How do I troubleshoot a server that won’t boot?

Troubleshooting a server that won’t boot is a systematic process. Start at the hardware layer and work up.

---

Step 1: Initial assessment

1. Observe and document symptoms
   1. Is there power to the server (fans spinning, LEDs lit)?
   2. Are there beep codes or error messages on screen?
   3. Is the server stuck in a boot loop or frozen at a specific point?
2. Determine the last known good state
   1. Was there a recent hardware or software change (firmware updates, new hardware, OS updates)?
   2. Did the server shut down unexpectedly or suffer power loss?

---

Step 2: Hardware troubleshooting

1. Power supply
   1. Verify the power cable is securely connected and the server is receiving power.
   2. If available, test with a known working power supply unit (PSU).
2. Connections
   1. Check all internal and external cables (SATA, power, network) for loose connections or damage.
3. Memory (RAM)
   1. Remove and reseat the RAM modules.
   2. If the server has diagnostic LEDs or error codes for memory, use them to identify faulty modules.
   3. Test with one RAM module at a time to rule out bad sticks.
4. Storage devices
   1. Disconnect all storage drives (HDDs, SSDs) and boot the server to see if it reaches the BIOS/UEFI.
   2. If the server boots without drives, a failed drive may be causing the issue.
5. Peripheral devices
   1. Disconnect all unnecessary peripherals (USB devices, additional graphics cards, RAID controllers) to isolate the problem.
6. CMOS/BIOS
   1. Reset the BIOS by clearing the CMOS. This often resolves boot issues caused by misconfigured settings.
7. Motherboard and components
   1. Inspect for physical damage (bulging capacitors, burns) or loose components.
   2. If you suspect motherboard failure, test with backup hardware if available.

---

Step 3: BIOS/UEFI diagnostics

1. Access BIOS/UEFI
   1. If the server boots to the BIOS/UEFI, check for error logs or boot settings.
   2. Ensure boot order is correctly configured (e.g., boot from the main OS drive).
2. Run built-in diagnostics
   1. Many servers (e.g., Dell PowerEdge, HP ProLiant) have built-in diagnostics tools. Run these tests to identify hardware faults.
3. Firmware updates
   1. Check if the BIOS/UEFI firmware is up to date. Corrupt or outdated firmware can cause boot issues.

---

Step 4: Software troubleshooting

1. Boot media
   1. Ensure the server can boot from a bootable USB/CD/DVD. Use a live OS (e.g., Linux or Windows recovery media) to test.
2. Operating system issues
   1. If the server reaches the OS boot screen but fails to load, there may be file system corruption or missing boot files. Use recovery tools to repair the OS.
3. Logs
   1. Check system logs for clues (e.g., /var/log in Linux or Event Viewer in Windows).

---

Step 5: RAID or storage configuration

1. RAID controller
   1. Verify the RAID controller status. A degraded or failed RAID array can prevent boot.
   2. Rebuild the array if necessary, but ensure you have backups before performing any RAID changes.
2. Disk health
   1. Use manufacturer tools (e.g., SMART diagnostics or vendor utilities) to verify the health of the drives.

---

Step 6: Advanced troubleshooting

1. Virtualization issues
   1. If the server hosts virtual machines, ensure the hypervisor (e.g., VMware ESXi, Hyper-V) is intact and the boot partition isn’t corrupted.
2. GPU servers
   1. If the server uses GPU cards for compute workloads, ensure the drivers and BIOS settings for GPUs are properly configured.
   2. Remove GPUs to test if they’re causing boot interference.
3. Kubernetes nodes
   1. For Kubernetes servers, verify that the system can at least boot into the OS.

---

Step 7: Backup and restore

1. Restore from backup
   1. If the server’s boot partition or OS is irreparable, restore the system from a recent backup.
2. Disaster recovery plan
   1. If the server is mission-critical, activate your disaster recovery plan and failover to a secondary server.

---

Step 8: Engage vendor support

1. Hardware warranty
   1. If the issue persists and hardware failure is suspected, contact the server vendor (e.g., Dell, HP, Lenovo) for support or replacement.
2. Software vendor
   1. If the issue is related to virtualization, Kubernetes, or other specialized software, reach out to the appropriate vendor for assistance.

---

Preventative actions

• Regularly update firmware, BIOS, and drivers.
• Conduct hardware health checks.
• Maintain a backup strategy for quick recovery.
• Use redundant PSUs to avoid power-related issues.