Problem overview

The main challenge consisted in migrating a large number of legacy applications, which ran on very outdated and unsupported operating systems, but which couldn't be simply updated due to functional dependencies on those operating systems. At the same time, the variability between servers made each team dependent exclusively on one person.

Solution overview

We removed the dependencies between the applications and the base operating system and we implemented a private cloud based on Kubernetes. We also standardized and automated the way in which applications were deployed over the private cloud. With this, we managed to disable a large number of physical servers, eliminating the dependency between each server and a technician, and substantially speeding up the deployment of applications.

Detailed solution

After evaluating the organization’s situation, not only in relation to the project but also thinking about its future management, we decided that the best solution was to create a private cloud and to move there all the applications and services that, up to then, had been provided on the servers to be replaced.

For this, we divided the work into two parts:

• Containerize applications so that all their dependencies would be self-contained.
• Design and implement the IT infrastructure that would constitute the private cloud.

In order to reduce times, we created two work teams to make progress in parallel with the development of the solution.

We developed the private cloud with Kubernetes on VMware Vsphere, which is the organization’s virtualization platform, and its generation was fully automated using Terraform, Ansible and Cluster API. This allows it to be replicated and restored quickly and identically.

On this infrastructure, with fully updated operating systems and libraries, we deployed containerized applications.

In addition, we also:

• Developed CI/CD pipelines, which allow anyone on the team to deploy the applications, as opposed to the previous need of an expert for each system.
• Installed a monitoring stack based on Prometheus, Grafana and Alert Manager to have a view of the historical and present states of the infrastructure, and to receive alerts in case of problems that deserve attention.