Founded as a blog for high school students in 2008, ScuolaZoo has evolved into a tour operator for young people aged 16 to 25 and a merchandise provider for schools. Today, ScuolaZoo offers three digital products for clients and three management tools, all developed internally.
ScuolaZoo has accumulated numerous digital web projects, experiencing significant traffic spikes that vary with the seasons.
In 2019 ScuolaZoo began exploring the cloud to better understand how to leverage its full potential, initially opting for a local cloud solution. However, the company soon realized that this option wasn’t scalable, and required significant effort to identify infrastructure issues and connection-related bugs. As a result, ScuolaZoo decided to implement Google Cloud technology and partnered with an innovation expert to support their migration and provide deep knowledge of Google services—leading to the choice of Huware. Since ScuolaZoo didn’t have in-house DevOps, their collaboration with Huware allowed them to evolve in the Cloud through the dockerization of their projects.
The main goals were to make the business scalable in terms of operations and volume, reduce in-house DevOps costs and complexity by leveraging Google-managed solutions and Kubernetes infrastructure, provide developers with manageable tools for monitoring and debugging ScuolaZoo’s applications, and ensure business continuity throughout the release lifecycle.
All this to aim at becoming faster and more efficient in problem-solving.
ScuolaZoo and Huware began a service transformation process, migrating ScuolaZoo’s technology to what is called a “Container Native” environment, a technology that enables scalability. Initially, this technology was deployed on Kubernetes clusters and then evolved into a serverless architecture using Cloud Run and a CDN to enhance performance.
"The move to Kubernetes was a natural evolution for us to manage the project’s scalability. However, we encountered a common issue faced by many development teams: the lack of solid internal DevOps expertise, which led to significant traffic and connection problems between services and databases. Together with Huware, we were able to identify what was missing in our configuration. "
Andrea Orselli
Head of Digital
The identification of the issues with the old configuration was made possible by isolating them. ScuolaZoo then migrated to Cloud Run technology, delegating to Google the management of traffic, connections and all activities that were previously slowed down. By moving all projects on Cloud Run applications, ScuolaZoo and Huware were able to accurately identify the source of the problems.
To achieve the final result, the project followed several steps.
First, Huware and ScuolaZoo defined and mapped all high-impact projects on K8s clusters. Then, the CI/CD pipeline was redefined, unified and redesigned for the Cloud Run project, planning weekly migrations pipelines from K8s to Cloud Run. Finally the migration was divided into four steps: updating the deployment code, moving SSL certificate management to Google-managed certificates and assigning Cloud Run domains, migrating the project to a test environment, and migration and going live with the project in the production environment.
Thanks to Huware and Google Cloud, ScuolaZoo has significantly reduced the effort required from DevOps engineers. Previously, the development team was heavily dependent on in-house DevOps, and every structural change or feature required effort from a single DevOps engineer, creating bottlenecks during peak seasons. Additionally, there has been an increase in expertise and responsibility among developers working on individual projects.
Thanks to Google Cloud technology, ScuolaZoo successfully offloaded the workload of connectivity and networking from internal applications to Google, using load balancers. Now ScuolaZoo developers have greater control over the status of each project.
"“This was our first experience in a completely serverless application context, so it could have posed a risk. In reality, it turned out to be a huge advantage, as we transitioned from service delivery in Kubernetes clusters—which had a certain processing delay—to Cloud Run with an average processing time of 900 milliseconds during peak traffic. This resolved a number of complexities, making our business scalable.” "
Andrea Orselli
Head of Digital
