Just a while ago, AWS was organizing a thrilling, action packed, bootcamp (Delivery Best Practices) at their Berlin offices. The focus was on cloud adoption readiness, migration planning & execution. As an added bonus we also covered a topic everyone loves right now, serverless. I'll be going through some of the key things we covered and share them with you.
Berlin is an amazing city with so much to see. Unfortunately, we (that's me and our Cybercom penguin) only had a couple of hours to spare around the city. I really hope to go back there as soon as possible and have some more spare time.
The bootcamp itself is typically spanning over the course of three days, but due to time constrains we only had two. To be honest, that's quite tight schedule as there's more than enough topics to expand the whole bootcamp to even five days. Well, as said, we only had two and this naturally lead to quite long days and skipping some non-vital parts.
One of the biggest topics in the bootcamp was surely the AWS CAF, short for Cloud Adoption Framework. The CAF is a framework created by AWS Professional Services to guide organizations navigate through their cloud adoption journey, efficiently and effectively.
The framework is organized in six different domains, called perspectives, which are further divided into technical and business perspectives.
Source: AWS Website
As we spent quite a lot of time with CAF, I started to like it quite much. It covers all the needed aspects of cloud transition and operations reasonably well. I think this makes cloud adoption much more controlled and less intimidating.
The CAF with an imaginary company
So, you're currently running your stuff on-premise and have decided to start a project to migrate that to cloud. Great idea, but how should you plan and execute all that?
AWS has created many migration related best practices and strategies you could utilize to start planning for your cloud migration. The resources cover wide variety of migration aspects including:
- Business drivers
- Stages of Adoption
- Migration Process
- Migration Strategies
- Tools and services used
The migration process can be divided into five distinct phases:
- Preparation & Business planning
- Portfolio discovery & planning
- Application Design
- Application migration & validation
The migration strategies AWS is proposing are based on "The 5 R's" made famous by Gartner in 2011 with some modifications along the way.
Source: AWS website
The bootcamp included various smaller hands-on workshops. One of them was around cloud migration, where we planned a migration for apps of an imaginary company.
We were divided into groups and each group had to present their design and vision. Fun fun fun…
Without a doubt, serverless is a topic everyone is talking about right now and to be honest, they have a really good reason to boast about it.
The way AWS approaches serverless is called Lambda. Lambda was released at AWS re:invent 2014 and since then it's been taking the tech world by storm and evolving constantly. An example of how it’s been evolving, is the number of supported languages. Initially it started with node.js only, and today Lambda is supporting a ton of languages like python, C# and Go to name a few.
With Lambda you don't need to worry about compute instances, VMs or containers. The thing to concentrate on is functions, which are the unit of deployment and scaling. You're not paying for idle resources as you only pay for executed functions.
Source: AWS website
Typical use cases for Lambda tends to gravitate towards real-time stream processing, real-time file processing, ETL workloads, IOT & Mobile backends and web applications. Obviously, there are use cases where Lambda might not be the perfect fit. For example, Lambda might not be the solution for really low latency workloads. The same goes for really high payload workloads. I really suggest you read the Lambda FAQ before starting your serverless journey.
During the bootcamp we did some studies on various serverless architectures and techniques, but due to the time constrains, we had to skip most of the actual labs. That was really a shame as I really like working with Lambda.
As an interesting anecdote, we’re currently seeing a lot of companies who have missed the container-train completely and are skipping directly to serverless where appropriate. Makes sense to me!
Once again Berlin delivered wonderful experiences and massive amount of information. The bootcamp was a success! It was a shame, we couldn’t have the full 3-day workshop as I really would’ve liked to dig much deeper into the serverless architectures and do the full-blown cloud migration exercises. Maybe next time…
If you’re interested in cloud adaption, migrations or virtually anything cloud related, we at Cybercom are at your disposal. Get in touch with us.