May 2009

You are currently browsing the monthly archive for May 2009.

I wasn’t expecting to be writing another item about customer service so soon after my lost rope issue, but here’s another story about excellent support from an online shop.

Sunday, 10th of Mayimage

I made an order at  The order included four Yondi travel pillows for our kids and the Flippin’ Penguins game, for our kids as well :-) .

Monday, 11th of May

I received a confirmation that the order had shipped using TNT.

Wednesday, 13th of May

Shipment delivered at our neighbours (on our request as we were not at home during the day).

So in the evening we started unpacking the box.  All items were well protected using airbags to fill up the box.  There’s was even a free edible goodie, a mint Aero bar.  As far as I know this is not sold in Belgium so it was an interesting bite.

The penguin game requires two batteries, luckily we still had some waiting in our battery drawer.  Upon insertion of the batteries – exciting moment – we swapped the on/off switch and … nothing happened.  Hmm, bad batteries ?  Too old ?  Tried another set.  Still nothing.  All of them bad ?

To rule out an issue with the batteries we decided to try the device using freshly-bought batteries before filing a complaint.

Some days went by…

Tuesday, 26th of May

We finally had bought new batteries and tried them out.  But still no go so the engine really is broken.

The Firebox website has a very clear explanation on what to do when faulty goods are received.  The customer has 30 days time after receipt of the order so I wrote them an email explaining in short what we already tried out.

As this is an overseas order (UK > Belgium) I was quite curious to see how they would handle this.  After all, it costs money to ship such a large item and the sale price is £9.95.

Wednesday, 27th of May

Around noon the following reply was sent:

Hello Valentino,

I’m sorry to hear that you are having problems with the Flippin’ Penguins. Ordinarily we would get you to return your faulty product and then we would replace it.

However, rather than delay the matter further we are going to send you out a replacement ref: 1847702. You can dispose of the faulty ones that you.

Please do not hesitate to contact us again if you require any further assistance.

Kind Regards,
Sarah Kabunga
Firebox Team
Tel: 0844 922 1010

WINNER – Retailer of the Year (2008) Stuff Awards
WINNER – Best Retailer (2007) T3 Awards

Bored? Take a look at the Firebox Lunchtime Challenge, guaranteed to liven up your day…

“…we don’t stop playing because we get old; we get old because we stop playing.” Ltd is registered in England and Wales, company number 3874477
Registered Company Address: 41 Welbeck Street London W1M 8HD

In the evening of that same day I received a confirmation email to let me know that the “order” has shipped.

Friday, 29th of May

Shipment received.  Again we opened the box and inserted the batteries in the device.  Another exciting moment … flipped the switch … IT WORKS!

The box contained another Aero bar and a “£5 off on next order” voucher.


  • Excellent customer service
  • Fast delivery
  • Free goodies
  • If I keep writing these types of post I should start to ask for commission


Here’s a quick tip on the usage of the Merge, one of the Data Flow Transformation components in Integration Services.  When merging data from two different sources one of the issues that you’ll need to cope with are the data types.  The Merge component may start complaining about certain columns when the metadata has been changed higher in the flow.  The error would look like this:

Validation error. Data Flow Task: Data Flow Task: The metadata for “input column “YourColumn” (590)” does not match the metadata for the associated output column.

A common issue is the length of a string field.  My favorite way to solve these types of issue is by opening up the Advanced Editor, locating the particular field in the Output Columns and increasing its Length property to match the length of the longer field.

But unfortunately the Merge component does not have an Advanced Editor option.  The easiest way to solve it in this case is by first deleting the column in the Merge Transformation Editor and then adding it again.  Luckily the dropdown only shows the input fields that haven’t been matched yet. :-)


Tags: , ,

A while back when I started using a preconfigured server with Windows Server 2008 and SQL Server 2008 I came across the following.

My reports were deployed to the server, they were working fine when using Internet Explorer directly on the server.  But when I tried to connect remotely to the Report Manager (http://server.domain/Reports) I got the well-known “Page cannot be displayed” message.  Then I tried the Report Server (http://server.domain/ReportServer) and that seemed to work fine.  Until I asked it to generate a report, same issue.

