Quiz: Server Hardware¶
5 questions
L0 (1 questions)¶
1. What do amber and blue LED indicators mean on most server hardware?
Show answer
Amber/orange LED = fault or warning (failed disk, degraded PSU, over-temp). Blue LED = identification (lit by admin via iDRAC/iLO/ipmitool to locate a server in a rack). Blinking amber often indicates a predictive failure. Solid amber = component has failed. Green = healthy. No LED on a PSU = no power input. Check vendor docs for exact meanings — they vary by manufacturer.L1 (2 questions)¶
1. How do you identify a failing disk on a running server?
Show answer
smartctl -a /dev/sdX (check Reallocated_Sector_Ct, Current_Pending_Sector). dmesg | grep -i error for I/O errors. For RAID: megacli/storcli -PDList -aAll.2. How do you diagnose ECC memory errors and map them to a physical DIMM slot for replacement?
Show answer
1. Check system event log: ipmitool sel list | grep -i mem.2. Check dmesg: dmesg | grep -iE 'edac|ecc|memory|mce'.
3. Check EDAC counters: find /sys/devices/system/edac/ -name ce_count -exec cat {} + (correctable errors).
4. Use dmidecode --type 17 to list all DIMMs with locator strings (e.g., DIMM_A1).
5. Cross-reference the EDAC channel/slot with dmidecode output to identify the physical slot. Correctable ECC errors accumulating on one DIMM = replace it proactively before it becomes uncorrectable.
L2 (2 questions)¶
1. How do you troubleshoot ECC memory errors on a running server?
Show answer
1. edac-util -s or mcelog --client for correctable error counts.2. ipmitool sel list for hardware events.
3. dmidecode -t memory to map DIMM slot. Correctable errors are OK short-term; rising count = replace the DIMM before it goes uncorrectable.
2. What is the difference between UEFI and legacy BIOS boot and why does it matter for automation?