We are happy to announce our first user group session of 2011…CSI style!

Tom Van Zele will bring us SQL Server Forensic Analysis – Playing CSI with databases.

You get a call from a colleague who thinks that his financial information system has been hacked. Of course this system uses a SQL Server Database back-end. In this session we’ll uncover the basics of digital forensics: how to find out whether the system has been accessed by the intruder? What data has been accessed? Was there anything modified? We’ll learn on how to find evidence of what happened and also on how to create a report that can be served as evidence in a court of law.

About Tom Van Zele

Tom is a seasoned IT professional with more than 10 years experience, first in large scale Active Directory implementations, later moving on to SQL Server. His focus lies on infrastructure & security. You can follow Tom on his blog www.synsol.eu/blog and Twitter @tvanzele

When

February, 28th

Agenda
18:00 Registration
18:30 Session Start
21:00 Session End
Location

Microsoft België
Corporate Village
Leonardo Da Vincilaan 3
1935 Zaventem

Click here to register

Hope to see you at the event!
The SQLUG Team

I just tried to rename one of the reports, what should be an easy thing to do, but I got this really weird error message:

System.Web.Services.Protocols.SoapException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database. —> Microsoft.ReportingServices.Diagnostics.Utilities.ReportServerStorageException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database. —> System.Exception: For more information about this error navigate to the report server on the local server machine, or enable remote errors
at Microsoft.ReportingServices.WebServer.ReportingService2005Impl.GetItemDataSources(String Item, DataSource[]& DataSources)
at Microsoft.ReportingServices.WebServer.ReportingService2006.GetItemDataSources(String Item, DataSource[]& DataSources)

This error did not make any sense to me, so I looked into the Report Server Trace Log for more info and found this:

library!ReportServer_0-11!6028!07/01/2010-14:15:39:: e ERROR: Throwing Microsoft.ReportingServices.Diagnostics.Utilities.ReportServerStorageException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database., ;

library!ReportServer_0-11!6c18!07/01/2010-14:15:39:: e ERROR: Throwing Microsoft.ReportingServices.Diagnostics.Utilities.ReportServerStorageException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database., ;

library!ReportServer_0-11!6028!07/01/2010-14:15:59:: Call to GetPermissionsAction(http://hictatriuap017/sites/DashboardsICTRA/ReportsLibrary/SLA PRD/Report Management Summary.rdl).

library!ReportServer_0-11!6c18!07/01/2010-14:15:59:: Call to GetItemDataSourcesAction(http://hictatriuap017/sites/DashboardsICTRA/ReportsLibrary/SLA PRD/Report Management Summary.rdl).

library!ReportServer_0-11!6c18!07/01/2010-14:15:59:: e ERROR: Throwing Microsoft.ReportingServices.Diagnostics.Utilities.ReportServerStorageException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database., ;

library!ReportServer_0-11!6c18!07/01/2010-14:15:59:: e ERROR: Microsoft.ReportingServices.Diagnostics.Utilities.ReportServerStorageException: An error occurred within the report server database.  This may be due to a connection failure, timeout or low disk condition within the database. —> System.Data.SqlClient.SqlException: Cannot insert duplicate key row in object ‘dbo.Catalog’ with unique index ‘IX_Catalog’.

The statement has been terminated.

For every report in the report library, reporting services creates a record in the Catalog table of the report server database. But when you delete a report, that record doesn’t seem to be deleted.

The catalog table has a unique index on the column “path”, that contains the full path to the report. When I tried to rename the report, it failed because a previously (deleted) report existed with that same name.

You can work around this problem by changing the value of the path column of the old record.

If anyone has more details about this problem or maybe a solution, please leave a comment.

SQLUG teams up again with Microsoft to organize this panel of hardcore SQL Server experts with Wesley Backelant(MSFT), Frederik Vandeputte(MVP), Dirk Gubbels(MSFT) and Gunther Beersaerts(MSFT).

Starts: 19:00
Location: Microsoft België

Send your SQL Server related questions to asktheexperts@sqlug.be.

