Azure vs AWS vs Google Cloud (GCP): The State of Cloud Technology
Amidst a mind-blowing array of new cloud technology service offerings, worldwide cloud infrastructure spending topped $80 billion in 2018 according to the research firm Canalys, and many analysts predict more than 20% growth for the market in 2019.
Cloud architects have grown more savvy in recent years, but the always-changing landscape makes it difficult to keep abreast of trends, strengths and weaknesses of cloud providers, or even new partnerships among cloud providers that may be of value. When it comes to Azure vs AWS vs Google, it can be difficult to know where to start.
To help, here’s an overview of major cloud infrastructure providers, considerations and recommendations for technical planners, and a peek into the future of cloud technology service offerings that might impact your decisions now.
Top 3 Cloud Technology Providers
Let’s look at the three market leaders for cloud infrastructure according to Canalys: Amazon Web Services (AWS, 32%), Microsoft Azure (16%), and Google Cloud Platform (GCP, 9%).
Pros: Mature platform, more affordable, vast number of services and options, e-commerce leader
Cons: Overwhelming number of choices, and retailers and businesses that compete with Amazon might not want to leverage a competitor
Amazon has been around since the mid-1990s and is currently one of the Big Four technology companies in the world alongside Google, Apple, and Facebook. More than an e-commerce platform, Amazon is also known for its digital streaming services and artificial intelligence offerings as well as its cloud offering: Amazon Web Services or AWS. Amazon wants to enable customers to build and sell using its technologies and has wisely productized the lessons it learned in becoming a juggernaut of e-commerce on its journey into AWS. The company is constantly innovating and looking to increase efficiencies in its own platform, passing the savings it obtains from economy of scale to customers in the form of lower prices. When considering Azure vs AWS vs Google cloud, it’s also important to remember that Amazon has very mature best practices around self-healing, fault tolerance and other essential elements that are assumed from a top cloud provider.
Pros: Strong in .NET/Microsoft technology stacks, enterprise-focused, robust in integration between application components and data components
Cons: More expensive and not as mature as AWS
Microsoft and .NET have been around a long time, and anyone deeply bound to the .NET framework will derive the greatest benefit from using Azure because the .NET offering in Azure is far superior to offerings in other environments. Simultaneously, enterprises that manage on-prem authentication and security access through Active Directory have tight integration with Azure Active Directory domains. Additionally, the pervasive Microsoft SQL Server database backend works very well in the Azure SQL infrastructure.
Pros: Analytics, Kubernetes knowledge, low pricing model, attractive to startups and small businesses because of free offerings
Cons: Lack of networking and application engineers with GCP knowledge for tuning the platform to get the most from the service, not as mature a platform as its competitors
Simply stated, Google knows data. With 20 years of mind-boggling amounts of data at hand and the brains to weave it into innovative offerings, Google has the potential to take a huge bite of the cloud architecture market share in coming years. GCP also sports a healthy maturity in containerization services, having developed and open-sourced the container orchestration service Kubernetes, or K8s for short.
Note: A major drawback for any provider is becoming too dependent and deeply invested in one provider’s technology. For example, when migrating applications to the next-generation serverless architecture model in AWS, Lambda, it’s easy to get married to those functions to where it’s difficult and very costly to migrate away from them.
Other Cloud Technology Providers
Numerous additional cloud architecture providers serve the market with a wide variety of cost models and offerings, including Alibaba Cloud, IBM, Oracle, SalesForce, Rackspace, and more.
Considerations and Recommendations
For technical planners, making decisions in the continually morphing cloud infrastructure landscape can be like trying to hit a bullseye on a moving dartboard. Here’s an overview of critical facets to consider before making major bets on your enterprise cloud architecture:
A huge driving point when building or maturing your cloud infrastructure is your application platform — is it Java, .NET, or Unix/Linux? As mentioned, for .NET platforms, Azure is a safe choice. For Java stacks, many believe AWS is the best provider. However, there’s no set rule to follow — for example, Java, .NET and Unix/Linux stacks can run on all three major cloud providers, but some will be more inclined for integration than others.
Containers (such as Docker containers) are software packages that include everything necessary to run an application and are great equalizers because they can be leveraged across all three major cloud platforms. While containerization is becoming more widespread, a strong delineation exists between the aptitude, skillset and appetite for its use. Price points, which can be based on CPU and memory usage and might include data transfer surcharges, differ among providers.
