Definition

High availability means a system is continuously operational and accessible with minimal downtime (often 99.99% uptime or better).

Core concept

HA designs avoid single points of failure by using redundancy (multiple nodes, replicas), automatic failover, and distributed architectures. The goal is close to “always on” operation.

Use cases

Critical services like banking, healthcare, cloud services (AWS, Azure) use HA to meet SLAs. Databases can use clustering, replication, and load balancing for HA.

Trade-offs

HA adds cost and complexity: extra hardware, more complex architecture, and sophisticated monitoring. But it prevents service outages.

Example

A highly available web app might run on multiple servers behind a load balancer; if one fails, traffic shifts to others seamlessly.

References

HA defined as being “accessible and reliable close to 100% of the time”[[37]](https://www.ibm.com/think/topics/high-availability "What is High Availability? | IBM").