Amazon AWS Review

Accelerates innovation through experimentation cycles in a scalable platform

What is most valuable?

Accelerates innovation through fast experimentation cycles in an agile, flexible, and scalable platform.

How has it helped my organization?

Enables fast prototyping, simulation, and rapid deployment of infrastructure configurations. Has low risk exploration of new architectural paradigms and technologies (FaaS, Containers, IoT, and Machine Learning) and is easy to integrate with current solutions.

What needs improvement?

Considering the rate of innovation of AWS and the vast range of services offered (over 15+ categories, 50+ services in 2017) the learning path of customers on the platform is something that can always be improved. Usability through simplification of the interface for the use cases chosen by the customer can be a possible improvement.

The current interface offers several options to select services, solutions, or learning paths. However, the ability to simplify the interface to focus on customer use cases could have an impact on productivity and ease of use.

This is a challenge that I’ve seen all cloud vendor share: Usability and different user experience on their platform is difficult when the span of services is so vast. However, some design thinking “persona” kind of approach could help offer alternative perspectives.

For how long have I used the solution?

  • Since 2012, in prototypes and proof of concepts
  • Since 2015, in production applications, advise, and support to some clients.

What do I think about the stability of the solution?

I’ve never experiences issues with stability related to the AWS infrastructure. The services are very resilient and there are constant reporting and monitoring tools available, a open status dashboard, and a personal health dashboard to receive news on any issues being investigated or sorted out. Even if there have been outages reported in AWS history their technical response capabilities have proven outstanding.

What do I think about the scalability of the solution?

I’ve never experience issues with scalability. AWS services offer very flexible set of tools to architect solutions that give the best performance and economic advantages. Combined solutions using elastic computing capabilities, containers, APIs, and even more innovative server-less capabilities (FaaS) can be leveraged to tackle the most challenging use cases.

Which solutions did we use previously?

I previously favored RackSpace and Digital Ocean for simplicity and focus for certain use cases (development prototypes, proof-of-concepts, etc.). I prefer to concentrate investment and training on the same platform when solutions scale and require more complex setups. Leveraging the learning curve on the service offering is increasingly specialized.

How was the initial setup?

The setup is easy and greatly supported by the learning paths offered through the platform. Expertise is required to take full advantage of AWS tools and continuous innovations.

Some customers can become overwhelmed by the range of services, so training and assistance from specialized third-parties is strongly recommended. Even experimented managed service providers can complement internal capabilities and help in the training of internal teams.

One of the advantages of AWS is their high rate of innovation. However, in order to leverage this, internal or external expertise is required. A good partnership is recommended.

What's my experience with pricing, setup cost, and licensing?

On demand, pay-as-you-go pricing is powerful to optimize expenses, but it’s important to keep a technical cost controlling function aware of usage and scale patterns to choose the best pricing mix.

Massive migration to cloud without analyzing the right service for the right usage can lead to higher cost than expected. It is important to get the right advice to match each use case needed to the optimum cloud economics.

Even if a lot of decisions to go to the cloud are based in the promise of lower costs, the true power of cloud services is their flexibility, rate of innovation, and avoiding vendor lock-in if architected consciously.

Even if a lift and shift approach with short schedules can lead to mistakes in choosing services and paying more than optimum, the speed in which you can correct the mistake is not comparable to any other infrastructure option.

This is forcing even the traditional hardware vendors to reinvent their business models and develop financial offerings that include operating expense based financing (pay-as-you-grow) or services based agreements (pay-as-you-go) to make their private cloud offerings competitive.

The other aspect to consider is the managed service required to get the most of this platform. Don’t underestimate the quality of the advice and support required. But at the same time, consider your core business management time released by adopting a platform instead of managing the components internally.

The internal expertise should evolve to understand how to use it best for the business outcomes pursued instead of the technicalities of how to make it. That’s where the right partnerships can be leveraged.

Which other solutions did I evaluate?

We evaluated Azure, RackSpace, Google Cloud, SoftLayer, DigitalOcean, and Linode.

What other advice do I have?

Test drive it with prototype applications, reproduce development and testing environments, and standardize your stacks to be able to move them easily, if needed. The deeper that the infrastructure-as-code approach is part of your culture, the easier it will be to leverage hybrid opportunities and gain agility.

This solution has been consistently in the top of the IaaS market for the last 10 years.

Disclosure: I am a real user, and this review is based on my own experience and opinions.


13 July 17


13 July 17

Sign Up with Email