Choices: AWS v Azure v Google Cloud
Cloud services are now an accepted, well used feature of nearly every organisation. It was estimated that even in 2019, 94% of all organisations used either public or private clouds (Flexera) and cloud usage has grown significantly since then. In one form or another, whether it be simply the use of SaaS, almost all organisations now use the cloud. The following article provides an overview of the differences between each of the three main cloud providers, AWS (Amazon Web Services), Microsoft’s Azure and Google Cloud. For more information, and what to consider when deciding whether or not to move existing services to the cloud, see our recent article “Should you be moving your software to the cloud?“.
A public cloud is one made available to many customers by a cloud provider. It runs on remote servers managed by the provider, access being over the internet. It shares computing services amongst those customers even though their data is kept separate from each other. A private cloud, on the other hand, is not a service shared with other customers. The responsibility for management and maintenance of the data falls to the company, the owner of the data. Using the public cloud means that your data is stored in a provider’s data centre, they being responsible for the centre’s management and maintenance. A hybrid cloud environment is one that extends a secure private cloud to a public cloud when a demand for additional resources arises, without the risk of an internal overload. In other words, a mix of the two.
The advantages of the public cloud include resources not having to be used to manage servers; stronger security is available to small and medium sized enterprises (i.e. those with smaller budgets); and there is an opportunity to reduce IT costs by outsourcing to an efficient provider. Economies of scale also mean public clouds often cost less to run than private clouds for individual users. The downside can sometimes be with compliance and security: some organisations need to adhere to strict rules, one example being healthcare organisations in the US with HIPAA 1996.
The most important advantage for many is that a public cloud strategy allows an organisation to grow at an almost infinite rate (although this is possible with an on-premises data centre it can be expensive) without the need to buy more hardware, nor with the responsibility to maintain a growing network.
Fortune Business Insights is forecasting significant growth in the global cloud computing market size over the coming years, starting at US$ 219bn in 2020 increasing to US$ 791 in 2028, a compound annual growth rate of 17.9% over the period 2021-2028.
The market leaders
The three leading public cloud providers are AWS (Amazon Web Services), Microsoft Azure and Google Cloud.
Choosing between the three (or others, such as IBM, Oracle etc.) is not only about pricing, one needs also to consider, for example, the size of organisation and its projected growth rate in addition to the budget available. There are also technical differences, which means that before making a decision you would be well advised to refer to an expert – a significant amount of investigating is time well spent when about to invest in a cloud partner
Market shares and growth rates
Amazon Web Services (AWS) was launched in 2006, fundamentally transforming the IT industry and becoming the market leader, in 2020 with a 40.8% share of the IaaS (Infrastructure as a Service) market according to Gartner. Following behind, but growing rapidly, are Microsoft, Google and Alibaba. Gartner predicts worldwide end-user spending on public cloud services will jump from $242.6bn in 2019 to $692.1bn in 2025, attaining a 16.1% compound annual growth rate.
AWS’s biggest strength is that it dominates the public cloud market – being the most mature enterprise-ready provider, with the deepest capabilities for governing large numbers of users/resources. Azure and Google are behind Amazon, but they are rapidly narrowing the gap.
It should not be difficult for an organisation to access the services of any of the main providers, although each has a different structure and terminology.
The AWS Cloud spans 84 Availability Zones within 26 geographic regions around the world, with announced plans for 24 more Availability Zones and 8 more AWS Regions in Australia, Canada, India, Israel, New Zealand, Spain, Switzerland, and United Arab Emirates (UAE).
AWS is hosted in multiple locations around the World. They are known as “Regions” and “Availability Zones”. Each Region is a defined geographic area, and is completely independent and designed to be isolated from other Regions. This achieves the greatest fault tolerance and stability. Each Region then has multiple, isolated locations known as Availability Zones (AZ). Each AZ can have multiple data centres (typically 3, and can consist of hundreds of thousands of servers), each one having their own power infrastructure, each AZ physically separated with a meaningful distance between them, up to 100km.
