For those of you not familiar with Azure, it’s Microsoft’s cloud computing platform made up of three parts:
- Windows Azure, the operating system in the cloud where your applications run
- SQL Azure, the database engine in the cloud where your data lives, and
- AppFabric, which connects cloud, hosted and on-premises services together.
As for the Azure development patterns that fit common business scenarios, they are:
- Transparence: Simply moving applications and data from servers to the cloud. The benefits are cost savings, not having to manage servers, cost-effective scaling and opportunities to prototype without having to invest in additional hardware and software.
- Scale-in multi tenancy: On-the-fly scaling by creating new Azure instances when demand increases. It’s hard to predict what demand for an online service will be; this “just in time” approach does an end run around having to make such forecasts and purchases based on them.
- Burst compute: This is scaling based on known peak periods, such as the Christmas rush for retailers or the Superbowl for pizza delivery. A cloud-based system like Azure lets you acquire more server capacity during those known peak periods and release them once the peak period is over.
- Elastic storage: This is data scaling – you can use Azure to extend your storage instead of purchasing more on-premises disk arrays.
- Inter-organization communication: Using Azure to host an API to connect to your company’s services or data (which may live in Azure, on some hosted system or on-premises). It’s a good way to provide services to the outside world while keeping your infrastructure manageable.
There’s more in the article, and even more in the interview, which you can either: