Customer Reliability Engineering: SRE for Your Customers
CRE extends SRE principles to your customers, making their reliability a shared goal. A provider partners with a key customer to co-engineer resilient services on their platform.
WHY IT EXISTS: Cloud providers and their largest customers have a shared fate. If a major customer's application goes down, it reflects poorly on the underlying platform, even if the platform itself is healthy. CRE was created to bridge this gap, aligning the provider and customer on a joint mission of reliability.
THE MENTAL MODEL: Think of CRE as embedding a Site Reliability Engineer from your cloud provider directly into your team. They don't just fix things when they break; they share the operational burden, consult on architecture, and teach your team how to use the platform's features for maximum resilience. It's a partnership, not a transaction.
HOW IT WORKS: A CRE team engages directly with a customer's engineering team. They might start by establishing shared Service Level Objectives (SLOs), conducting production readiness reviews (PRRs) for new features, and participating in the customer's incident response process. The goal is to transfer SRE knowledge and best practices, making the customer's team self-sufficient in the long run. The engagement is consultative and hands-on.
WHEN TO USE IT: CRE is most effective for critical, large-scale applications running on a third-party platform like a public cloud. It's for situations where the complexity is high and the cost of failure is significant for both the customer and the platform provider. It's a strategic investment for both parties, not a standard offering.
WHEN NOT TO USE IT: CRE is not a substitute for a standard support contract. It's not for small-scale applications or teams that just need break-fix help. Customers must be willing to invest significant engineering time and be open to changing their architecture and operational practices. It's a deep commitment, not a casual help desk.
ONE CANONICAL EXAMPLE: Google Cloud's CRE team partners with major customers like The Home Depot. Instead of just providing infrastructure, the CREs work with Home Depot's engineers to re-architect their e-commerce platform for Black Friday scale, define meaningful SLOs for their services, and co-manage operational tasks, ensuring the site stays up during peak traffic.
Read the original → cloud.google.com
Get five bites like this every day.
Tezvyn delivers a daily feed of 60-second tech bites with quizzes to lock in what you learn.