• Shlomi Perets

Raycatch Achieves 50% Reduction in Cloud Spending by AWS Cloud Migration



About Raycatch

Founded in 2015, Raycatch is an Israeli startup that has developed an innovative, AI-driven digital asset management system that automates and optimizes solar PV assets, making renewable energy more efficient, accessible and profitable.


Raycatch’s system delivers a 360° view of the solar assets and shows exactly where are the

performance issues, how to solve them and whether fixing them is a worthwhile investment. This enables asset owners, managers and operators to effectively manage their portfolio

and optimize their yield.

The Motivation

While having a working production environment running on Kubernetes in GCP, Raycatch wanted to improve the performance, scalability and cost of the solution. Raycatch contacted Automat-IT with the following goals:

  1. Lower cloud spending.

  2. Design and deploy a flexible, scalable and manageable environment to run the current workloads, as well as be able to re-architect some of the platform services and deploy them on the same infrastructure with minimal production impact.

  3. Zero downtime, including “live” databases and monitoring systems.

The Solution

Raycatch and Automat-IT decided to implement migration to AWS cloud with a re-architecture migration strategy.

The chosen solution was to migrate the Kubernetes workloads to AWS EKS managed Kubernetes. In order to provide different workload sizes to run on different machine types, the Kubernetes was built with several affinity node groups, while Raycatch’s internal application deploys specific workloads on specific node groups per demand. This solution provided high flexibility when services and workloads re-architectured and could be deployed on the specific node group as required.


Automat-IT worked with Raycatch to re-architect its infrastructure to adopt the new system requirements, then migrated Raycatch’s workloads to AWS while keeping the production system running, and upon cutover, the databases were migrated.


The migration was done in stages:

  1. Create a staging environment to test the workloads on AWS before migrating the production workloads.

  2. Create a side-by-side production environment in AWS.

  3. Migrate the workloads to AWS while monitoring that the system functions properly.

  4. Software optimizations on Raycatch’s behalf, to adapt more efficiently to the use of relatively short microservices and cloud broadcasting, and to improve performance on spots.

Benefits

  1. Lower cloud dynamic services spend by 70% through the use of Spot Instances.

  2. Lower overall cloud spending by 50% after migration from GCP to AWS.

  3. A scalable, redundant, secured, stable and monitored environment.

  4. Improved performance and scalability.

  5. A seamless, low-effort migration without downtime.