Skip to content

Customer story · Ecommerce infrastructure

HomeTiger: modernizing infrastructure around a growing marketplace.

CloudHero helped HomeTiger review its infrastructure, plan container-based application delivery and migrate business-critical search to AWS. The work connected platform engineering with the reliability of the shopping experience.

Workspace pictured in the original HomeTiger case study
From HomeTiger’s original project story.
Business
Home and garden marketplaces
Work
Assessment and modernization
Cloud migration
Elasticsearch on AWS

The challenge

Modernize the platform without losing sight of the marketplace.

HomeTiger builds home and garden marketplaces under the Moebel24 and HomeTiger brands. Its infrastructure has to support product discovery across a large catalogue while the engineering team continues to deliver application changes.

The original engagement began with an infrastructure assessment. HomeTiger was operating across Hetzner and Google Cloud, using a mixture of self-hosted and managed services. It wanted a clearer view of performance, availability and security, alongside a more consistent approach to deploying its applications.

CloudHero reviewed the platform, infrastructure provisioning, observability, CI/CD, backup strategy and security. Looking at these areas together matters: a faster deployment process is of limited value if recovery, credentials or operational visibility remain fragile.

Engineering decisions

Turn the assessment into a practical modernization plan.

The assessment recommended containers and Kubernetes for application deployment, with Helm for packaging releases, upgrades and rollbacks. It also identified work in the surrounding infrastructure. The table separates those recommendations from the later AWS migration described below.

Priorities from the original infrastructure assessment
AreaRecommendationEngineering value
Application deliveryContainerize workloads and use Kubernetes with Helm.Make application packaging and release configuration easier to repeat and review.
Infrastructure codeRefactor Terraform into maintainable modules.Make shared infrastructure changes easier to understand and reuse.
Operational visibilityMonitor network failures in data pipelines and hardware health; centralize consistently formatted logs.Give engineers a clearer starting point when investigating a failing service or data flow.
Recovery and secretsEstablish a backup strategy and centralize secrets management.Make recovery and credential handling explicit parts of the operating model.

“CloudHero transformed our infrastructure, boosting performance, security, and cost efficiency.”

Guilherme Braga, CTO of HomeTiger · original CloudHero case study

AWS-published migration

A focused search migration with a documented outcome.

In a separate AWS-published HomeTiger story, CloudHero migrated Elasticsearch to Amazon EC2 within a VPC. AWS reports less than three months of preparation, a 30-minute cutover with no downtime, and initial website load times improving by 10%.

Cutover
30 minutes
Migration downtime
Zero
Initial load times
10% faster

The migration plan included a rollback path. Product updates were temporarily paused, a snapshot was transferred through Amazon S3, and legacy and AWS replicas were retained while the team monitored the new environment for two weeks.

Source: AWS Partner Success, HomeTiger and CloudHero. These are historical results for that migration, not a service-level guarantee or measurements of the current platform.

For your own platform

What should an ecommerce modernization assessment resolve?

Use the assessment to connect engineering work to a specific customer journey. For a marketplace, that may be product discovery; for another retailer, it may be checkout, availability or catalogue ingestion. Agree how that journey will be tested before deciding the migration is complete.

  1. Map the dependencies. Identify which applications, data feeds and external services rely on the component you want to change.
  2. Define the change window. Distinguish pausing background updates from interrupting the service customers use.
  3. Make rollback testable. Agree the decision point, data-reconciliation approach and responsibility for returning to the previous environment.
  4. Plan the operating period. Set acceptance criteria and a monitoring window before retiring the old infrastructure.

These are planning questions for your project. The target architecture should follow the workload’s requirements and the team’s ability to operate it.

Your next step

Plan modernization around the service your customers depend on.

Bring us your architecture, operational constraints and business priorities. We will help define the assessment and an achievable migration plan.

Contact CloudHero

Tell us what you're working on

Our sales team replies by email within 1 business day.

Topic

We use your details only to reply. Privacy policy