Legacy Cloud was not enough
Plytix is the all-in-one Product Information Management platform for Omnichannel Marketing. It combines PIM with Digital Asset Management (DAM), Product Data Syndication, Catalog Creation Software and Analytics. Affordable, easy to use and with an amazing Customer Support team.
Plytix was already working in AWS cloud, but using a classical approach, nowadays is called Legacy Cloud. Everything was running on Virtual Machine, so they were able to scale up and down when it was required, but the time response and the cloud resource utilization were far from being optimal. They were going to grow and they knew that approach was not enough therefore, we tried to establish some foundation for the upcoming cloud infrastructure, which would be based on Kubernetes.
CONTINUOUS INTEGRATION AND CONTINUOUS DEPLOYMENT
The first step was focusing on the Continuous Integration and Continuous Deployment processes (CI/CD). Applications had to be containerized and image creation should be automated. It implied to set up a right environment where managing the code, integrate building actions and running test were scalable, easy to use and as independent as possible from third parties.
Next step was infrastructure related. We developed a system, based on Terraform, to provide the necessary platform on demand (Kubernetes cluster with the necessary cloud resources). The system helped to define the infrastructure as code (IaC), taking advantage of a declarative approach, and offered the possibility of deploying platforms in the same way for production, staging, testing and so on. Having a platform, based on Kubernetes, decoupled the applications from the cloud provider in many ways, but especially, in the deployment approach, getting unlocked from the cloud provider.
Once the platform provider system was ready, we worked in a flexible way of deploying applications and the necessary tooling around them. We used Helm to builf it, tacking advantage of their templating system and therefore, being able to deploy the entire system into the desired platform but adapting it to the chosen environment (production, staging, testing,…). This system only needed a ready and accessible Kubernetes cluster to work with.
Automation has accelerated the delivery where the same code is tested in a development phase and promoted to different stages up to reaching production. Now, Plytix tech team is more aware of how their infrastructure works. They own the necessary tooling to observe, troubleshoot and improve their applications.