We will answer as many questions as possible within the available timeframe.

About The Panel

Wesley Backelant
In his role as a Technology Advisor at Microsoft, Wesley is responsible for helping customers understand the capabilities of SQL Server and the Microsoft Business Intelligence stack. Before joining Microsoft, Wesley was a Database Architect working on some of the largest implementations of SQL Server in Belgium. Wesley started his professional career in the SQL Server 6.5 timeframe and remained true to his passion ever since. Wesley is active on Twitter where he handles topics related to his favorite technology.

Frederik Vandeputte
Frederik Vandeputte is a Senior Consultant and partner at Kohera, the Microsoft SQL Server/Business Intelligence Competence Center of the Cronos Group. Frederik has been working with SQL Server since version 6.5. In his free time Frederik collects Microsoft certifications. His collection includes MCTS, MCSA, MCSE, MCDBA, MCITP, MCT, ranging from Windows 2000 and SQL Server 2000 up to SQL Server 2008. Frederik is one of the co-founders and the President of the Belgian SQL Server User Group (SQLUG.BE). In January 2008 Frederik, became the first Belgian MVP on SQL Server. Follow Frederik on his website and twitter.

Dirk Gubbels
Dirk Gubbels is a senior consultant at Microsoft, and has been working with SQL Server since version 4.2. As one of the few Microsoft Certified Database Architects he has been involved in the most demanding SQL server based applications in Belgium and all over the EMEA region. His main focus areas are Design, Performance and Availability for both OLTP and Business Intelligence environments

Gunther Beersaerts
Gunther Beersaerts joined Microsoft in 1998 (on the launch day of SQL Server 7.0) as a Technical Marketer for MSDN/TechNet road shows and has held a number of technical roles during his career, including Systems Engineer, ATS, TSP roles covering a broad set of Microsoft Application Platform solutions. Over the past few years, Gunther has been active in technical roles for Databases and Business Intelligence platforms in EPG Belgium & Luxemburg. He then became a Strategist in the CATM (Customer Advocacy and Technology Management) organization which is a key-connection between the Microsoft Development teams and Customers /Partners. In this role, Gunther focuses on the Microsoft Data Platform, including SQL Server and Business Intelligence solutions. Prior to Microsoft, he was a developer and messaging engineer at a large financial institution in Brussels.

Agenda:

Location:
Microsoft België
Corporate Village
Leonardo Da Vincilaan 3
1935 Zaventem

Tel: +32.2.503.31.13
Fax: +32.2.704.35.35

This Event is brought to you by Microsoft

Click here to register

SQL Server Management Studio makes it very easy for you to create XMLA-scripts to create or alter cubes, dimensions, etc… But it’s not possible to use Management Studio to create an alter-script that only contains the MDX calculations. Because I needed to resolve a bug in the calculations but the rest of the cube was not ready to be deployed to production yet, I had to create an alter script just for the MDX.

Of course I could have started learning the XMLA-syntax to write a script for this, but I decided to go for the easy way: the BIDSHelper has an option to deploy the MDX to a cube without having to deploy the entire cube.

If you don’t know BIDSHelper, or don’t already have it installed, go to http://www.codeplex.com/bidshelper. This is a must have add-in for all SSAS and SSIS developers!

But as I was allowed to use Management Studio to makes changes to the cubes in the development environment only. So I needed to capture the XMLA script the BIDSHelper executes on the cube, to deploy the same script to the production environment.

That what the profiler is for, not? J

Open the SQL Server Profiler, create a new trace and connect to the analysis services instance. Use a blank template and switch to the tab “Events Selection”. The only event we want to capture here is “Command Begin”.

Click Run.

Now the profiler is ready to capture the XMLA-script, so switch back to Management Studio and use the BIDSHelper to deploy the MDX calculations. When the deploy is finished you can stop the trace.

The profiler should have captured 2 events. The event with subclass “Alter” is the one we need.

