] = @environment_name
AND f.name = @folder_name
)
BEGIN
SET @message = 'Environment “' + @environment_name + '” is being created.'
RAISERROR(@message , 0, 1) WITH NOWAIT;
EXEC [SSISDB].[catalog].[create_environment] @folder_name, @environment_name, @environment_description;
END
IF @environment_name = N'DEV'
BEGIN
INSERT INTO @environment_variables_table
( [create_only], [variable_name], [data_type], [sensitive], [value], [description] )
VALUES
-- Just a few examples of environment variables, add your own here or generate this part from existing environments or deployed packages, see scripts below.
( @true, N'Parameter1', N'Int32', @false, N'1', N'Example parameter 1 (DEV).' )
, ( @true, N'Parameter2', N'String', @false, N'Hello world from script (DEV).', N'Example parameter 2 (DEV).' )
, ( @true, N'Parameter3', N'Decimal', @false, N'9.95', N'Example parameter 3 (DEV).' )
, ( @true, N'ProjectParameter1', N'String', @false, N'Hello world from script (DEV).', N'Example projectparameter 1 (DEV).' )
, ( @true, N'ProjectParameter2', N'Int32', @false, N'4', N'Example projectparameter 2 (DEV).' )
END
IF @environment_name = N'TST'
BEGIN
INSERT INTO @environment_variables_table
( [create_only], [variable_name], [data_type], [sensitive], [value], [description] )
VALUES
-- Just a few examples of environment variables, add your own here or generate this part from existing environments or deployed packages, see scripts below.
( @true, N'Parameter1', N'Int32', @false, N'1', N'Example parameter 1 (TST).' )
, ( @true, N'Parameter2', N'String', @false, N'Hello world from script (TST).', N'Example parameter 2 (TST).' )
, ( @true, N'Parameter3', N'Decimal', @false, N'9.95', N'Example parameter 3 (TST).' )
, ( @true, N'ProjectParameter1', N'String', @false, N'Hello world from script (TST).', N'Example projectparameter 1 (TST).' )
, ( @true, N'ProjectParameter2', N'Int32', @false, N'4', N'Example projectparameter 2 (TST).' )
END
IF @environment_name = N'PRD'
BEGIN
INSERT INTO @environment_variables_table
( [create_only], [variable_name], [data_type], [sensitive], [value], [description] )
VALUES
-- Just a few examples of environment variables, add your own here or generate this part from existing environments or deployed packages, see scripts below.
( @true, N'Parameter1', N'Int32', @false, N'1', N'Example parameter 1 (PRD).' )
, ( @true, N'Parameter2', N'String', @false, N'Hello world from script (PRD).', N'Example parameter 2 (PRD).' )
, ( @true, N'Parameter3', N'Decimal', @false, N'19.95', N'Example parameter 3 (PRD).' )
, ( @true, N'ProjectParameter1', N'String', @false, N'Hello world from script (PRD).', N'Example projectparameter 1 (PRD).' )
, ( @true, N'ProjectParameter2', N'Int32', @false, N'4', N'Example projectparameter 2 (PRD).' )
END
SELECT @id = 1, @max_id = MAX([id]) FROM @environment_variables_table
WHILE @id <= @max_id
BEGIN
SELECT
@variable_name = v.variable_name,
@data_type = v.data_type,
@sensitive = v.sensitive,
@value = v.value,
@description = v.[description],
@create_only = v.[create_only],
@exists = 0
FROM @environment_variables_table v
WHERE [id] = @id;
IF EXISTS (
SELECT 1
FROM
[SSISDB].[internal].[environment_variables] v
JOIN
[SSISDB].[internal].[environments] e ON e.environment_id = v.environment_id
JOIN
[SSISDB].[internal].[folders] f
ON e.folder_id = f.folder_id
WHERE
v.[name] = @variable_name
AND e.environment_name = @environment_name
AND f.name = @folder_name
)
BEGIN
IF @create_only = 1
BEGIN
SET @message = @variable_name + ' already exists and is not replaced.'
RAISERROR(@message , 0, 1) WITH NOWAIT;
SET @exists = 1;
END ELSE BEGIN
SET @nsql = N'EXECUTE [catalog].[delete_environment_variable] '
+ N'@folder_name = N'''+ @folder_name + ''', @environment_name = N''' + @environment_name + ''', '
+ N'@variable_name = N''' + @variable_name + ''''
PRINT @nsql;
EXEC sp_executesql @nsql;
END
END
IF @exists = 0
BEGIN
SET @message = 'Creating variable “' + @variable_name + '”.';
RAISERROR(@message , 0, 1) WITH NOWAIT;
SET @nsql = N'EXECUTE [catalog].[create_environment_variable] '
+ N'@folder_name = N'''+ @folder_name + ''', @environment_name = N''' + @environment_name + ''', '
+ N'@variable_name = N'''+ @variable_name + ''', @data_type = N''' + @data_type + ''', '
+ N'@sensitive = ' + CONVERT(NVARCHAR, @sensitive) + ', @description = N''' + @description + ''', '
+ CHAR(13) + CHAR(10) + N'@value = ' +
CASE UPPER(@data_type)
WHEN 'String' THEN 'N''' + CONVERT(NVARCHAR(1000), @value) + ''' '
ELSE CONVERT(NVARCHAR(1000), @value)
END + '; '
-- PRINT @nsql;
EXEC sp_executesql @nsql;
END
SET @id = @id + 1
END
RAISERROR('Creating folder, environment and settings has completed ..', 0, 1) WITH NOWAIT;
Finally:
GO
--\
---) ON a server where the VARIABLES already exist, you can use this query to generate the 'VALUES' part of the INSERT command used above.
