Two power supplies do not automatically create end-to-end power redundancy. The full path from utility supply to server PSU must be considered.
Practical example: small business virtualisation server
Two PSUs are strongest when the upstream paths do not share the same single point of failure.
Scenario: A 35-user business wants one physical server for directory services, an accounting application, file services and three small virtual machines.
The design starts with the workload rather than a model number. Memory is budgeted for the hypervisor, each virtual machine and growth. Storage is separated into performance and capacity requirements. RAID is selected for availability, while a separate backup target is retained for recovery. Dual power supplies and multiple network interfaces are considered if the required server supports them.
Decision: The final server should be chosen only after checking processor capacity, DIMM population rules, drive bays and backplane, RAID controller, network interfaces, PSU configuration and the supported expansion path.
Common mistakes and selection checklist
Common mistakes
- Connecting both redundant PSUs to the same PDU or upstream power source and assuming the path is fully redundant.
- Using two PSUs that cannot individually support the required server load.
- Ignoring PSU matching and platform-specific redundancy rules.
What happens if you get it wrong?
A single upstream PDU, UPS or circuit failure can still shut down a dual-PSU server. An incorrectly configured PSU pair may also run without the redundancy expected by the administrator.
Selection checklist
- Confirm the server's supported PSU configuration and redundancy mode.
- Ensure each required power path can support the intended load.
- Separate upstream feeds, PDUs or UPS paths where the availability requirement justifies it.
- Check PSU type and rating compatibility.
- Test failure behaviour during planned maintenance where operational procedures allow it.
1+1 redundancy
In a typical 1+1 configuration, either power supply can support the server load. Dell documents systems where identical PSUs can operate in redundant 1+1 mode, while HPE and Lenovo also specify matching power supply requirements for supported platforms.
A/B power paths
For stronger resilience, PSU A and PSU B should not both depend on the same failed component. They may connect to separate PDUs, UPS outputs or upstream circuits where the electrical design supports that arrangement.
Capacity during failure
One remaining PSU and one remaining power path must be able to carry the full server load. A configuration that requires both PSUs for capacity is not redundant against one PSU failure.
Test and monitor
Power supply alerts, UPS health, PDU loading and battery condition should be monitored. Redundancy that has silently degraded provides false confidence.