Grading Checklist: Power Supply Redundancy Lost

A good response must include:

• Verified PSU status via BMC (ipmitool sensor list, ipmitool sdr list)
• Confirmed the remaining PSU can handle the full load (520W < 800W capacity)
• Checked the power cable and PDU outlet for PSU 2 (rule out cable/PDU issues)
• Assessed the risk: server running without redundancy means any PSU 1 issue = downtime
• Considered draining workloads from the Kubernetes node as a precaution
• Checked warranty status and initiated replacement process
• Verified PSUs are hot-swappable (they are on DL380 Gen10)
• Planned the PSU replacement -- can be done live without downtime
• Attempted reseating PSU 2 before declaring it failed
• Documented the failure for asset management and trending
• Assessed whether other servers with same-age PSUs should be proactively checked
• Verified that after replacement, redundancy is restored and alerts clear