--/
/*
DECLARE @environment_name nvarchar(128) = N'DEV'
SELECT
m.*,
v.name,
N', ( N''' + v.name + ''', N''' + m.[ssis_data_type] + ''', '
+ CONVERT(NCHAR(1), v.sensitive) + ', N'''
+ CASE v.sensitive
WHEN 0 THEN CONVERT(NVARCHAR(4000), v.value)
ELSE '< type sensitive value here TODO >' END
+ ''', N''' + v.description + ''' )' AS [sql_values_part]
FROM
[SSISDB].[internal].[environment_variables] v
JOIN
[SSISDB].[internal].[environments] e
ON e.environment_id = v.environment_id
JOIN
[SSISDB].[internal].[folders] f
ON e.folder_id = f.folder_id
JOIN
[SSISDB].[internal].[data_type_mapping] m
ON m.sql_data_type = v.base_data_type
AND ( m.sql_data_type != 'tinyint' OR m.ssis_data_type = 'Byte')
WHERE
e.environment_name = @environment_name
*/
--\
---) ON a server where the PACKAGES are ALREADY DEPLOYED, you can use this query to generate the 'VALUES' part of the INSERT command used above.
---) Manually check the output before running the sql script!
--/
/*
SELECT
N', ( N''' + q.[variable_name] + ''', N''' + q.[data_type] + ''', '
+ CONVERT(NCHAR(1), q.[sensitive]) + ', N'''
+ CASE q.sensitive
WHEN 0 THEN CONVERT(NVARCHAR(4000), q.value)
ELSE '< type sensitive value here TODO >' END
+ ''', N''' + q.[description] + ''' )' AS [sql_values_part]
FROM
(
SELECT TOP 2147483647
prm.[parameter_name] AS [variable_name],
prm.[parameter_data_type] AS [data_type],r
prm.[sensitive],
COALESCE(prm.[default_value], prm.[design_default_value]) AS [value],
prm.[description],
ROW_NUMBER() OVER (PARTITION BY prm.[parameter_name] ORDER BY prm.[parameter_name]) AS [Occurrence]
FROM
[SSISDB].[internal].[object_parameters] prm
-- [variable_name], [data_type], [sensitive], [value], [description]
WHERE
prm.[value_type] = 'V'
AND prm.[parameter_name] NOT LIKE 'CM.%'
AND LEFT(prm.[parameter_name], 1) != '_'
AND NOT ( prm.[object_type] = 30 AND LEFT(prm.[object_name], 1) = '_') -- Naming convention for internally used SSIS Packages: start with _
ORDER BY
prm.[parameter_name]
) q
WHERE
q.Occurrence = 1
*/
GO
Example output of this script, when executed using Management Studio.
And the second time, when a parameter is added and “create_only” is true for the existing parameters.
This is about how to deploy an .ispac file in an automated way.
An .ispac file is a zipfile that contains all packages in a SSIS project that can be deployed all at once. You create it when you chose Build or Rebuild in Visual Studio when you have the SSIS project open.
By the way, if you rename it and give it a .zip extension, you can simply look what is inside.
Amongst the files you can download here there are demo SSIS projects that I used for this on SQL Server 2012, 2014 and 2016.
The demo SSIS project ..
Menu-option to rebuild the SSIS solution ..
After rebuilding the solution from the menu you will find the ispac files in the bin\<configuration name> folder of each project. In the screenprints above you can see how the bin subfolder is related to the chosen configuration in Visual Studio.
To install an ispac file, I use the command line options of the Integration Services Deployment Wizard.
For a default installation you can find the Integration Services deployment wizard (ISDeploymentWizard.exe) here:
| Version | Path |
|---|---|
| SQL Server 2012 | C:\Program Files\SQL Server\110\DTS\Binn |
| SQL Server 2014 | C:\Program Files\SQL Server\120\DTS\Binn |
| SQL Server 2016 | C:\Program Files\SQL Server\130\DTS\Binn |
You will find the exact command line in one of the following paragraphs, where all steps are glued together.
On 64-bit machines there is also a 32-bit version of this tool, but to be honest I do not see a reason why you would want to use the 32-bit version instead of the 64-bit, therefore I did not list the paths here.
After installation you can find the folder, project and environment here ..
After the packages/SSIS projects are deployed, the next thing to do is adding an environment reference to the project(s) and connecting environment variables to package parameters and project parameters, based on the name.
I use this script for that:
--\ -------------------------------------------------------------------
---) Script to add environment reference to project(s) and to connect
---) environment variables to package parameters, based on the name
--/ -------------------------------------------------------------------
/*
(c) Copyright 2016 - hansmichiels.com
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see http://www.gnu.org/licenses/.
*/
USE [SSISDB]
GO
DECLARE @object_parameter_value_table TABLE(
[id] int identity(1, 1),
[object_type] smallint,
[object_name] nvarchar(260),
[parameter_name] nvarchar(128),
[project_name] NVARCHAR(128)
)
DECLARE @project_names_table TABLE(
[id] int identity(1, 1),
[project_name] NVARCHAR(128)
)
DECLARE @message nvarchar(255)
DECLARE @id int
DECLARE @max_id int
DECLARE @folder_name sysname = N'unknown'
DECLARE @environment_name nvarchar(128) = N'unknown'
DECLARE @environment_description nvarchar(128) = N'unknown'
DECLARE @project_name NVARCHAR(128)
DECLARE @object_type SMALLINT
DECLARE @object_name NVARCHAR(260)
DECLARE @value_type CHAR(1)
DECLARE @parameter_name NVARCHAR(128)
DECLARE @parameter_value SQL_VARIANT
DECLARE @reference_id BIGINT
--\
---) Environment settings
--/
IF @@SERVERNAME LIKE 'VWS63-SQL1%1'
BEGIN
SELECT
@folder_name = N'DEV',
@environment_name = N'DEV',
@environment_description = N'My development environment'
END
IF @@SERVERNAME = '<SQL INSTANCE NAME>'
BEGIN
SELECT
@folder_name = N'TST',
@environment_name = N'TST',
@environment_description = N'My test environment'
END
IF @@SERVERNAME = '<SQL INSTANCE NAME>'
BEGIN
SELECT
@folder_name = N'PRD',
@environment_name = N'PRD',
@environment_description = N'My production environment'
END
IF @folder_name = N'unknown'
BEGIN
RAISERROR('Unknown environment', 16, 1)
-- Make sure that the script does not continue when the environment settings could not be set due to an unknown @@SERVERNAME.