Copy the script from the profiler to a new XMLA script in Management Studio.

The last 3 lines (PropertyList) can be deleted.

In the node “ObjectDefinition” you will find your script.

Enjoy!

A Parent-child hierarchy is one example of this: When you browse Parent-child hierarchy from an Analysis Services cube with Excel, Excel automatically displays it with nice drill-down functionality.

But when you add it to a Reporting Services report, you get something like this:

That’s not how we like to display a Parent-Child hierarchy to our users!

To display it with the expected drill-down you need to do some extra configuration:

The following example uses the AdventureWorks example cube available on CodePlex:

Create a new report and add a dataset with the following MDX query:

SELECT NON EMPTY { [Measures].[Amount] } ON COLUMNS,

NON EMPTY { (DESCENDANTS([Organization].[Organizations].[Organization Level 01].ALLMEMBERS) ) }

DIMENSION PROPERTIES MEMBER_CAPTION, MEMBER_UNIQUE_NAME, PARENT_UNIQUE_NAME,

LEVEL_NUMBER ON ROWS

FROM [Adventure Works]

CELL PROPERTIES VALUE, BACK_COLOR, FORE_COLOR, FORMATTED_VALUE, FORMAT_STRING,

FONT_NAME, FONT_SIZE, FONT_FLAGS

Add a Tablix to your report and drag the parent-child hierarchy “Organizations” en the fact “Amount” to the Detail Row of the Tablix.

To enable drill-down you need to edit the Group Properties of the Row Group:

Configure the group to iterate on the UniqueName of “Organizations”:

Go to the Advanced options and set the property “Recursive Parent” to the ParentUniqueName of “Organizations”.

Change the visibility of this Group to “Hide” and check the option “Display can be toggled by this report item”. The toggle item should be the Textbox containing the Parent-Child hierarchy.

Now optimize the layout by increasing the indent for the lower levels of the hierarchy:

In the Textbox properties: set the following expression for “Padding – Left”:

=Cstr(Fields!Organizations.LevelNumber * 10) & “pt”

Click Preview to test the report.

Look’s way better isn’t it?

This session focuses on the out-of-the-box capabilities of System Center Operations Manager 2007 R2 but also shows you how a distributed application can easily reuse these capabilities to provide an extra step in monitoring.

About Ricardo Noulez
Ricardo Noulez is a Senior Consultant at Microsoft Consulting Services in the Benelux. Been with Microsoft since January 2002 as a subject matter expert in Microsoft Operations Manager (now System Center Operations Manager) and Systems Management Server (System Center Configuration Manager). Ricardo has a great focus around the System Center product family and he is currently expanding his horizons towards the System Center Service Manager product. He is always interested in finding ways to improve the way that customers manage their IT environment and has certifications and great affinity with ITIL good practices and the MOF framework to help them achieve this. In the first third of his 18 year career, Ricardo was a developer but gradually converted himself to Systems Management and IT operations. This gives him a broad understanding in the different needs and view-points of all technical players that make IT ideas come true.

Agenda:

• 18:00 – Registration
• 18:30 – Session Start
• 21:00 – Session End

Location:
Microsoft België
Corporate Village
Leonardo Da Vincilaan 3
1935 Zaventem

Tel: +32.2.503.31.13
Fax: +32.2.704.35.35

This Event is brought to you by Microsoft

Click here to register.

I came to the conclusion that the server always returned this error after exactly 2 minutes, so it had to be a timeout exception.

The report was kind of big, so tuning it would not really help (enough).

I took me some time to find out which timeout setting to change, because there seem be a lot of them:

There is the Query Execution timeout of the Dataset, a Report Execution timeout, the Connection timeout of IIS, other timeouts in the Report Server Configuration file, …

But the real cause of this timeout is an Execution timeout of Sharepoint.

You can change this timeout setting in the web.config of your Sharepoint site, by default in c\:inetpub\wwwroot\wss\VirtualDirectories\80.

