What is a single point of failure (SPOF)?
Options
- A bug that is found in a single line of code
- A component whose failure takes the system down (Correct option)
- The first server that is set up in a cluster
- A failed test somewhere in the build pipeline
Answer
A component whose failure brings the whole system down because there is no backup
Why
If your app runs on one server with one database, and that database crashes, everything stops: it is a single point of failure.
You remove SPOFs with redundancy: several app servers behind a load balancer, a database with a replica ready to take over, more than one data center.
High availability is designed by removing SPOFs.
Remember
SPOF = one part that takes everything down. Add redundancy.