GOTO Finally
END
INSERT INTO @project_names_table([project_name])
SELECT
p.name
FROM
[SSISDB].[internal].[folders] f
JOIN
[SSISDB].[internal].[environments] e
ON e.folder_id = f.folder_id
JOIN
[SSISDB].[internal].[projects] p
ON p.folder_id = f.folder_id
WHERE
f.name = @folder_name
AND e.environment_name = @environment_name
-- Optional project_name filter here:
-- AND p.name IN ('Project_1', 'Project_2')
--\
---) Add environment reference to project(s).
--/
SELECT @id = 1, @max_id = MAX([id]) FROM @project_names_table
WHILE @id <= @max_id
BEGIN
SELECT
@project_name = v.[project_name]
FROM @project_names_table v
WHERE [id] = @id;
IF NOT EXISTS( SELECT 1
FROM
[SSISDB].[internal].[folders] f
JOIN
[SSISDB].[internal].[environments] e
ON e.folder_id = f.folder_id
JOIN
[SSISDB].[internal].[projects] p
ON p.folder_id = f.folder_id
JOIN
[SSISDB].[internal].[environment_references] r
ON r.environment_name = e.environment_name
AND p.project_id = r.project_id
WHERE
f.name = @folder_name
AND e.environment_name = @environment_name
AND p.name = @project_name
)
BEGIN
SET @message = 'An environment reference for project “' + @project_name + '” is being created.'
RAISERROR(@message , 0, 1) WITH NOWAIT;
EXEC [SSISDB].[catalog].[create_environment_reference]
@environment_name=@environment_name, @reference_id=@reference_id OUTPUT,
@project_name=@project_name, @folder_name=@folder_name, @reference_type='R';
--Select @reference_id
END
SET @id = @id + 1
END
--\
---) Connect environment variables to package parameters, based on the name
--/
INSERT INTO @object_parameter_value_table (
[object_type],
[object_name],
[parameter_name],
[project_name]
)
SELECT
prm.[object_type],
prm.[object_name],
prm.[parameter_name],
prj.name
FROM
[SSISDB].[internal].[folders] f
JOIN
[SSISDB].[internal].[environments] e
ON e.folder_id = f.folder_id
JOIN
[SSISDB].[internal].[environment_variables] ev
ON e.environment_id = ev.environment_id
JOIN
[SSISDB].[internal].[projects] prj
ON prj.folder_id = f.folder_id
JOIN
@project_names_table prjsel
ON prjsel.project_name = prj.name
JOIN
[SSISDB].[internal].[object_parameters] prm
ON prj.project_id = prm.project_id
AND prm.parameter_name = ev.name
WHERE
prm.[value_type] != 'R'
AND prm.value_set = 0
AND prm.[parameter_name] NOT LIKE 'CM.%'
AND LEFT(prm.[parameter_name], 1) != '_' -- Naming convention for internally used parameters: start with _
AND NOT ( prm.[object_type] = 30 AND LEFT(prm.[object_name], 1) = '_') -- Naming convention for internally used SSIS Packages: start with _
AND f.name = @folder_name
AND e.environment_name = @environment_name
ORDER BY
prm.object_name, prm.parameter_name
SELECT @id = 1, @max_id = MAX([id]) FROM @object_parameter_value_table
WHILE @id <= @max_id
BEGIN
SELECT
@object_type = v.[object_type],
@object_name = v.[object_name],
@parameter_name = v.[parameter_name],
@project_name = v.[project_name]
FROM
@object_parameter_value_table v
WHERE
[id] = @id;
SELECT @value_type = 'R', @parameter_value = @parameter_name;
SET @message = 'Parameter “' + @parameter_name + '” (of object “' + @object_name + '”) is mapped to environment variable.'
RAISERROR(@message , 0, 1) WITH NOWAIT;
EXEC [SSISDB].[catalog].[set_object_parameter_value]
@object_type, @folder_name, @project_name, @parameter_name,
@parameter_value, @object_name, @value_type
SET @id = @id + 1
END
RAISERROR('Adding environment reference to project(s) and mapping environment variables to package parameters has completed ..', 0, 1) WITH NOWAIT;
Finally:
GO
Also important is that I use a naming convention to decide whether or not to create a link between a package parameter and an environment variable.
For this I borrowed a naming convention from c# coding, where a variable starting with an underscore is a private variable within a class.
So what I have done is the following:
For the demo I use a solution with two projects, one of them with three packages, and the second one with just one package. This is done to demonstrate how connecting package- and project parameters to environment variables (step 3) works.
Example output windows when this script is run after having deployed the ispac files.
And the result: the environment is “selectable” and the parameters show that they are mapped to an environment variable by having an underlined environment variable name in the “Value” column.
Alright, what do we have so far?
So our PowerShell script simply needs to execute a sql script, then deploy the packages, and then execute another sql script. That shouldn’t be too difficult.
This is the script that does exact those steps. If I have some time left I want to split the configuration and the actual script over different files/scripts.
