System overview: single points of failure
The following section identifies some of the potential SPOFs in a high availability configuration. The occurrence of any one of these failures results in that node’s failure or isolation.
Potential SPOFs include:
- Operating system crash or failure
- Network failures
- Failure of the physical servers
- Storage subsystem failures
- Power supply failures
The objective is to keep the applications running and highly available for the clients. The strategy is to keep the host operational with as much internal redundancy and operating system safeguards as possible or practical.