Reliability

Hardware faults

• Hard disk crash

• RAM becomes faulty

• Network device is not accessible

• ETC.

Potential solutions

• Add redundant hardware components. e.g. RAID for disk

• Add standby machine, fail over to standby if primary fails (HA)

Software errors

• Software bug

• Run out of shared resource

• ETC.

Potential solutions

• Comprehensive testing

• Process isolation

• Process life cycle management (auto restart)

• Monitoring

Human errors

• Miss configured the system

• Operation failure

Potential solutions

• Provide the sandbox environment

• comprehensive tests which cover the negative cases

• Provide rollback mechanism

• Provide telemetry metrics

• Training