######################################################################
# CONFIGURATION
######################################################################
# Configure SQL instance name and (SSIS) foldername here.
$SQLInstance = “VWS63-SQL161”
$FolderName = “DEV”
# SQL Server 2016 default location:
$ISDeploymentWizard = “C:\Program Files\Microsoft SQL Server\130\DTS\Binn\ISDeploymentWizard.exe”
# My location:
# $ISDeploymentWizard = “C:\MSSQL\Shared features\130\DTS\Binn\ISDeploymentWizard.exe”
######################################################################
# MAIN SCRIPT
######################################################################
# \
# 1. Calls pre_deployment sql script
# /
$File = $ScriptPath + “\pre_deployment.sql”
sqlcmd -S $SQLInstance -i $File
# \
# 2. Installs each ispac file found in the folder where the powershell script is located.
# /
Get-ChildItem $ScriptPath -Filter *.ispac |
Foreach-Object {
$ProjectName = $_.BaseName
Write-Host “Deploying project $ProjectName ..”;
& $ISDeploymentWizard /Silent /ModelType:Project /SourcePath:”$ScriptPath\$ProjectName.ispac” /DestinationServer:”$SQLInstance” /DestinationPath:”/SSISDB/$FolderName/$ProjectName” | Out-Null
Write-Host “Deployment of project $ProjectName completed ..”;
}
# \
# 3. Calls post_deployment sql script
# /
$File = $ScriptPath + “\post_deployment.sql”
sqlcmd -S $SQLInstance -i $File
And this is the result of the execution of the Powershell script. From now on no more manual deployments!
Powershell output window ..
The goal of this blog post was to explain to you how to automate the deployment of SSIS packages including the configuration and environment setup.
I have explained how:
You could also read that I use a naming convention to be able to exclude individual package parameters or all parameters of a package from being connected to environment variables.
(c) 2016/2017 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
Important, please read!
Hi, I noticed something weird when upgrading my packages as described below, and thought it would be fair to share this with you.
There seems to be an issue with upgrading packages with Script Tasks in it. The packages I upgraded with just plain SSIS stuff did not have that problem.
The following warnings are given:
Warning 0x40016044: DFT_Export_Person: Found SQL Server Integration Services 2012 Script Component SCR_Add_RowNumber that requires migration!
Warning 0x40016044: SCR_Foo_Bar: Found SQL Server Integration Services 2012 Script Task “ST_a1fbd3cd5d924758a2c98397289876d0” that requires migration!
However, the wizard still says that the upgrade is successful:
The message that the upgrade is successful
I think these warnings should be big fat errors and I will show you why.
Before the upgrade: the script task has a ReadOnlyVariable and a ReadWriteVariable
Before the upgrade: the script task has code in public void Main()
After the upgrade: the script task does not have the ReadOnlyVariable nor the ReadWriteVariable
After the upgrade: the code in public void Main() has disappeared
As you know I am not anti-Microsoft, in the past I have thanked them for being so cool towards developers.
But as far as this upgrade process is concerned I wonder: what must the developer at Microsoft have thought when making this a warning instead of an error? Was it monday morning, was he or she drunk or on drugs? I don’t know, the package is now cripple and will not work. That is not a warning.
So I “warn” you, pay attention to your script tasks, if you use them.
After upgrading, my builds started to take a lot of time (up to 4 minutes, I suspect an on-the-fly upgrade was being done). While investigating this, I found that in the project file all packages are also listed with a PackageFormatVersion per package in it, which was still 6!
I am not sure if this is a Visual Studio bug, if the PackageFormatVersion of the dtsx file says 8, while the dtproj file says 6 for the same package.
Anyway, after I modified the procedure slightly, my builds were fast again (20 seconds instead of 4 minutes).
To achieve this, I opened the SSIS 2012 project using Visual Studio 2015, then chose Upgrade All Packages from the contextmenu.
After that I changed the TargetServerVersion as described below.
Sorry for any inconveniece caused by my initial article, and I hope this update will help.
(The original post starts below)
If you are intending upgrading your SQL Server installations to 2016, you’ll probably want to do the same with your SSIS packages.
And I can tell you, coming from SQL Server 2012 or 2014 this is fairly easy to do.
But with my experience I want to help you doing this not the wrong way.
I made the mistake by using the SSIS Package Upgrade Wizard while the projectversion still was SQL Server 2012.
Funny thing is that the packages seemed to upgrade, however I got all kinds of errors and they just did not work.
While the solution seemed so easy.
For this article I prepared a SSIS project made with SQL Server 2012.
In the picture below you can how it looks like when opened with SQL Server Data Tools 2015 on a Server with SQL Server 2016 installed.
After the project name, between brackets you can read that this is a SSIS 2012 project.
Also, if you open a package in Notepad (be careful not to change it) you can see that the PackageFormatVersion is 6. This means SQL Server 2012.
The simplest way to upgrade all packages in the project is to change the version of the SSIS project! Please be sure to backup your entire project and all SSIS packages first.
To upgrade all packages, right click on the project node (in the picture on SSIS-E09-Packages (SQL Server 2012) and select Properties.
Then select Configuration Properties and change the TargetServerVersion to SQL Server 2016.
Changing the TargetServerVersion ..
Then read the message, follow the instructions and click Yes.
Upgrade instructions and warning
The good news is, that’s basicly it! When you open a package with Notepad again, you can see that it is upgraded to PackageFormatVersion 8. This means SQL Server 2016.
Package which is upgraded to PackageFormatVersion 8 (SQL Server 2016).
In my experience with more complex packages than used for this demo, they all upgraded with no problems. I used all the standard SSIS components. Script Tasks and Script Components did not have to be recompiled.
Still a warning about the Upgrade All Packages menu item, available from the SSIS Packages node. You don’t need it if you simply change the TargetServerVersion of the project as just described. If you use Upgrade All Packages with the wrong TargetServerVersion it seems to work, but the output packages do not work.
The Upgrade All Packages menu-item
When you do it the right way, upgrading SSIS Packages from 2012 to 2016 is easy to do. However I did not test this with 2014, I am quite sure it will work in the exact same way.
As you might have noticed I have changed the frequency of my blog posts, due to several circumstances. I still try to do 2 or 3 a month.
Last updated: 27 November, 2016.
(c) 2016 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
Hi, as announced in my previous post of July, 14, I took a summer break. I was really useful and I enjoyed it.
From now on I will try to write some interesting posts weekly again.
Today I want to explain the project deployment model of SSIS, introduced already in SSIS 2012.
