Title

AWS re:Invent 2022 - Camada Zero: A real-world architecture framework (PRT268)

Summary

Camada Zero (C0) is an architecture framework designed for extreme resilience, particularly for mission-critical applications.
Itaú Unibanco, one of the largest financial institutions in Latin America, has adopted C0 to serve its 65 million customers, including 20 million digital-only customers, with over 4,000 business services.
The bank has a 10-year partnership with AWS and is modernizing its applications by moving to the cloud, aiming for 45% of applications on AWS by the end of the year and 60% by the next year.
C0 applications are designed for 100% uptime and are critical for Itaú, which processes 30% of Brazil's daily transactions.
The C0 framework includes principles across three pillars: architecture, engineering, and operations.
The architecture is based on a scale unit-based design, with cells (applications) that are self-contained and decoupled, and shards (system partitions) for fault tolerance and scalability.
Engineering principles focus on small services with well-defined purposes and no shared resources, which simplifies testing and troubleshooting.
Operations principles emphasize observability and alerting based on Service Level Objectives (SLOs), with a focus on minimizing Mean Time to Recovery (MTTR) and customer impact.

The C0 framework is a response to the need for financial institutions to have highly resilient and available systems, especially in the face of increasing digital customer interactions and regulatory implications.
Itaú Unibanco's approach to cloud migration is not just re-hosting but re-architecting and rebuilding applications for the cloud, which is a significant undertaking that requires careful planning and execution.
The architecture principles of C0, such as scale unit-based architecture, demonstrate a sophisticated understanding of cloud-native design patterns that prioritize fault isolation and controlled scalability.
The engineering principles of C0 advocate for microservices architecture, where services are small, with single responsibilities, and do not share resources, aligning with modern software development best practices.
The operations principles of C0, particularly the use of SLOs and error budgets, show an advanced approach to operational excellence, where the focus is on proactive management of system health and rapid response to issues.
The C0 framework's emphasis on control and minimizing the blast radius of failures reflects a shift in mindset from reactive to proactive system management, which is crucial for maintaining customer trust and meeting regulatory requirements.