In this web.config you will find an XML node called “httpRuntime”. In this node you have to add an extra tag: executionTimeout=”9000″:

<httpRuntime maxRequestLength=”51200″ executionTimeout=”9000″ />

Save and close the file and do an IISReset. When you run the report again, you won’t get this error anymore.

I also checked the web.config of the Report Manager, but in Native Mode this seems to be the default:

Good luck!

The component that enables this is the service broker, a message based system which allows for dialogs between two SQL server instances. In this session, we will first introduce the terminology and features of this framework, then we will discuss some of the scenarios in which it can be used, and we will conclude by implementing a small service broker application. Dr. Nico Jacobs was a researcher at the Computer Science department, KULeuven for seven years, where he obtained his PhD in machine learning (data mining). Since 2004, he is trainer and consultant at U2U, where he focuses primarily on SQL Server and the Business Intelligence back-end.

Free registration is now open.

• Opening Keynote by Wesley Backelant and Ritchie Houtmeyers
• Best Practices for SQL Server Consolidation by Tom Van Mulders
• Breaking the ETL world record with Integration Services by Henk van der Valk
• Designing Effective Aggregations in Analysyis Services 2008 by Chris Webb
• High availability for BI DWH by Patrick Tack
• How New.NET Development technologies could change the live of the SQL DBA guy/girl by Gill Cleeren & Kurt Claeys
• Implementing best practices with Policy Based Management by Nico Jacobs
• Introducing to PowerPivot or How I Learned to Stop Worrying and Love Excel by Chris Webb
• Tracking Changes and Data movement in SQL Server 2008 by Dirk Gubbels
• Handling XML in SQL Server by Nico Jacobs

Have fun!

bing brought me to the following great article, where Mosha explains how you can use MDX to create a custom format:
http://sqlblog.com/blogs/mosha/archive/2008/09/26/displaying-duration-values-mdx-expressions-in-format-string.aspx

When you apply this for a calculated member you get something like this:

    CREATE MEMBER CURRENTCUBE.[Measures].[DurationInDays]
     AS [DurationInDays],
    FORMAT_STRING =  
        IIF([Measures].[Duration] < 1
         ,’”0 days” hh:mm:ss’
         ,’”‘ + cstr(int([Measures].[Duration])) + ‘ days” hh:mm:ss’)
    ,
    VISIBLE = 1;

But this brings us to another problem:
This custom format will only display the correct result if [Measures].[Duration] contains decimal seconds.
For more information about the conversion from seconds to decimal seconds, read the following article on wikipedia: http://en.wikipedia.org/wiki/Decimal_time
In short, you need to divide the number of seconds by 86.400. The resulting number presents the number of days on the left side of the decimal separator and the remaining decimal seconds on the right side.

So the final result will be something like this:

    CREATE MEMBER CURRENTCUBE.[Measures].[DurationInDecimalSeconds]
     AS
        [Measures].[DurationInSeconds] / 86400,
    VISIBLE = 0;  

    CREATE MEMBER CURRENTCUBE.[Measures].[DurationInDays]
     AS [Measures].[DurationInDecimalSeconds],
    FORMAT_STRING =  
        IIF([Measures].[DurationInDecimalSeconds] < 1
         ,’”0 days” hh:mm:ss’
         ,’”‘ + cstr(int([Measures].[DurationInDecimalSeconds])) + ‘ days” hh:mm:ss’)
    ,
    VISIBLE = 1;

If you want this custom format to be displayed in Excel, make sure the connection properties are set correctly to retrieve the number format from the server:

Go to Connection Properties, tab Usage, and check Number Format in the OLAP Server Formatting options.

There is one more thing I would like to tell you about these options: Normally all these options will always be checked by default, but I’ve noticed that when you start from an existing odc-file, this is not the case. So, when you’re used to deliver an odc-file to your users to start analyzing from Excel, you will have to replace this odc-file by an Excel template that connects to the cube, to make sure this option is always on.

There might be a better way to change this behavior for odc-files, but I couldn’t find any. If you know it, please let me know

Enjoy!