If you are contemplating hosting some of your applications and data in one of the large public clouds, you are probably wondering what the performance and availability is like. 

Up to now, the only way has been to trust the vendors claims or to sign-up to the services and test them yourself. 

If you are considering hosting your application(s) with a cloud service provider …. read this first. 

There is now a service that compares the performance and availability of some of the larger solution providers such as Rackspace, Amazon, Google and Microsoft Azure. In a future release, they plan to add the capability for you to test your preferred supplier.

It’s interesting to see Microsoft Azure consistently at or near the top, with Amazon EC2 Ireland and Singapore lagging behind. 

• Over a 30 day period, Microsoft Azure’s typical response time is 0.63 seconds from Chicago, 8.44 from London and 40.55 from Mumbai.
• Amazon EC2 hosted in Singapore has response times of 5.89 from Tokyo up-to 40.52 from Boston USA.

These differences in latency are a real eye-opener. If you are considering hosting your application(s) with a cloud service provider, make sure you visit this site first. If your preferred provider isn’t listed, make sure you test it yourself before making the big move. 

Click here to try out the CloudSleuth service.

Also read ….

• Clouds and data jurisdiction
• Software as a Service (SaaS)
• Infrastructure as a Service (IaaS)
• Platform as a Service (PaaS)
• Desktop as a Service (DaaS)

Barclays Capital had to file a legal relief motion relating to the acquisition of Lehman Brothers’ US assets. This was after a reformatting error with an Excel spreadsheet resulted in 179 contracts being mistakenly included in the purchase agreement.

September 2008

Barclays Capital’s law firm, Cleary Gottlieb Steen & Hamilton, filed a motion with the US Bankruptcy Court that requested the contracts to be excluded from the deal. The motion was that on 7:48pm 18 September 2008, Barclays sent an Excel spreadsheet that contained a list of contracts to be included in the asset purchase agreement to their law firm, Cleary Gottlieb Steen and Hamilton.

The spreadsheet contained approximately 1000 rows and more than 24,000 individual cells and had to be   reformatted and converted into a PDF document before it was posted on the bankruptcy court’s Web site. However, a junior associate who reformatted the spreadsheet was unaware that the original Excel document had hidden rows that contained contracts marked with an “n” to signify they should not be part of the deal.

The reformatted document was posted on the court’s web site, but the hidden cells had become visible when the rows in the spreadsheet were globally resized – without taking into account the original “n” designations.

The error was not spotted until 1 October, when Barclays posted a revised list of contracts and advised the affected parties.

Cleary Gottlieb Steen & Hamilton filed a relief motion for Barclays Capital, asking the court to “correct the record to accurately reflect Barclays’ actual designations and confirm that those agreements that were erroneously posted are not closing date contracts”.

This is a big lesson to those companies who make critical decisions based on spreadsheets. Spreadsheets are databases that contain calculations and sometimes code. They must be tested like any other business system.