Server Won't Power On: Troubleshooting Checklist

No power is a physical-layer problem until you prove otherwise. Do not start by reseating random parts — work this order, document each result, and you will either fix it or hand the engineer a case that needs no follow-up questions.

The fixed order

1. Safety first. Burning smell, scorch marks, or a tripped smell near the rack: stop, keep hands off, escalate to your lead/facilities.
2. Read the PSU LEDs. Dark PSU = no feed reaching it. Amber/blinking = PSU fault. Note which PSU (1/2) and the exact state — this single observation splits the problem in half.
3. Verify the PDU feed. Is the outlet live? Check the PDU display, a neighbor on the same feed, and whether feed A or B is down.
4. Reseat power cables at both ends. PSU end and PDU end. If nothing changes, swap in a known-good cable.
5. Check front panel and management. Any standby power signs? Does the management controller (iDRAC/iLO/IPMI) respond? Standby present but no power-on points inside the chassis; nothing at all points at the feed.
6. Ask what changed. PDU maintenance, a rack move, a neighbor's RMA. Sudden no-power after nearby work is rarely a coincidence.
7. Swap the PSU if authorized. Known-good, same rating. Fault follows the PSU → PSU. Fault stays → feed or chassis.
8. Document and escalate. PSU LED states, feed verified how, cable/PSU swap results, and your specific ask.

Two traps

• Dual-PSU masking: one dead feed can hide behind the other PSU for months. Check both feeds even when the server is running.
• Swapping before observing: if you reseat everything first, you destroy the evidence. Read the LEDs first — they take five seconds.

See RackReady   Free sample

More guides: Fiber no-link checklist · Escalation ticket guide · Interview questions

Support: support@hasherventures.com