Overview
MyHeritage, a global genealogy and DNA platform serving tens of millions of users, operates Legacy Family Tree Webinars, a high-traffic WordPress-based platform that delivers live and recorded educational content to genealogy enthusiasts worldwide. The platform was running on traditional infrastructure that struggled to handle traffic spikes during popular webinar broadcasts and lacked modern deployment practices.
The Challenge
Legacy Family Tree Webinars had outgrown its hosting environment. The WordPress platform needed to handle unpredictable traffic surges during live webinar events, maintain high availability for a global audience, and support a content library of thousands of recorded sessions. Manual server management was consuming engineering time, backups were unreliable, and deployments carried significant risk with no automated rollback capability.
Live webinars attract thousands of concurrent viewers. Any downtime during a broadcast means lost audience engagement and direct revenue impact. The migration had to be seamless, with no interruption for viewers.
Our Solution
Dcode migrated the entire Legacy Family Tree Webinars platform to AWS, containerizing WordPress on ECS Fargate to eliminate server management overhead. We designed a highly available architecture with Amazon EFS for shared WordPress filesystem storage across containers, Amazon RDS MySQL in a Multi-AZ configuration for database high availability, and ElastiCache Redis for session management and object caching to reduce database load during peak traffic.
For the deployment pipeline, we built a fully automated CI/CD system using AWS CodeBuild and AWS CodePipeline. Every code push triggers an automated build, test, and deploy sequence with blue/green deployments on ECS for zero-downtime releases. AWS Chatbot was integrated to send real-time notifications to Slack for every pipeline stage: build started, tests passed, deployment completed, or any failure. This gave the team immediate visibility without checking the AWS console.
Automated backup pipelines were built with AWS Lambda and Amazon EventBridge, running on schedule to snapshot the database and filesystem to S3 with lifecycle policies for cost-efficient long-term retention. The backup pipeline also reports status to Slack via AWS Chatbot, so the team knows immediately if a backup fails. All infrastructure was codified with Terraform managed through Spacelift, and AWS SSO was configured for secure team access to the AWS environment.
The combination of CodeBuild, CodePipeline, and AWS Chatbot creates a fully observable deployment workflow. The team receives Slack notifications for every step: source change detected, build in progress, tests passing, deployment rolling out, and deployment complete. Any failure triggers an immediate alert with a direct link to the pipeline logs.
Results
- Seamless migration to AWS with zero user-facing downtime during live webinar broadcasts
- Fully automated CI/CD with CodeBuild and CodePipeline for push-to-deploy blue/green zero-downtime releases
- Real-time Slack notifications via AWS Chatbot for deployments, backups, and pipeline failures
- 100% infrastructure as code with Terraform and Spacelift, making environments reproducible on demand
- Automated backup pipelines via Lambda + EventBridge with S3 lifecycle policies and failure alerting
- ECS Fargate with ElastiCache Redis and Multi-AZ RDS handles webinar traffic spikes without manual intervention