markrubens - Fotolia
Azure PaaS strategy hones in on hybrid cloud, containers
Microsoft's PaaS offerings might have a leg-up in terms of support for hybrid deployments, but the vendor still faces tough competition in a quickly evolving app-dev market.
The Azure PaaS portfolio continues to offer a compelling story for companies that need a development environment where legacy applications can move freely between on premises and the cloud. But even as the vendor increasingly embraces hybrid cloud, open source and emerging technologies, such as containers and IoT, it still faces tough competition from the likes of Google and AWS.
Strong foundation
Azure App Service is Microsoft's flagship PaaS offering, enabling developers to build and deploy web and mobile applications in a variety of programming languages -- without having to manage the underlying infrastructure.
But App Service represents just one of many services that Microsoft has rolled out over the years to help developers create, test, debug and extend application code. The company's Visual Studio line, for example, now includes four product families: Visual Studio Integrated Development Environment, Visual Studio Team Services, Visual Studio Code and Visual Studio App Center, which includes connections to GitHub, Bitbucket and VSTS repositories to support continuous integration.
Microsoft has also created a vast independent software vendor and developer community, and has tightly integrated many of its development tools, according to Jeffrey Kaplan, managing director at THINKstrategies, Inc. Visual Studio and SQL Server, for example, support common design points and feature high levels of integration with App Service.
Microsoft's Azure PaaS strategy is also unique in its focus on hybrid cloud deployments. Through its Hybrid Connections feature, for example, developers can build and manage cloud applications that access on-premises resources. What's more, Azure App Service is also available for Azure Stack -- Microsoft's integrated hardware and software platform designed to bring Azure public cloud services to enterprises' local data centers and simplify the deployment and management of hybrid cloud applications.
Missing pieces
But despite its broad portfolio and hybrid focus, Azure PaaS is not a panacea. While many traditional IT departments have embraced the offering, it hasn't been as popular in business units, which now drive development initiatives in many organizations, according to Larry Carvalho, research director for IDC's PaaS practice.
What's more, organizations that don't have a large footprint of legacy systems often prefer open source development tools, rather than tools like Visual Studio. Traditionally, Microsoft hasn't offered support for open source technology as quickly as other cloud market leaders, such as AWS, according to Carvalho. This is likely because competitors like AWS are not weighed down by support for legacy products.
But while, historically, Microsoft's business model has been antithetical to the open source approach, that's started to change. The company has made an effort to embrace more open source technologies and recently purchased GitHub, a version control platform founded on the open source code management system Git.
The evolving face of PaaS
The PaaS landscape is evolving rapidly. Rather than traditional VMs, developers increasingly focus on containers, and interest in DevOps continues to rise. In an attempt to align with these trends, Microsoft now offers a managed Kubernetes service on its public cloud and recently added Azure Container Instances to enable developers to spin up new container workloads without having to manage the underlying server infrastructure.
Additionally, enterprises have a growing interest in application development for AI, machine learning and IoT platforms. And while Azure PaaS tools offer support for these technologies, Microsoft still needs to compete against fellow market leaders, AWS and Google -- the latter of which has garnered a lot of attention for its development of TensorFlow, an open source machine learning framework.
