Finding the Right Fit in Microsoft Azure

Posted June 19, 2017

In Cloud, IT Planning, Microsoft, Microsoft Azure

Getting the benefits of cloud requires choosing the right cloud. That means not just the right cloud provider, but the right size cloud. If you choose too small a cloud, you won’t get the performance your applications need. Choose too large a cloud, and you’ll spend more than you need and lose the financial benefits of cloud.

Fortunately, cloud providers offer a variety of instance types to meet the performance needs of any application. Here’s how to select the right instance type in the Microsoft Azure cloud.

Microsoft Azure Instance Choices

Start by understanding your application and what factors drive its performance. Azure has instances optimized for compute-bound, storage-bound, memory-bound, and other application types. Once you’ve determined the type of instance you need, you can choose from various sizes that offer different processing, memory, and storage capacity. For some instance types, the underlying hardware may vary, while others commit to specific processor types such as Intel Xeon or Nvidia GPUs.

You may find it makes more sense and is more cost effective to choose smaller machines and scale by increasing the number of servers rather than running a smaller number of higher-powered machines.

General purpose: The A, D, and DS series offer matched CPU and memory capacity to serve general computing loads such as non-production environments, small databases, and low traffic web servers. The different series vary in the number of cores and CPU types, offering a range of performance options.
Compute optimized: F series instances offer a higher ratio of CPU to memory. These servers are useful for application servers and batch processing applications.
Memory optimized: G and M series instances come with more memory, making them appropriate for larger databases and in-memory databases and analytics.
Storage optimized: L series instances are optimized for low latency I/O, serving large databases and big data analytics.
GPU: The N series instances have specialized GPU processors to support graphics-intensive applications.
High performance compute: The H series have the most powerful CPUs and network interfaces that support high performance computing.

Monitor Usage to Adjust Sizing

Your initial Azure environment should be sized based on your expected capacity requirements, but capacity needs are never static. You should continue to monitor usage of your environment even after it’s live. This can help you detect instances where utilization is growing and you need to add capacity as well as instances which are underutilized and should be consolidated or shut down, such as development and test environments that are no longer needed.

You should also monitor usage of self-service features that let departments bring up additional instances, as that can lead to unwanted sprawl. Although you don’t want to put too many roadblocks in the way, there should be a relatively small number of users who are authorized to bring up new servers.

Need Help Deciding Which Microsoft Azure Instance is Right for You?

Do you need help deciding which instance is right for you? Prescient Solutions can help you evaluate your requirements and size your Microsoft Azure cloud so that you aren’t either over- or under-provisioned. As Microsoft Partners, we have the expertise to help you use Microsoft Azure effectively, whatever your cloud computing goals are. Contact us to learn how you can get the most benefit from your Microsoft Azure environment.