Single Points of Failure Identification
What You'll Learn
You'll learn how to systematically identify and document single points of failure (SPOFs) in system architecture diagrams to improve availability.
What Is a Single Point of Failure?
A single point of failure is any component in your system whose failure would cause the entire service—or a critical part of it—to stop working. Think of it like a single bridge connecting an island to the mainland: if that bridge fails, no one can get across.
In cloud infrastructure, SPOFs are availability killers. They're the weak links that can take down your carefully designed system, often at the worst possible time.
How to Identify SPOFs
Start with Architecture Diagrams
Walk through your system architecture drawing and ask at each component: "If this fails right now, what breaks?"
Common SPOF Locations
- Single database instance with no replica or failover
- One load balancer without a backup
- Single availability zone deployment
- Single network gateway or NAT instance
- One storage volume without replication
- Single authentication service with no redundancy
Documentation Strategy
For each SPOF you identify:
- Mark it on your diagram (often with a red icon or highlight)
- Document the impact: What fails if this component goes down?
- Estimate the blast radius: How many users or services are affected?
- Note the RTO/RPO implications: How quickly can you recover?
The "What If" Test
Mentally (or actually) remove each component and trace the failure cascade. If removing one piece brings down a critical path, you've found a SPOF.
Key Takeaway: Identifying SPOFs requires systematically examining each component in your architecture and documenting any single element whose failure would cause service degradation or outage—this inventory becomes your roadmap for improving high availability.