Cloud Computing is pretty much taken for granted in 2021. It is everywhere you look. From Software as a Service (SaaS) like email and calendars to Platform as a Service like SAP, Salesforce or Google App Engine to Infrastructure as a Service (IaaS), including virtualized compute, storage and other services, if you have a need, somewhere in the cloud there is a solution.
The COVID pandemic has spurred an even greater adoption of the cloud. More remote work means more virtualization. As enterprises learned that remote is a real possibility, there became less of a reluctance against pushing ever more workloads into the cloud. As someone who has been in this business for a long time, I can say the push to cloud migrations has been huge in the last year. That is one of the reasons cloud and data engineers have become so in demand.
When you choose a provider, you want to choose the best. How do you decide which is the best? Now that is completely dependent on your needs and budget. If all you need are some web servers and a database, there are many who offer many different kinds of pricing. However, most enterprises need more than a service or two.
An enterprise will typically look for a few great and fast needs in a cloud provider:
- The environment must be secure.
- The platform must have a choice of database.
- It must have the ability to run compute-intensive workloads.
Learn more:Â More Business Leaders Are Betting on Cloud Amid Elusive ROI: PwC ReportÂ
The Major Players
How you judge a significant player will depend on who you talk to and the criteria you use. Regardless of the requirements, you will generally see the same few players, usually in the same order, with an occasional provider offering a stand-out service but lagging overall.Â
Generally speaking, the top three providers will be Amazon, Microsoft, and Google (in that order). Some ordering of Alibaba and IBM will usually follow these. Some analysts believe Alibaba will overtake Google (which is still losing money) and become the 3rd largest cloud vendor. If you look at specific offerings, you will sometimes see Oracle, Salesforce, SAP, and Tencent.
Of the big three, Amazon and Microsoft are usually quite a bit ahead of Google. While Google has some excellent offerings (BigQuery especially), this post will look at the two leaders (by a large margin) – Amazon and Microsoft. They have much in common but just as much in terms of difference. Their history defines their approach to the cloud and their journey to hyperscaling.
Amazon’s History
Amazon started as an electronic book store from Jeff Bezos’ garage. In the early 2000s, as Amazon grew, a hard and fast rule around automation started. Everything must be automated and automation requires code. The team there decided early on that the application needed to be decoupled from infrastructure and so they adopted a deep API and web service mindset.
Automation was required for the teams to explore and extend without bottle-necking on infrastructure upgrades. This led to application and deployment automation and infrastructure automation (which in itself led to infrastructure as code).
By 2005, Amazon was utilizing tools and platforms that would eventually lead to Amazon Web Services (AWS). In 2006, Amazon officially launched AWS to specific customers with storage (S3), compute (EC2), and simple queuing (SQS). In 2007, AWS was launched publicly, and it has iterated and grown in its offerings ever since.
Amazon has continued to add features, tools and new services (many adopted from open source) as fast as they can. If a customer asks for a service, Amazon adds it. Sometimes Amazon will offer several different tools to accomplish the same functionality. More on that later.
Microsoft’s History
I won’t dwell on Microsoft’s history. If you are not familiar with MS and where they come from, you probably wouldn’t be reading this. I do want to point out the origins of their cloud service, Azure.
While Amazon grew out of a need to support the burgeoning Amazon web store, Azure was a planned and methodically designed solution for enterprises. Azure was announced (2008) not long after AWS, but the first Azure service was not available until later. Azure SQL Database became available in 2009 but was not available commercially until 2010. Cloud by committee.
Learn more: Cloud Wars: Azure Closes the Gap on AWS with 50% YoY GrowthÂ
Major Differences between AWS and Azure
How AWS and Azure differ in key features, market share
When considering the differences in cloud providers, especially providers of scale-like Microsoft and Amazon, you don’t need to waste time on the commodities. Virtual CPU is pretty much Virtual CPU. Object storage is object storage. Containers are containers. All providers offer a choice of OS, and all providers offer a selection of relational databases. While those are important to an enterprise, they are unimportant in choosing your provider (as long as they are offered and any provider of any reasonable size will have those).
No, the essential points to consider when choosing your cloud provider are how well the provider meets your expectations in services (i.e., does the platform support you), do they meet your budget, and if you need big data to support, does that support meet your needs.
For budget, Azure and AWS are competitive, and which has the best prices fluctuates over time and need. If you plan to employ this kind of hyperscale cloud platform, the price will be competitive.
For the platform, this is where history comes into play. AWS is a jungle of services and tools. Azure is a platform designed by a committee. While both offer closed and open source options, Azure is more of a guided tour that holds your hand.
Amazon created many of its solutions in the early days and offered those. Then over time, as customers needed (or wanted) specific solutions, Amazon started offering more open source solutions. As the Amazon engineering team grew, they started incorporating open source solutions, making them amazon specific, and creating entirely new solutions.
Azure is planned. When you use Azure, there is a much more definitive flow and logical progression through the platform’s services. They do offer open source, but you are more likely to end up in a Microsoft-designed solution.
AWS offers choices, maybe too many. It is sometimes difficult to see a “best†way forward, and you almost randomly choose one of the solutions available. Azure has a way to do almost anything you want to do, but it can sometimes feel constricted on choice. Neither method is better; you need to pick the one that fits your culture better. I lean towards too many over the limited choices.
When it comes to big data, the platforms follow the same pattern. The real differentiating factor here is the analytic warehouse/database. Azure offers Synapse Analytics, and Amazon offers Redshift. I have to give the point here to Azure currently. Even though they were the leader, Amazon is playing catch-up. Snowflake and Google BigQuery now have the best offering, IMHO. Synapse is just a step behind, and Redshift is a few steps behind.
Redshift has been improving, though, so it is not a done deal. Now that Redshift is moving towards separating computing from storage (with RA3), they are moving back into a competitive landscape. It’s still a managed landscape and not serverless, but they may move back into the leader position if they keep bringing costs down.
Microsoft won’t stand still and let that happen, though. Microsoft is a database company (and OS, tools, office, etc.), and they have the engineering team to push Synapse Analytics beyond even Snowflake. I am keeping my eye on them.Â
Pros and cons of AWS and Azure
One final note on Azure over AWS:Â
Microsoft tends to partner with other software providers and truly integrate with them. An excellent example is the Databricks offering. On Azure, databricks feels like a built-in tool. AWS does partner, but it always feels additive and not native. It’s just a part of the Azure culture that your organisation’s culture may be concerned with or not. Databricks works just fine on either platform, but it feels more tightly integrated on Azure. This is true for most integrations, at least in my experience.
Learn more: Keeping Up With the Cloud: What Companies Can Learn From Amazon’s Leadership Shift
Summary
Choosing a cloud provider is not an easy choice or should be made without a lot of thought and due diligence. In the early days of cloud computing, choosing a provider required you to make sure your choice of OS or database would be included at a price you could afford. In 2021, cloud computing is a commodity offering. Prices are competitive, and all of the larger providers will give you all of the standard commodity service offerings. It’s more important to make sure the platform matches the culture your environment requires.
The simplest way to look at it is like this: does your company value a holistic solution over a plethora of options. Both platforms will get the job done, and they are priced competitively. Do you prefer the wild west of AWS where you build a platform from a Lego-like stack of parts, or do you like to order from a menu and buy the meal plan to get the meal, dessert, side and drink all in one?
What are your thoughts on the AWS and Azure rivarly and who has an upper hand. Comment below or tell us on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We would love to hear from you!