Is AWS (Amazon Web Services) Over-Hyped?

Amazon Web Services (AWS) is a subsidiary of Amazon that provides cloud-computing services and application programming interfaces for its users. AWS offers its services on-demand, meaning that users only pay for the services they use. While the pricing of all AWS services is usage-based, each service may usage in its own way.

AWS dominated cloud services as of 2017, with a market share of 34 percent. The next three closest competitors are Microsoft, Google and IBM at 11 percent, eight percent and six percent respectively. Amazon markets AWS as a way to obtain large-scale computing capacity faster and cheaper than building a physical infrastructure. One of its distinguishing features is that AWS services are available to users of all sizes, from individuals to large enterprises. However, the disparity of this user base causes some critics to wonder if AWS can really deliver on its promise to meet the needs of so many different users.

Overview

AWS services generally provide users with a virtual infrastructure, distributing computing capability and tools for leveraging this platform. The AWS emulates the attributes of a physical infrastructure, including processing, memory and storage. It also includes software components such as operating systems (OSs), networking and application software such as databases, web servers and customer relationship management (CRM) systems. AWS offers a range of subscription plans depending on the type of infrastructure the subscriber wants, including a dedicated physical computer, a single virtual machine (VM) or clusters of either type.

The large number of services that AWS is one of the means it uses to meet the needs of its many users. The total number of AWS services in 2019 was 165, which included tools in the following cloud-computing areas:

• Application services

• Analytics

• Computing

• Database

• Deployment

• Developer tools

• Internet of Things (IOT)

• Management

• Mobile

• Networking

• Storage

Amazon Simple Storage Service (Amazon S3) and Amazon Elastic Compute Cloud (EC2) are two of AWS’ most popular services. Like most of AWS’ services, they provide their functionality through application programming interfaces (APIs) instead of exposing it directly to users. APIs are therefore critical components for AWS users, who access services with the Hypertext Transfer Protocol (HTTP). Older APIs use the representational state transfer (REST) architectural style and Simple Object Access Protocol (SOAP) for older APIs, while the newer ones use JavaScript Object Notation (JSON) exclusively.

Advantages

The sheer number of services that AWS offers is one of its greatest benefits for users. AWS’ dominance in cloud services also makes it particularly beneficial for large organizations. On the other end of client size, smaller organizations are more likely to benefit from AWS because they lack the technical expertise needed to maintain their own hardware.

Number of Services

Of the 165 services that AWS currently provides, very few AWS developers are likely to be intimately familiar with more than about 30. These services perform strictly cloud-based functions that deal with VMs, file storage and load balancing, but it also includes developer functions such as API gateways, databases and content delivery networks (CDNs). Additional AWS services perform analytical functions such as machine learning, data warehousing and identity management that provide insight into data, system performance and user behavior.

These services allow AWS to solve numerous problems common to many infrastructures, including developing REST frameworks, monitoring, caching, scaling and upgrading software. Most analysts find that the percentage of infrastructure problems specific to their organization is usually quite small, typically less than 20 percent. Client organizations can thus offload 80 percent of their infrastructure maintenance workload onto AWS, allowing them to focus on the remaining 20 percent of work that AWS can’t handle automatically. Of course, clients still need to perform their own cost analysis to ensure the AWS service fees are providing a positive return on investment (ROI).

Large Clients

The fact that AWS market share is significantly greater than that of its three largest competitors combined makes especially attractive enterprises that require a large infrastructure. These organizations are able to rent Amazon’s infrastructure rather than buying and maintaining their own, allowing them to pay for that infrastructure as an operational expense instead of a capital expense. This benefit alone adds a lot of value for large organizations.

AWS allows clients to run applications and store data faster, easier and cheaper, especially for large organizations. Renting infrastructure from Amazon also allows enterprises to save time since they no longer have to perform their own maintenance tasks such as processing, data management, application upgrades and security. The economy of scale that AWS provides drives down the unit cost of each server, meaning larger organizations save more money by using AWS.

The advanced security features of AWS and the large number of security experts that it keeps on staff makes AWS VMs highly secure. This is an increasingly important advantage for organizations that keep payment information on millions of customers. The well-publicized data breaches of corporations like Target, Sony and Home Depot illustrate the value of this information to criminals. All of these organizations maintained their own servers, and it’s highly unlikely that these same attacks would have succeeded on AWS.

AWS’ reliability is also of particularly great benefit for enterprises. It maintains copies of applications and data in multiple data centers that are geographically separate from each other. Client information thus remains safe even when an individual data center is affected by a natural disaster or manmade attack.