AWS not surprisingly has a larger global footprint than any other cloud provider, however its competitors are catching up.
Azure operates in multiple datacenters around the world that are grouped in geographic regions. Within each region, multiple datacenters exist to provide for redundancy and availability, allowing flexibility to design applications to create virtual machines closest to users and to meet any legal, compliance, or tax purposes.
An Azure “Geography” is a discrete market, typically containing two or more Regions that preserve data residency and compliance boundaries. This allows a customer’s data and applications to be kept close by, if this is required. They are designed to be fault tolerant, to withstand complete region failure. “Availability Zones” are physically separate locations within an Azure region. Each Availability Zone is made up of one or more datacentres equipped with independent power, cooling, and networking.
Like Azure, Google Cloud is also expanding rapidly. Its resources are hosted in multiple locations known as Regions and Zones. The former is a specific geographical location containing one or more Zones (US-West1 has three zones, for example, US-West1a, US-West1b and US-West1c), each containing its own resources. Again, choice of location is available, and using multiple zones and regions will tolerate outages. Each Zone is isolated from other zones, and most single failure events will therefore affect only a single zone.
Who uses each provider?
As you would expect, with AWS being the oldest and biggest player in the cloud market, it has a bigger community support and user base. Therefore AWS probably has more high-profile and well-known customers, such as Netflix, Airbnb, Adobe, Hitachi, Johnson & Johnson, Capital One, etc.
However, with the multicloud it is fair to say all the major providers have household names that also overlap. Azure has and is also gaining more high-profile customers over time. The majority of Fortune 500 companies in the US use Azure, Microsoft itself stating “more than 95 percent of Fortune 500 companies trust their business on Azure today”. Some of its major customers include BP, Intercontinental Hotels, Uber, Johnson Controls and BMW. In the UK, as well as major commercial enterprises, Azure is widely used by numerous Government departments including the MoD, NHS and HMRC.
Google has many high-end companies now using Google Cloud. Major clients include Twitter, Deutsche Bank, Goldman Sachs, PayPal, P&G, Airbus, UPS, and more.
Services from each provider
Amazon has over 200 services for cloud applications. They can be broadly categorised as follows:
• Databases that offers security and scalability
• Compute services that enable developers to build and deploy applications
• Security tools
• Developer tools
• Management tools that enable users to control costs, minimise risks and improve efficiency
• Networking and delivery of content
Not surprisingly, Azure offers similar services to Amazon. Its computer services cover IaaS, PaaS as well as Web Apps. In addition, there are mobile, storage, and communication services as well as data management, messaging and media services. Azure ML also provides tools for developers to create machine learning and artificial intelligence services.
Google Cloud has over 150 products ranging from Compute Engine (virtual machines running in their data centres), through to storage, operations and data flows.
How do they compare with pricing?
Making a price comparison is difficult as each provider changes their prices frequently and new products as well as improvements are being continually introduced. Pricing models range from a subscription-based model that is billed per month/mailbox/licence (irrespective of whether resources are used or not), through to pay-as-you-go, allowing subscribers to pay only for the resources that they use.
AWS has a reputation for complex billing: many enterprises find it difficult to understand their cost structure and hence manage those costs. The platform does however provide its customers with a cost calculator. Because the pricing structure is complex, some assistance will probably be necessary to navigate through the options available and thus enable costs to be controlled.
Azure also has its own pricing calculator to help potential customers understand what various combinations of services will cost.
Pricing is one area where Google does try to differentiate itself, by making its pricing structure more transparent and customer-friendly. It strives to undercut the list prices offered by most cloud services providers and make available discounts and other incentives to win business.
It is important to choose the right cloud vendor. It’s about choosing the best-suited cloud provider according to the needs of your own organisation.
AWS through its sheer size and the length of time it has been around wins on establishment, availability, market share and number of services. It brings with it a stable, reliable service with a long track record. It has a wide range of services and is particularly suitable for larger companies.