When considering your containerization strategy, it’s important to determine the level of maturity and complexity you want for containerization in your organization and what features best fit your model. Most containers are administered by Kubernetes, installed through the open-source technology Kops or through a managed service such as Elastic Container Service (ECS) in AWS, Azure Kubernetes Services (AKS) in Azure, or Google Kubernetes Engine (GKE) in GCP. Also worth considering are technologies like Fargate, an AWS offering that further simplifies ECS management by removing the need to manage servers or clusters.
Rising in popularity, serverless architecture is a model in which server management overhead is essentially removed from the software development lifecycle, so the only management organizations are responsible for is managing their own code. It’s highly important now to assess your appetite for leveraging serverless in your enterprise, as it can result in tremendous infrastructure cost savings down the road. Top serverless providers include AWS (Lambda), Microsoft (Azure Functions) and Google (Google Cloud Functions).
Be aware, however, that making a decision toward serverless will probably involve planting a stake in the ground for one provider, as leveraging multiple providers will call for having multiple skillsets in your environment. That’s not a drawback if you have those skillsets already on board, but it can be costly to acquire them or train for them with existing personnel.
Also keep in mind that in some circles, containerization and serverless are considered as competing technologies, but they can be a powerful combination when integrated.
Cloud Infrastructure: The Future is Multi-Cloud
The future of cloud infrastructure is not necessarily Azure vs AWS vs Google; rather, it will undoubtedly involve leveraging multiple public clouds simultaneously, often termed “multi-cloud.” A recent survey of 310 technical and business executives found that AWS and Azure are being paired together in multi-cloud deployments, and the move to multi-cloud might be occurring faster than previously thought. That said, global organizations are more likely at this time to move to multi-cloud scenarios than smaller companies that might not have the financial resources and skillsets.
Leveraging a multi-cloud infrastructure enables greater choice, diversity, and hedging of bets versus going with only one provider. Those leveraging multi-cloud architectures are wise to use provider-agnostic tools such as Kubernetes and HashiCorp’s Terraform Infrastructure as Code (IaC) toolset. Terraform can also be utilized on-premises, allowing enterprises to leverage hybrid computing platforms much more easily.
New partnerships will continue to evolve the cloud infrastructure landscape as well. A great example is the recent partnership between traditional mortal enemies Microsoft and Oracle for direct high-speed connectivity between Azure and Oracle Cloud, which will enable organizations to continue to leverage best-of-breed cloud technologies from both organizations simultaneously. Because partnerships like this are win-win for everyone, expect cloud providers to continue to investigate these opportunities.
No discussion of cloud futures is complete without a mention of machine learning and AI. Cloud providers will continue to reduce the learning curve for integrating data and analytics technologies into application stacks, lessening the need for deep knowledge in these areas within IT staffs. And as more of these technologies are plugged into the foundation of cloud infrastructures, they will fuel more innovation and transformation for organizations far into the future.
Consider Cloud Consulting
Cloud has greatly matured in the last five years, yet a solid and healthy cloud technology solution can still be a moving target for many organizations. AIM Consulting’s Cloud Consulting Services can help with cloud strategy (including an assessment of the architecture, cost, and security as well as building proof of concept to pilot a recommended cloud solution within the parameters of your requirements), migration and workload management, and DevSecOps.
Cloud consulting services include:
- Helping you choose a cloud provider (Azure vs AWS vs Google)
- Cloud readiness assessments for existing applications
- Architecture, cost, and security reviews
- Building a proof of concept in a pilot program to validate the cloud strategy
- Improved business stability and a roadmap for business continuity
- Elastic infrastructure that scales up or down depending on demand
- Cloud architecture and application reviews
- Improving the speed and quality of application deployments
- Reducing duplication and waste to make environments leaner and more efficient
- Evaluating the skills and maturity of your team and filling in gaps
- Assistance with cloud migrations and workload optimization
Cloud consulting can guide challenging decisions and help your organization get the greatest value from its cloud investment.
Mark has over 20 years of experience working within the Cloud and DevOps implementation space in the government, healthcare, education, and technology sectors. Mark is passionate about helping clients realize the benefits of leveraging the latest Cloud and DevOps technologies and practices.