The other war between Microsoft, Google and Amazon: the battle to control cloud services for developers
In mid-2006, Amazon launched AWS, its cloud platform for developers. At that time, it was already a successful e-commerce website and had learned a great deal about internet service infrastructure, but no one imagined that it would become one of the benchmarks in cloud computing, surpassing technological giants of the time such as IBM or Oracle. , both with decades of experience behind them.
Today, Amazon is the clear dominator of the cloud services market, followed by Azure from Microsoft and Google Cloud. Although you already know that in technology nobody can rest on their laurels.
The cloud computing business grows every year. And if we look at the statements of Eyal Manor, VP of Engineering at Google Cloud: "Around 80% of the workload of the current servers is not yet in the cloud", we can imagine what can still continue to grow this industry that moves billions.
The current landscape in the cloud: leaders and laggards
No one would think today of starting a tech startup or internet business without using any of the cloud providers we've mentioned. In fact, Netflix uses AWS and Spotify just migrated to Google Cloud, for example. And they spend a very good peak of money year after year.
Amazon has a vast amount of available services that continue to grow exponentially, covering practically all the IT infrastructure of companies and developers. As of today, according to estimates from reports such as the Synergy Research Group, AWS has about 33% of the entire cloud computing market in the world. A figure that increases year after year to reach $ 7.7 billion in Q1 of 2019, almost 41% more than Q1 of 2018, according to CNBC.
Nobody would think today of starting a technology startup or an internet business without using the cloud.
Meanwhile, Microsoft is extremely strong on the whole SaaS (Software-as-Service) issue with its Azure cloud platform. Mainly because many developers still prefer to have an end-to-end ecosystem, with all the tools on the Microsoft family servers. In addition, it has a large number of SaaS services such as Office 365 or Dynamic, where it continues to earn a lot of money.
On the other hand, there is Google with Cloud Platform that, as we discussed in the last Cloud Next’19 conference, continues to try to convince us with its multi-cloud strategy. According to forecasts, 88% of companies in 2024 will combine different providers, that's where Google, with fewer services than its competition, sees the opportunity to use its extensive experience with cloud containers and Kubernetes to be the administrator of the different infrastructures.
We cannot forget an old acquaintance who is looking for his part of the market such as IBM. Reinforced by the acquisition of Red Hat, which strengthens its strategy based on creating private cloud infrastructures for companies, which moves away from public cloud services from AWS, Microsoft or Google. It is difficult to compete today in the public cloud of its competitors.
And finally, we should mention the surprising emergence of Alibaba, another known for its e-commerce business, which has been growing a lot in China in recent years. Although seeing how the panorama is, we do not anticipate that it will be able to enter with a lot of market share in the US or Europe due to trade problems with the Asian giant.
We are going to review each of the three main competitors in the cloud business such as AWS, Microsoft Azure and Google Cloud Platform to discover how they differ and what are their advantages and disadvantages.
Amazon AWS, its vast amount of services is the foundation of the IT infrastructure of any company in the cloud
Amazon's cloud service is the clear dominator with a large catalog of IT services for companies and developers. It has been the undisputed leader as an IaaS service provider for the past ten years.
Its objective has always been to replace the traditional infrastructure of any IT company, in addition to providing all kinds of tools for system administrators and developers.
It all started with EC2, Elastic Compute Cloud, which allowed to “rent” and scale virtual machines in the cloud paying only for the processing time, supporting both Linux and Windows. It has a large number of configurations and options, having the user full power to install any software on those machines. Which has evolved to pre-established images of certain most common servers and services. A good part of this marketplace of tools is provided by companies like Bitnami, facilitating that task and creating a new business model as providers of cloud utilities.
Another of Amazon's best-known services is Amazon S3, a storage service widely used by all kinds of users from websites to file repositories. Almost almost any file or images on the internet comes from there or from one of its associated CDNs. Elastic Block Storage (EBS) or Elastic FIle System (EFS) joined this service to build more complex data files. Nor can we forget the databases: Database Service (RDS), DynamoDB NoSQL or RedShift that promote the interconnection of services and the registration of information.
Of course, we leave dozens of these services in the pipeline. You can check on its official AWS page practically everything necessary for a technology company to scale from 0 to millions of users.
Microsoft Azure, leveraging the entire Microsoft cloud ecosystem
Microsoft was late to the cloud market, although it took the leap of the curve thanks to its infrastructure based on its own Windows Server, Office, SQL Server, Sharepoint or .Net servers. All this reusing it to be deployed in the cloud, Azure is a flagship.
Probably the big point in favor of Azure is that many companies deploy on Windows and have the entire Microsoft software ecosystem ready. The almost ubiquitous inclusion of Azure makes it easy for any Microsoft service to be used on Microsoft's cloud infrastructure without much friction.
Much of Microsoft's cloud computing is based on Virtual Machines with support for Linux, Windows Server, SQL Server, Oracle, IBM or SAP. Like AWS, it has a wide catalog of configurations. Apart from that, Azure has its infrastructure based on Kubernetes and Docker Hub to use containers that can be scaled independently without having to worry about the virtual machines we need. Really, the big trend of almost all cloud providers.
As for storage, Azure offers solutions such as Blob Storage for unstructured information, Queue Storage for large volumes of information, File Storage for shared resources and Disk Storage as HDD / SDD for companies.
Like AWS, it has a wide range of different types of databases such as SQL Database, Database for MySQL and PostgreSQL. In addition to NoSQL like TableStorage and Cosmos DB.
Google Cloud Platform was late, but Kubernetes has given it the necessary push
Since Google introduced Kubernetes within its own platform it has become a de facto standard. So one of their main business opportunities is in the extensive use of containers in the cloud rather than in the specific marketing of virtual machines.
In the case of Google Cloud Platform, Google has applied all its knowledge to scale and balance large system infrastructures, offering tools that are as similar as possible to those used internally. Thus, its main products stand out for the use of BigData, analytical tools or machine learning.
Google Cloud Platform is quite distant from its two main competitors due to its late entry into the IaaS business, its limited product catalog, its historical focus away from the business world or the fact that it does not have a data center infrastructure in so many regions like AWS or Azure. The typical use of GCP remains experimental or as a second provider to save costs on some secondary services.
As for cloud database, the highlight perhaps is Cloud Storage, its unified object storage system, whose success story is Spotify storing and serving music from there through its infrastructure. And as NoSQL databases we find BigTable and BigQuery widely used for BigData in Data Science.
The future of the next cloud leader will be who best uses application containers and serveless
All cloud service providers are converging on two key concepts that will set the trend in the sector in the coming years: the intensive use of containers (CaaS) to become independent from the platform where applications are deployed and the use of serverless, which It allows us to execute code directly without having to worry about infrastructure.
Both Google Cloud Engine, AWS EC2 Container Service or Azure Container service are already prepared so that the minimum unit of work is a docker container that we deploy in the provider's service on duty.
Google has gone even further with Anthos, based on Kubernetes, offering a service for the management, deployment and monitoring of the different cloud platforms that the user has, be it AWS, Azure, Google Cloud or their own infrastructure.
As for Serverless, AWS Lambda is one of the most popular FaaS (Function as a Service) implementations that we can find, although different providers have evolved to offer serious alternatives such as Google Cloud Functions, Azure Functions or Alibaba Cloud Functions.
The promise of a transparent and more productive cloud for developers is very close. Probably in the near future we can run away from the artificial cages of traditional service providers. Run our applications without marrying any cloud service.