Also you are gonna find a (hopefully useful) script below that creates folders, environments and environment variables, and creates connections from package parameters to environment variables.
But that’s for later, first a little introduction.
What I found very confusing when I started using the project deployment model, already some years ago, is that an environment in this context is not what I expected it to be.
From a software development perspective an environment is an installed and working version of your software, optionally using one or more databases and configuration setttings exclusively for that environment. For instance a Development or a Test environment.
But the project deployment model of SSIS does not agree with that definition.
For the project deployment model an environment is only a set of values for all environment variables, and nothing more.
Can you think of what that means? Suppose you have a server, both used for Development and Test, and you have databases and cubes with a “_DEV” or a “_TST” suffix to their names.
But as far as SSIS is concerned, you can only install one version of an SSIS package! Yeah, you can execute it with a different set of parameter values stored in SSIS environment variables, but that’s it.
You cannot have two different versions of the same package installed on the same instance of SQL Server – at least not using only SSIS Environments.
I found it not so intuitive that you need to use folders inside the Integration Services Catalog to achieve this.
So far the theory, let us try this stuff out together.
I’ll just click an example together, you do not need to follow me, my goal is that you understand the concepts so that you can apply it to your own projects.
By the way, I am using SQL Server 2016, but it should also work in SQL Server 2012 and 2014.
My test setup is very basic:
Playing around with the example above, I can demonstrate the use of folders and environments when the package is deployed using the project deployment model.
The initial SSIS Package contains just one Execute SQL Task, one parameter named FruitName and one OLEDB Connection manager named MyDatabase.connectionmanager.
The connection manager is configured by an expression so that the property ConnectionString gets the value of a project parameter also named ConnectionString as demonstrated in the images below.
The Execute SQL Task
The package parameter
The project parameter
The OLEDB connection manager configured with an expression
Then rebuild the Solution
Once the solution has been rebuilt, you can deploy the package with the Integration Services Deployment Wizard. You can also call this utility from the command line, but explaining that is out of scope for this blog post.
To execute this wizard, you have to locate the .ispac file, it is in a subfolder of your projects bin folder.
If you have created additional solution configurations in your Visual Studio solution, this subfolder might be different from the image below (where the subfolder is Development).
Locating the .ispac file, that was created when the solution was rebuilt.
First step of the wizard
The default deployment model is Project Deployment, the one you need. The other option is Package Deployment.
Select a server name and a path [1]
Select a server name and a path [2]: create a new folder
Select a server name and a path [3]: name the folder after your environment
Select a server name and a path [4]: then click Next
Review and deploy.
Deployment succeeded.
In the Integration Services Catalog you can find your project and package.
You could configure it now, but there are no SSIS environments yet (note: these are a set of environment variable values).
Don’t configure just yet. We will do this later.
Here you could create an environment .. don’t do it now !
--\
---) A script to create a folder and environment in an SSIS Catalog.
---) Author: Hans Michiels
--/
/*
(c) Copyright 2016 - hansmichiels.com
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see http://www.gnu.org/licenses/.
*/
USE [SSISDB]
GO
DECLARE @environment_variables_table TABLE(
[id] int identity(1, 1),
[variable_name] nvarchar(128),
[data_type] nvarchar(128),
[sensitive] bit,
[value] nvarchar(4000),
[description] nvarchar(1024)
)
DECLARE @object_parameter_value_table TABLE(
[id] int identity(1, 1),
[object_type] smallint,
[object_name] nvarchar(260),
[parameter_name] nvarchar(128)
)
DECLARE @id INT
DECLARE @max_id INT
DECLARE @folder_name sysname
DECLARE @environment_name nvarchar(128)
DECLARE @environment_description nvarchar(128)
DECLARE @project_name nvarchar(128)
DECLARE @variable_name nvarchar(128)
DECLARE @data_type nvarchar(128)
DECLARE @sensitive bit
DECLARE @value sql_variant
DECLARE @value_tinyint tinyint
DECLARE @description nvarchar(1024)
DECLARE @nsql nvarchar(max)
DECLARE @object_type smallint
DECLARE @object_name nvarchar(260)
DECLARE @value_type CHAR(1)
DECLARE @parameter_name nvarchar(128)
DECLARE @parameter_value sql_variant
--\
---) Environment settings
--/
SELECT
@folder_name = N'TST',
@environment_name = N'TST',
@environment_description = N'My demo Test environment',
@project_name=N'SSIS_S01E08'
--\
---) Script to create environment and settings.
