Automotive industry - Why Replatforming is crucial after Lift and Shift

 

Cloud Migration

With the global workspace undergoing a digital revolution, Cloud has emerged as the priced alternative to on-premise physical servers. Organizations are gravitating towards hosting their applications and services on Cloud for high security, to function in terms of pay-per-use model and flexibility.

Cloud migration is known as moving applications and data from an on-premise server to a cloud infrastructure. This transition from legacy infrastructure to a cloud platform could be either a full transition or partial.

Types of Cloud Migration


  • Lift and Shift – This approach is fruitful for migrating apps and the data stored in it to the cloud platform without making any modifications or redesigning the app. Organizations prefer to lift and shift model to sustain and protect their business and process workflows and data present in the on-premises infrastructure. This model helps organizations take steps towards modernizing IT solutions and creating adaptable architecture on Cloud.

  • Rebuilding – In this model, an application code needs to be re-written to make it compatible with the cloud environment. However, if the process of making the application compatible with Cloud or cloud-ready is too complicated, then the app is generally built from scratch.

  • Replacing – This model comes into play when an application doesn’t comply with the cloud environment and cannot be rebuilt or modified architecturally. In this scenario, SaaS (Software as a Service) offers all the essential features of the application hosted minus the infrastructure overhead costs, development, or management processes.

  • Re-platforming – This model is useful when the business application doesn’t comply with the Cloud for architectural reasons entirely or involves opting for the migration to facilitate app innovation and feature upgrades. In this scenario, the applications require few modifications or bare minimum code changes to ensure they make the best use of cloud infrastructure and are compatible.

  • Refactoring– This model involves moving applications to the Cloud after the re-architecting process, which involves modifying software and re-writing a major portion of the code. It helps the applications to become flexible and compatible with the cloud environment.

Lift and shift of a famous automotive industry first application



"Lift and shift" target architecture for an automotive industry with AWS

We have a N-tiers application with the web tier, the app tier and the database tier. There are few improvements that can be done in a cost, security, design, availability and performance perspectives.
In detail :
  • The web tier could be hosted in S3 behind a CloudFront origin to benefit from the cache improving cost and speed
  • The app tier could be serverless but it could require some refactoring. This is why we could propose some containers to fit the exploitation needs and enable a cost-effective scaling as we will add more applications like this one over the time
  • The database tier is handling failover strategies but what about read intensive requests and future scaling with more apps. Moreover, this is a monolithic app with a shared database for many backends and we didn't have the budget to refactor it. So we proposed a Serverless Aurora for handling scaling and read intensive requests.

Re-platformed solution

"Re-platformed" target architecture for an automotive industry with AWS

As we can see :
  • We created 1 bucket per frontend application
  • We hosted our backends in containers within some Kubernetes pods inside a cluster based on a node group of EC2. We could have choosen Fargate but our client wanted to install Centreon agents on underlying EC2 to monitor in his central dashboard on-premise
  • We used an Aurora Serverless single endpoint for each backend
Finaly, this re-platforming enabled us to grow and scale without having a huge impact on our costs, scalability, resiliency, availability which would have been the case with the initial architecture from the lift and shift strategy. As a conclusion, always think about re-platforming and if you can, refactoring application if you predict a significant growing to take benefit from the Cloud !

Comments

Most viewed posts

My tips for the AWS Solution Architect Associate certification

AWS Solution Architect Professional certification : the hardest one !

How to become an AWS Authorized Instructor