Aztec Logo
Migration Completed

WordPress to Kubernetes Migration

From fragile VMs to auto-scaling containers

Complete migration of multiple WordPress sites hosted on EC2 instances and VMs to a Kubernetes cluster, achieving automated backups, local environments identical to production, centralized monitoring, and intelligent routing.

aws, cloudflare
8 weeks
2024
2 engineers

The Problem

Multiple WordPress sites spread across EC2 instances and VMs with no standardization. Manual and inconsistent backups, no local development environment, deployments with downtime, and zero visibility into site health. Each server was a snowflake impossible to replicate.

The Solution

Containerized each WordPress site with Docker, defined Helm charts for declarative deployment, migrated to an EKS cluster with EFS for shared storage, implemented Ingress with automatic certificates, and configured monitoring with CloudWatch and alerts via SNS. Every developer can spin up the complete environment locally with a single command.

The Results

Hourly automated backups, local environment identical to production, zero-downtime deployments, centralized monitoring of all sites, and ability to horizontally scale during traffic spikes.

Measurable Results

Deploy time

45 min

10 min

78% improvement

Recovery time

2-4 hrs

10 min

97% improvement

Local env setup

1-2 días

~0 min

99% improvement

Sites migrated

12+

Want results like these?

Let's scope your project — 30 min, no commitment.

Schedule assessment

Project Phases

Audit & inventory

1 week

Inventory of all sites, PHP/WordPress/plugin versions, databases, and server dependencies.

Containerization

2 weeks

Creation of optimized Dockerfiles, multi-stage builds, and docker-compose configuration for local development.

K8s infrastructure

2 weeks

EKS provisioning, EFS configuration, Ingress Controller, cert-manager, and Helm charts.

Migration & DNS

2 weeks

Data migration, DNS cutover with zero downtime, and validation of each site.

Monitoring & hardening

1 week

CloudWatch setup, SNS alerts, pod and node metrics, and network policies.

Tech Stack

Technologies

kubernetesdockerwordpressnginxmysqlhelmphp

Cloud Services (AWS, CLOUDFLARE)

EKSEFSCloudflare DNSCloudFrontECRRDSCloudWatchSNS

Tools

helmgithub-actionskubectl

Implementation Details

Architecture

The migration transformed an infrastructure of independent servers into a standardized Kubernetes cluster where each WordPress site runs as a pod with its own declarative configuration.

Before

  • Individual EC2 servers per site
  • Manual backups with cron and bash scripts
  • No local development environment
  • Deployments via SSH + rsync

After

  • EKS cluster with per-site Helm charts
  • Automated hourly backups to S3
  • docker-compose up for full local development
  • CI/CD with GitHub Actions → Helm upgrade

Have a similar technical challenge?

Let's talk about your infrastructure, architecture or pipeline. No commitment.

Schedule a Technical Assessment