According to a recent study, Microsoft Azure grew its revenues by 136% on a rolling annual basis, making it the fastest growing cloud IaaS vendor, coming in second in size to the market leader, AWS. As part of our plan to support multi-cloud strategies, we’ve ventured out to launch Cloudyn for Azure and have taken the liberty of investigating the Azure pricing models over the past few months. In this series of articles, we will share what we’ve learned about Azure’s pricing.
Check out the Synergy report.
As with every online cloud service, Azure supports free tiers. Simply set up your Azure account and start your free usage. You will then be able to download reports about server, application and website usage. In addition, you can use Azure’s pricing calculator to estimate how much your servers will cost before you actually deploy.
Azure Cloud Dashboard
Azure is different from many providers in that it initially was designed as a PaaS. However, due to market demand, Microsoft introduced its classic IaaS offering. The Azure Virtual Machines offering supports Windows and Linux, as well as databases such as SQL Server. All machines and services are made up of two components: tier and size. Prices vary between regions for the various machine families, which include:
- General Purpose/Basic Tier (Series A) – for dev/test
- General Purpose/Standard Tier (Series A) – for more options and sizes on top of the basic tier
- Optimized Compute (Series D) – for faster CPU, more memory and SSD
- Network Optimized (Series A) – 40Gbit/s InfiniBand network
- Compute Intensive (Series A) – 8/16 cores, 56/112 GB RAM
Instead of dealing with the machine itself, Azure enables you to package and deploy your application. Azure Cloud Services provides you with another PaaS experience, allowing you to run your Java, Node.js, PHP, Python, .NET, and Ruby applications. Pricing is based on the machine you use for your application.
Cloud Services offers two roles: worker and web. The web role runs IIS allowing you to run simple web applications, while the worker role does not. Instead, the worker role is used for advanced capabilities such as processing.
You have to select different machine sizes, as well, with varying prices for each worker (machine) that you use. If you deploy your application using two machines and specify that you want autoscaling capabilities, additional workers will be provisioned automatically if necessary. Azure charges differently per hour based on worker or web roles. The main advantage of Cloud Services is that you do not need to manage and administer the infrastructure, yet you will still have control over the virtual machine. Instead of dealing with the machine itself, Azure enables you to package and deploy your application as is.
Azure’s Websites is a fully managed PaaS that supports a higher level of abstraction in multiple pricing tiers. With Azure Websites, you do not control the infrastructure – everything is managed by Azure. With the virtual machines behind the scenes you can deploy and publish a static website or simple enterprise application. Azure Websites is available in four pricing models: Free, Shared, Basic and Standard. You can host an unlimited amount of websites or applications with Websites with varying sizes of storage and SLAs depending on the model you choose. In order to package and deploy a service, you need to use one of the following tools: Microsoft’s Developers Studio, an Azure plugin for Eclipse, or PowerShell.
Choosing between the different options depends on the level of control you wish to have. Using virtual machines is the most natural, straightforward cloud consumption approach. You can create a virtual machine using Azure’s management console or REST APIs. Creating services requires a bit of a deeper understanding about Azure because they can’t be created from the management console.
As in every cloud environment, pricing doesn’t end here. Additional components contribute to overall charges, including storage, network, backup and even security (MFA). Cloudyn currently supports cloud management and cloud usage and utilization tracking and analysis for Amazon Web Services, Google Cloud Platform and OpenStack clouds. In our next article we will elaborate more on Azure’s enterprise enrollment, discussing its structure and how it can help you allocate and monitor resource costs across your organization.
You can see for yourself how to improve your Azure deployment and cut costs easily with our 20-day free trial.