When Advanced Telemetry, an energy consumption monitoring and management provider, realized they could not accommodate...
their rapid growth with on-site infrastructure, they looked at cloud computing options to expand their computing resources.
As it moved to the cloud, Advanced Telemetry decided to move away from the relational database. The relational database was not able to sufficiently handle the company's growing stores of data—one table receives 60,000 to 100,000 records a day—as fast as was needed. "Relational databases are not good at managing and accessing that much data in real time," said Naylor.
Advanced Telemetry first looked at Google BigTable to meet its needs, but eventually chose Microsoft Azure. "All Google BigTable offered was offloading our data storage requirement, which was our primary need initially," said Naylor. "We became aware of Azure and started to give the table storage a good look. To my surprise, the framework Microsoft was developing offered many more benefits."
For Naylor, those benefits include the building blocks for new cloud applications. Prior to migrating to Azure, Advanced Telemetry developed most applications from scratch. "There's all kinds of data being queued, services that run, background services that run—all things we had to essentially design and develop on our own," said Naylor. "In the Azure framework, you get replacements for these things."
The migration, though, has not been without difficulty. "The biggest issue that we've run into is with the table storage itself," said Naylor. "They probably shouldn't have used the term table. It's not a table in the traditional sense of the relational database table."
When a problem arose, it was difficult to pinpoint. "The type of errors you get when you're doing insertions can be very vague," said Naylor.
Naylor attributed some problems to the immaturity of the product. "That's the reality when you're pre-beta," said Naylor. "We have a customer technical preview account that we used until this Monday." Advanced Telemetry is now transitioning to the commercial version of Azure.
Visual Studio extension preps applications for cloud migration
Before it began to migrate its data, Advanced Telemetry migrated its desktop and Web applications to Azure. To do this it used Visual WebGui from Gizmox. Visual WebGui is a Web application development tool designed to simplify the development process by abstracting code with a drag and drop graphical interface. It is used on top of Microsoft Visual Studio.
Created over four years ago, Visual WebGui was not intended to assist cloud deployment. But certain characteristics make the product a good fit for such a process.
Itzak Spitzen, vice president of research and development at Gizmox, located outside of Tel Aviv, says the platform requires low CPU usage, thus saving money on cloud-based servers that are rented by the instance. "Visual WebGui is fully stateful, so the server doesn't have to create and destroy objects. The amount of data sent to the server is very low," said Spitzen.
Spitzen also says that Visual WebGui eases migration of legacy applications. The platform uses the WinForms API, so existing .NET code can sometimes be pasted into Visual WebGui and then enhanced.
Naylor said that using Visual WebGui to migrate applications to Azure was not without difficulty. "There were a lot of configuration issues that had to be accessed. It was not just plug-and-play."
Naylor said the company relied on Gizmox support to resolve these issues. "The support was there when we needed it," said Naylor.
Visual WebGui fit Advanced Telemetry's need to make familiar applications. "It made sense for us to use Gizmox because [applications made with it] look like a windows desktop app," said Naylor.
The migration to Azure was not the company's first effort with cloud computing. "About a year and a half ago we offloaded some of our critical duties to Rackspace," said Naylor.
Advanced Telemetry is now migrating resources hosted on Rackspace to Azure to take advantage of Azure's billing mechanism. "With Rackspace, you have to pay for the maximum possible that you could use," said Naylor. "With Azure you could scale as you needed it."