Sergey Frolov
X5 Tech
Apelsin began as a startup within a corporation, then grew into an independent company and eventually returned to X5 Tech. Today, it is a fintech service with a substantial workload, a large team, and numerous integrations.
As the product evolved, so did its infrastructure. The team completed five migrations driven by increasing workloads, information security requirements, outdated technologies, and corporate changes. In this talk, I will walk through these migrations and explain the engineering and organizational challenges of moving a live product to a new environment.
I will explain how IaC, Helm, GitOps, and Argo CD helped us standardize our infrastructure and reduce migration time from approximately three months to one. I will also cover traffic switching, migrating a production database within a limited downtime window, and preparing a rollback plan.
Using real incidents, I will demonstrate why teams need infrastructure code reviews, environment documentation, checklists, and up-to-date technical documentation.
The sixth migration does not have a confirmed date yet and should become part of our Disaster Recovery Plan, so the infrastructure must remain ready for the next stage. The central idea of the talk is simple: a successful migration should not require heroics but should become a predictable and deliberately uneventful engineering task.
X5 Tech