Hardware Independence

The ability of AWS to free clients from the responsibility of maintaining their own server farm is a great advantage, especially for smaller organizations. AWS clients can also easily scale their infrastructure as their operation grows, which would require the purchase of new hardware for an on-premises data center. They only need to pay the required recurring cost, and AWS will build the necessary hardware needed to support a client’s requirements. This advantage is particularly useful for a tech startup with a high requirement for hardware.

Disadvantages

The limitations of AWS include factors that generally apply to all cloud services, while others are disadvantages of AWS compared to its competitors. These include the following:

• Technical support

• Loss of control

• Resource limitations

• Integration

• Containerization, analytics, and scaling

Technical Support Fees

AWS fees for technical support tend to be significantly higher than those of either Microsoft Azure or Google Cloud Services (GCS), especially as an organization’s size increases. It provides three tiers of support, including developer, business and enterprise. The developer tier is for individuals, the business tier is for small and medium businesses, and the enterprise tier is for large businesses. The support fee for developers is a flat rate of $29 per month.

However, the rates for the higher tiers can be much higher, since they’re based on the usage of AWS resources. For example, the business support fee starts at $100 per month, assuming your month AWS usage cost is no more than $1,000 per month. The fee then increases to ten percent of monthly usage costs between $1,000 and $10,000. This rate then drops as the monthly usage increases, with the top end being three percent for monthly usage costs over $250,000.

For enterprise customers, AWS technical support is $15,000 per month if the usage fee is no more than $15,000 per month. The support fee then changes to seven percent of the usage fee if it’s between $150,000 and $500,000 per month, and five percent if the usage fee is between $500,000 and $1 million per month. The top technical support pricing tier for enterprise customers is three percent for usage fees over $1 million per month.

In contrast, Azure offers technical support for a flat rate based on the service tier. It’s free for the Basic tier, $29 per month for Developer, $100 per month for Standard and $1,000 per month for Professional Direct. The Premier tier for Azure requires custom pricing on customer support. Google Cloud charges per user for its support plans. The Basic plan is free, while the Development is $100 per month per user. The Production plan is $250 per month per user, and the Premium plan requires custom pricing.

Loss of Control

The other side of the coin for offloading infrastructure management to AWS is the subsequent loss of control over that infrastructure. AWS isn’t going to be a good fit for an organization that needs to keep its data in-house, typically due to regulatory compliance or other special security requirements. Very large companies with the necessary resources may be better running their own private cloud by using solutions such as VMware, Red Hat Enterprise Linux (RHEL), Docker, Kubernetes and OpenStack. Furthermore, AWS may be overkill if all you need is a website and email.

Another reason to consider alternatives to AWS is the risk of vendor lock-in, which occurs when a client commits to a proprietary platform like AWS. A client in this situation may only be able to use AWS solutions, which could limit their capabilities down the road. Businesses that are already using these tools may find it disadvantageous to give them up, even when it means gaining the benefits of AWS.

Resource Limitations

AWS sometimes sets default limits on resource consumption, typically based on geographic location. This is most obvious with Amazon EC2, which has limits on images, snapshots and volumes. However, you can request an increase on your limits. AWS also provides only limited information on the resources that Amazon EC2 manages, which can make it difficult to identify the cause of a performance bottleneck.

Integration

One of the most common reasons for choosing Azure over AWS is Azure’s tight integration with other Microsoft applications gives it an advantage in this area over AWS. Many enterprises already use Microsoft OSs and other applications, giving Azure a strong advantage when these organizations move to the cloud. Microsoft also rewards brand loyalty with significant discounts off Azure services for its existing customers.

Containerization, Analytics and Scaling

Containerization, analytics and scaling are some of the reasons to choose GCS over AWS. Google is currently leading the trend towards containerization since it developed the Kubernetes standard that AWS and Azure also offer. GCS is also likely to be a better choice than AWS for clients that place a priority on analytics capabilities, including Big Data and machine learning. Other areas where GCS wins over AWS include scaling and load balancing, which are particularly important for clients that value response times.

Summary

AWS has a reputation for being the best overall cloud services provider that most experts feel is well-deserved. Its many services and dominant position in the cloud services make it unlikely that AWS will lose to its competitors any time soon. For customers who can afford its technical support, AWS is often the best choice when you need to move your operations to the cloud. If you are looking for a company experienced in AWS Migration consulting, Baytech can help you on your way.