--/
IF NOT EXISTS (
SELECT 1
FROM [SSISDB].[internal].[environments]
WHERE [environment_name] = @environment_name
)
BEGIN
EXEC [SSISDB].[catalog].[create_environment] @folder_name, @environment_name, @environment_description;
END
INSERT INTO @environment_variables_table
( [variable_name], [data_type], [sensitive], [value], [description] )
VALUES
( N'ConnectionString', N'String', 0, N'Data Source=.;Initial Catalog=SSIS_S01E08_TST;Provider=SQLNCLI11.1;Integrated Security=SSPI;Auto Translate=False;', N'Connection string for the database.' )
, ( N'FruitName', N'String', 0, N'Tamarillo', N'' )
--, ( N'Int32Example', N'Int32', 0, N'1', N'' )
--, ( N'BooleanExample', N'Boolean', 0, N'1', N'' )
--, ( N'ByteExample', N'Byte', 0, N'5', N'' )
SELECT @id = 1, @max_id = MAX([id]) FROM @environment_variables_table
WHILE @id <= @max_id
BEGIN
SELECT
@variable_name = v.variable_name,
@data_type = v.data_type,
@sensitive = v.sensitive,
@value = v.value,
@description = v.[description]
FROM @environment_variables_table v
WHERE [id] = @id;
IF EXISTS (
SELECT 1
FROM [SSISDB].[internal].[environment_variables] v
JOIN [SSISDB].[internal].[environments] e ON e.environment_id = v.environment_id
WHERE v.[name] = @variable_name
AND e.environment_name = @environment_name
)
BEGIN
SET @nsql = N'EXECUTE [catalog].[delete_environment_variable] '
+ N'@folder_name = N'''+ @folder_name + ''', @environment_name = N''' + @environment_name + ''', '
+ N'@variable_name = N''' + @variable_name + ''''
PRINT @nsql;
EXEC sp_executesql @nsql;
END
PRINT '/*'
PRINT @variable_name
PRINT CONVERT(nvarchar(128), SQL_VARIANT_PROPERTY(@value, 'BaseType'));
PRINT '*/'
SET @nsql = N'EXECUTE [catalog].[create_environment_variable] '
+ N'@folder_name = N'''+ @folder_name + ''', @environment_name = N''' + @environment_name + ''', '
+ N'@variable_name = N'''+ @variable_name + ''', @data_type = N''' + @data_type + ''', '
+ N'@sensitive = ' + CONVERT(NVARCHAR, @sensitive) + ', @description = N''' + @description + ''', '
+ CHAR(13) + CHAR(10) + N'@value = ' +
CASE UPPER(@data_type)
WHEN 'String' THEN 'N''' + CONVERT(NVARCHAR(1000), @value) + ''' '
ELSE CONVERT(NVARCHAR(1000), @value)
END + '; '
PRINT @nsql;
EXEC sp_executesql @nsql;
SET @id = @id + 1
END
--\
---) Add environment reference to project.
--/
Declare @reference_id bigint
IF NOT EXISTS(SELECT 1
FROM [SSISDB].[internal].[environment_references] r
JOIN [SSISDB].[internal].[projects] p
ON p.project_id = r.project_id
WHERE p.name = @project_name
AND r.environment_name = @environment_name
)
BEGIN
EXEC [SSISDB].[catalog].[create_environment_reference]
@environment_name=@environment_name, @reference_id=@reference_id OUTPUT,
@project_name=@project_name, @folder_name=@folder_name, @reference_type='R';
--Select @reference_id
END
--\
---) Connect environment variables to package parameters, based on the name
--/
INSERT INTO @object_parameter_value_table (
[object_type],
[object_name],
[parameter_name]
)
SELECT
prm.[object_type],
prm.[object_name],
prm.[parameter_name]
FROM
[SSISDB].[internal].[object_parameters] prm
JOIN
[SSISDB].[internal].[environment_variables] ev
ON ev.name = prm.parameter_name
JOIN
[SSISDB].[internal].[projects] prj
ON prj.project_id = prm.project_id
WHERE
prm.[value_type] != 'R'
AND
prm.value_set = 0
AND
prj.name = @project_name
ORDER BY
prm.object_name, prm.parameter_name
SELECT @id = 1, @max_id = MAX([id]) FROM @object_parameter_value_table
WHILE @id <= @max_id
BEGIN
SELECT
@object_type = v.[object_type],
@object_name = v.[object_name],
@parameter_name = v.[parameter_name]
FROM @object_parameter_value_table v
WHERE [id] = @id;
SELECT @value_type = 'R', @parameter_value = @parameter_name;
EXEC [SSISDB].[catalog].[set_object_parameter_value]
@object_type, @folder_name, @project_name, @parameter_name,
@parameter_value, @object_name, @value_type
SET @id = @id + 1
END
GO
My script that facilitates to set up multiple environments quickly (download all the used scripts here).
In the demo it would not be a problem to do all the configuration by hand, but if you have a lot of SSIS Packages with a lot of parameters, you really do not want to do this, especially not if you have different environments, like Development, Test and Production.
This is why I made a script that automates this work for me:
We run the script for the TST environment.
The script was executed successfully.
After the script has run, a TST folder with an TST environment inside exists. You can view or edit its properties.
When the Variables pane is selected on the left, the environment variables, created by the script, are shown.
View package configuration.
The package parameter FruitName is now connected to environment variable FruitName.
When I choose Execute ..
… all I have to do is check Environment, it will automaticly show .\TST
After execution of the package, a row exists in the table (and one from a previous test)
We can also execute the package with a SQL Agent Job ..
Execute the package with a SQL Agent Job: all I have to do is check Environment, it will automaticly show .\TST ..
.. and all values are connected properly ..
And yes, it works!
I must admit, I cheated a little. I skipped the part of deploying the first version of the SSIS Package to the DEV environment.
So now I am going to deploy to a new DEV Folder and Environment for the first time, while it is will be the second version of the package. This is not a recommended practice in real projects.
Anyway, the package gets an extra parameter FruitDescription.
Added FruitDescription
Modify the Execute SQL Task to insert also the value for @FruitDescription [1]
Modify the Execute SQL Task to insert also the value for @FruitDescription [2]
Then Rebuild
Then run deployment wizard again, but now create a DEV folder
Script modifications [1]: use a DEV folder and a DEV environment name.
Add FruitDescription to the list of environment variables and change the values: connection string must use the DEV database and fruitname and -description something different than in the TST environment.
Then run the script.
The package was deployed using the deployment wizard.
SSIS Environment was created using the script
Environment variables where added using the script
Environment variables were connected to package parameters using the script
Package execution test: note the new parameter FruitDescription
Test result
Final result: different versions of the same SSIS Package installed side by side on the same machine. But in addition to SSIS Environments with variable values, folders inside the Integration Services Catalog are used for that!
With the project deployment model of SSIS, introduced in SQL Server 2012, deployment and configuration of SSIS Packages has become easier.
However it is important to understand how folders and environments in the Integration Services Catalog work. I have tried to explain this, and provided a script to set up multiple environments quickly.
D=Development
T=Test
😉
Read more here:
Durian
Tamarillo
(c) 2016 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
This post is a follow-up on two older posts and will be the last one before my summer break.
Today I want to enhance the SSIS package made in post Fixing corrupt csv files in a SSIS Package (SSIS Series) by using the Plug and Play Logging Solution.