I then logged on to the server to have a look if anything was appearing in the SSRS-related log files.  I saw the GET requests with the ListChildren command from Report Server getting logged in the HTTP log, but no Render command.  Other log files didn’t show anything for the report render requests and neither did the ExecutionLog2 view in the ReportServer database.  Which is quite logical of course, if there’s no GET request, there’s no execution either.  But hey, I was exploring all options here.

Then I activated HttpWatch to have a look at what was happening with the GET request.  And guess what!  It was being redirected to HTTPS!  So the report server apparently did not accept remote unencrypted connections, it forced the user into HTTPS.  But that didn’t work either, probably because no one properly configured SSL.

In my situation there’s no need for the encrypted connection so I started looking for the switch to turn off SSL.  I had a look in the Reporting Services Configuration Manager, no checkbox that says “do not redirect to SSL” or similar.  Nothing in the Report Server properties either.

Apparently the only way to turn SSL off (or on for that matter) is by changing a setting in the RSReportServer.config file.  This setting is called SecureConnectionLevel and valid values are:

0 Least secure. The report server does not check for a secure connection when Web service methods are invoked. However, all calls to the Web service can still be made to the report server over a secure connection.
1 Minimally secure. All Web service calls that are made over an insecure connection and which might pass sensitive data such as user credentials are rejected. However, this setting does not guarantee security. It is still possible for sensitive data sent by the client to the report server to be exposed before the report server handles the request and rejects it.
2 Secure. All rendered reports and all Web service calls require a secure connection. This includes all calls to the Render method and requests for rendered reports made through URL access.
3 Most secure. All calls made to the Reporting Services SOAP API require a secure connection.

The descriptions in the table above are copied from the Using Secure Web Service Methods article, see there for more details.

In my situation the setting was set to 2.  Changing it to 0 solved my issue.  This also explains why the Report Server was partly working.  When the security level is set to 2, all calls to Render and other web service methods need to be secure while listing the children can be done unencrypted.

In case you’re interested in actually using SSL, have a look at the Configuring a Report Server for Secure Sockets Layer (SSL) Connections article.

I noticed that my blog item contains links to almost all relevant MSDN BOL pages related to Reporting Services configuration, interesting to keep in mind :-)


Tags: , ,

About a week ago Gill announced that he’d finished and published the new website so here’s my part of spreading the word.


Click the banner above to register for the free event or put your browser on



Lately I had been getting annoyed by Visual Studio crashing on me while working on ETL packages.  The crashes seemed somehow related to debugging a package because they usually occurred after I clicked the “Package execution completed. Click here to switch to design mode, or select Stop Debugging from the Debug menu.” link.  But I couldn’t really pinpoint why or when exactly they occurred.  Now I’ve got an explanation, and a solution!

How did I get to the solution?  Well, today I decided to click the “Send to Microsoft” button on the infamous crash pop-up window.  And from it I actually got a link to a fix for the issue.  Apparently the issue is not related to Integration Services in particular, it is a much broader Visual Studio shell issue that occurs in Visual Studio 2008 with SP1 when you’ve got a combination of undocking windows and changing window layout.  And that is indeed what I usually do when debugging a package!  When the package stops executing I often execute the following scenario:

  • double-click the title bar on the Output window to undock it
  • enlarge the undocked Output window to almost full-screen so I can have a good look at the errors
  • double-click the title bar to dock the window back to its original place
  • click the “Package execution completed.” link (or hit the Stop Debugging button)

And that’s exactly the second scenario described in this Microsoft Support article because when you stop debugging, the IDE switches the window layout back from debug to design.

I have now installed the hotfix and the issue is gone.  Beware though if you also use WPF, better read through the whole article and comments on the download page first.

Hmm, this may also explain some other unexplicable crashes I’ve seen lately.  Makes you wonder doesn’t it :-)


Tags: , , , , , , ,

« Older entries

© 2008-2019 BI: Beer Intelligence? All Rights Reserved