As long as a script runs well, you might not need logging inside. BUT .. when something GOES wrong, having logging in the script can reduce the time needed to troubleshoot the issue.
In the example for fixing csv files, anything can be wrong with a file, causing an error in the script. In that case it is for instance useful to know which file was being processed when the error occurred.
For this to work, you have to install my Plug and Play Logging Solution, which can be downloaded from this post. It is just one SQL Script you have to run, that creates a [logdb] database with objects inside. You can also add those objects to a different database that already exists.
It is however not mandatory to rebuild the Fixing corrupt csv files in a SSIS Package example, as you can also add logging to your own Script Task.
This post is a enhancement on the Fixing corrupt csv files in a SSIS Package post, so you might need to read that post first, if you did not do that already.
Therefore I will dive directly into the changes that are needed to add logging in the Script Task.
This will be done with lots of screenprints with some comment in between.
Adding a package parameter “LoggingConnectionString”
First you need to add a package parameter LoggingConnectionString that can be used in an expression of the OLE DB Connection and as input variable for the Script Task.
OLE DB Connection configuration ..
Then add an OLE DB Connection for the logging database.
OLE DB Connection properties ..
Use an expression to let the ConnectionString of the OLE DB Connection be equal to the value of the package parameter LoggingConnectionString.
OLE DB Connection – Expression configuration [1]
OLE DB Connection – Expression configuration [2]
Final result
By the fx icon you can see that the connection manager uses an expression.
Enable logging
Now enable logging for the package.
Enable logging [2]
Add a log provider for SQL Server and let it use the logdb.oledbConnection by selecting it under the Configuration column header.
Enable logging [3]
Then select all events. Filtering on what is actually logged is done by the logging solution (by the value of @MaxMessageClass, see this blog post for more details).
Select the Script Task
Select the Script Task and add the following Variables to ReadOnlyVariables:
The added ReadOnlyVariables in the red rectangles
Below you will find a number of screenprints of the script task to talk you through the changes.
You can download the C# script here.
Overview
First make sure the Namespaces region is as shown.
Then fold the namespace with the guid in the name, and paste the entire namespace HansMichielsCom.PlugAndPlaySSISLoggingSolution underneath it.
This code is put in a separate namespace, so that it could also be placed in a .NET assembly that is added to the GAC (Global Assembly Cache). When you would do this, you do not have to add the code to every Script Task.
For the example of today, we just put this namespace inside the Script Task to make things not too complicated for now.
Using the HansMichielsCom.PlugAndPlaySSISLoggingSolution namespace
As a result, you have to tell the first guid-like namespace, that you want to call code inside the second namespace. Therefore add the using statement as shown above.
Constant used for logging
Below you will see some printscreens with changed parts in the script.
Method GetSsisLogWriter to instantiate a SsisLogWriter object
Method Main is extended with logging.
Pass the logWriter as parameter to other methods ..
IMPORTANT: Bugfix in CheckAndReturnHeader!
IMPORTANT: Bugfix in CheckAndReturnHeader!
(header == null) is added to cope with empty files.
Test preparations [1]
Test preparations [2]
Test execution
Test result: logging rows done inside the script are in the log table.
In this post I have demonstrated how to implement logging in SSIS Script Tasks using my Plug and Play Logging Solution.
This type of logging gives more control on what to log and how to log it than when you implement logging using SSIS events.
The examples given are very basic. You can use your imagination to implement logging of errors using a try .. catch block, or use all available parameters of logWriter.AddLogEntry to change the Retention Class, Message Class, and so on.
In the summer I will take some time for study, reflection, holiday, and still .. work.
My next post will be early September at the latest, maybe earlier.
(c) 2016 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.
This article wasn’t planned. SSIS again? No I wanted to write a post about something else.
Nevertheless I got this question on one of my previous posts and thought, I might as well create a blog post about it.
“Hi Hans,
I have created my ssis package for flat file destination. My problem is that I want to add header and footer in my flat file destination to send it to FTP location format will be
0|MMDDYYYY|HHMMSS
1|Row1
1|Row2
1|Row3
1|ROW4
9|6
The last line contains no of row count including header and trailer
Please let me know how to do that.. I am completely stuck with that and no good luck.. “
So to summarize, both the header and the footer line of the csv file need to contain dynamic content.
Also each row starts with a ‘row type’: 0 for the header, 1 for every detail row, and 9 for the footer.
As an example I just use a simple Data Flow Task with an OLEDB Source connected to table DimCurrency from the AdventureWorksDW database.
The data then flows to a Script Component used as a Destination.
Here are a few printscreens to demonstrate what I mean. The Script Component will create the file.
The package is made using SSIS 2014.
The demo package just contains one Data Flow Task
Design of the Data Flow Task
OLEDB Source: DimCurrency from the AdventureWorksDW database
OLEDB Source: select all three columns
When you drag the Script Component to the Data Flow, you will get this dialog. Use the Script Component as a Destination
Variables that the package contains.
In the Script Transformation Editor, add both variables User::HeaderTimestamp and User::TargetCsvPathAndFileName to ReadOnlyVariables
Now comes the most interesting part! We use the endless possibilities of C# to fix our problem.
In the Script Transformation Editor window, press Edit Script ….
Please note that you can download the entire script here, so you do not have to copy the (sometimes malformatted) code from this webpage..
Then in the namespaces region add two namespaces, as indicated below.
Above public override void PreExecute() add the copyright notice and the following two regions: Configuration and Private variables.
Please note that these regions do not exist yet, so you have to copy the code including the #region and #endregion lines.
// Here you can change the column delimiter and the text qualifier easily.
private const string _delimiter = “|”;
private const string _textQualifier = “”;
// To set the text qualifier to ” use:
// private const string _textQualifier = “\””; // I recommend to use a text qualifier.
// Every x rows, write the buffer to the file, and clear the buffer.
// This is faster than writing every individual line to the file.
// The higher the value of _bufferRows is, the more memory will be used and the less write actions are done to the csv file.
// A higher value is a good idea if you have enough memory in your server.
private const int _bufferRows = 100000;
#endregion ***** Configuration *****
#region Private variables
// Variable to hold the path and filename
private string _fileName;
// Variable to count the rows
private int _rowCount;
// StringBuilder to hold a buffer of text that still needs to be written to the csv file.
private StringBuilder _fileContentsBuffer;
#endregion Private variables
PreExecute() is called before the data starts to flow through. Here we do some preparation and compose the headerline of the file.
&nbs p;/// <summary>
/// This method is called once, before rows begin to be processed in the data flow.
/// </summary>
public override void PreExecute()
{
base.PreExecute();
// Initialize variables
_rowCount = 0;
_fileContentsBuffer = new StringBuilder();
_fileName = Variables.TargetCsvPathAndFileName;
DateTime headerTimestamp = Variables.HeaderTimestamp;
// Append the new line to the buffer.
// This will be written to file with the first detail rows.
// If there are no detail rows, no file will be created.
_fileContentsBuffer.AppendFormat(“{1}0{1}{0}{1}{2:ddMMyyyy}{1}{0}{1}{2:HHmmss}{1}\r\n”, _delimiter, _textQualifier, headerTimestamp);
}
PostExecute() is called after the data has flowed through. Here we append the last lines in the _fileContentsBuffer variable and the footerline to the file.
using (StreamWriter sw = File.AppendText(_fileName))
{
// Append a footer line with the rowcount, example value: 9|105
// The rowcount is a count of the detail rows, but the footer count
// needs to include header and footer, so we add 2 to the value of _rowCount.
_fileContentsBuffer.AppendFormat(“{1}9{1}{0}{1}{2}{1}”, _delimiter, _textQualifier, _rowCount + 2);
sw.Write(_fileContentsBuffer.ToString());
_fileContentsBuffer.Clear();
}
}
Input0_ProcessInputRow(Input0Buffer Row) is the place where all the data processing takes place. As this script is the Destination for the data, we need to do something with those data. A Flat File Destination is not there and will not do this for us.
So here we append every detail row to the _fileContentsBuffer variable. Once in a while, we write the buffer to the file, and clear the buffer again.
/// <summary>
/// This method is called once for every row that passes through the component from Input0.
/// </summary>
/// <param name=”Row”>The row that is currently passing through the component</param>
public override void Input0_ProcessInputRow(Input0Buffer Row)
{
// Increase the rowcounter
_rowCount++;
// Append the new line to the buffer.
// The format needs to be like “{1}[middle part]{1}\r\n”, so that the row starts and ends with the text qualifier.
// Replace [middle part] with {1}{0}{1}{x} for every column that you want to append to the file, where x is 2 for column 1, 3 for column 2 and so on.
_fileContentsBuffer.AppendFormat(“{1}1{1}{0}{1}{2}{1}{0}{1}{3}{1}{0}{1}{4}{1}\r\n”, _delimiter, _textQualifier,
SafeValue(Row.CurrencyKey), // First column, {2} in the format
SafeValue(Row.CurrencyAlternateKey), // Second column, {3} in the format
SafeValue(Row.CurrencyName) // Third column, {3} in the format
// etc.
);
// Every x rows, write the buffer to the file, and clear the buffer.
if (_rowCount % _bufferRows == 0)
{
using (StreamWriter sw = File.AppendText(_fileName))
{
sw.Write(_fileContentsBuffer.ToString());
_fileContentsBuffer.Clear();
}
}
}
Multiple overloaded versions of the SafeValue method (for different datatypes like string, int, DateTime, you can add more when needed) add escape characters to the column values when needed.
if (value.Contains(_textQualifier))
{
// If the value contains one or more text qualifiers,
// escape them by replacing them with two text qualifiers.
return value.Replace(_textQualifier, _textQualifier + _textQualifier);
}
return value;
}
/// <summary>
/// Overloaded method for DateTime value (meant as an example)
/// For other data types, create new overloaded methods,
/// and in that overloaded method convert the value to string,
/// before calling the 'SafeValue' method with a string parameter.
/// </summary>
/// <param name=”value”></param>
/// <returns></returns>
private string SafeValue(DateTime value)
{
return SafeValue(value.ToString(“yyyy-MM-dd HH:mm:ss.fff”));
}
/// <summary>
/// Overloaded method for int value
/// </summary>
/// <param name=”value”></param>
/// <returns></returns>
private string SafeValue(int value)
{
return SafeValue(value.ToString());
}
}
When the script is ready, Choose Rebuild from the BUILD menu in the VstaProjects – Microsoft Visual Studio Window. Then close that Window. You return to the Script Transformation Editor, close it by pressing OK.
I simply run the package in debug mode. 105 currency rows are added to the csv file.
Including the header and footer, the csv file contains 107 rows, and this is the number that is written to the footer line.
Running the package in debug mode
First and last part of the created csv file.
Please note that to change the delimiter and text qualifier you only have to make two changes in the ***** Configuration ***** region.
// Here you can change the column delimiter and the text qualifier easily.
private const string _delimiter = “;”;
private const string _textQualifier = “\””;
. . .
#endregion ***** Configuration *****
By this change the file will look as follows:
First and last part of the created csv file with a different delimiter and text qualifier.
I also did a test with a source query that returned over 1.1 million rows, and the csv file of 25 MB was created in just 3 seconds (on a core i7 laptop with SSD).
So my conclusion is that the performance of this solution is very satisfying.
In this post I have demonstrated how to use SSIS 2014 to create a csv file with a dynamic header (with a timestamp) and footer line (including a rowcount) using a Script Component as a Destination in a Data Flow Task.
Big Data & Analytics insight 2016: save the date for an insightful conference
(as a speaker I was asked to do a bit of promotion ..)
(c) 2016 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.