troubleshooting practice questions

The UPS battery is depleted

Why it fails: The UPS battery status does not affect the server's ability to power on if utility power is present; the server powers on momentarily.

One of the power supplies has failed

Why it fails: A failed power supply would not allow the server to power on even briefly; it would remain off or show error LEDs.

The CMOS battery is dead

Why it fails: A dead CMOS battery would only reset BIOS settings and may cause a date/time error, not a failure to POST.

Shrink the existing partition and extend it onto an additional drive

Why it fails: This is not directly supported on a basic disk and would require conversion to dynamic disk, which often requires a reboot.

Add a new drive, initialize it, and configure a RAID 1 mirror with the existing drive

Why it fails: Mirroring does not increase usable space and may require a reboot to convert to dynamic disk.

Replace the current system drive with a larger one and restore from backup

Why it fails: This causes significant downtime and is not the best solution when a non-disruptive option exists.

The new replacement drive is faulty.

Why it fails: If the new drive were faulty, the rebuild would likely have failed or shown errors.

The motherboard has a faulty PCIe slot for the RAID controller.

Why it fails: Faulty PCIe slot would cause the controller to disappear or cause I/O errors, not parity errors.

The RAID controller cache module has failed.

Why it fails: Cache failure would cause write errors or performance issues, not parity errors on a mirrored pair.

Run a filesystem check utility from a recovery disk.

Why it fails: A filesystem check would only help if the error were due to filesystem corruption, not a missing boot device.

Replace the hard drive with a known good one.

Why it fails: Replacing the drive is drastic and should not be the first step without verifying other possibilities.

Reseat the hard drive data and power cables.

Why it fails: Reseating cables is reasonable but less likely than a boot order change after a power outage.

Disk fragmentation

Why it fails: Incorrect. Fragmentation increases seek times but typically does not cause queue lengths above 10 on modern RAID arrays.

Slow network link

Why it fails: Incorrect. A slow network link would affect network throughput, not disk I/O wait times and queue length.

Overloaded CPU

Why it fails: Incorrect. High CPU utilization does not directly cause high disk I/O wait; it might indicate a bottleneck elsewhere.

Memory leak in an application

Why it fails: Incorrect. While a memory leak could cause high usage, the exhibit does not indicate a specific application; the most direct interpretation is insufficient memory.

Virtual memory is misconfigured

Why it fails: Incorrect. Virtual memory settings appear normal; the issue is lack of physical memory, not a configuration problem.

Page file is too small

Why it fails: Incorrect. The page file size (16,384 MB) is equal to physical memory; that is typically sufficient.

Corrupted file system

Why it fails: Incorrect. File system corruption would show event IDs like 55 or 26, not a controller error.

Loose data cable

Why it fails: Incorrect. While loose cables can cause errors, event ID 11 typically points to a controller-level issue rather than a cable connection.

Failed hard disk drive

Why it fails: Incorrect. A disk failure usually generates different event IDs (e.g., 7, 50), and the error points to the controller.

The memory module is not seated properly

Why it fails: While improper seating can cause errors, the beep code for a memory error often points to incompatibility first; seating issues usually cause no beep or a different code.

The BIOS is not updated to support the new memory

Why it fails: BIOS updates are rarely required for standard memory; this is a less common cause of POST memory errors.

The memory is installed in the wrong slot

Why it fails: Most server motherboards support single-module configurations, so slot order rarely prevents boot. The beep code would be different if it were slot-related.

Replace the CMOS battery

Why it fails: A dead CMOS battery affects BIOS settings and time, but does not prevent the system from POSTing and booting.

Flash the BIOS firmware

Why it fails: BIOS corruption is rare and typically shows symptoms like boot loops or incorrect hardware detection; it is not the primary suspect after reseating components.

Replace the power supply

Why it fails: A failing power supply typically causes random shutdowns or failure to power on, not a complete lack of POST with fans spinning.

Move the server to a cooler location in the data center

Why it fails: The ambient temperature is within acceptable range, so moving the server would not significantly improve cooling.

Reinstall the CPU heat sink with more thermal paste

Why it fails: The thermal paste has already been replaced; adding more or reapplying is unlikely to resolve the issue.

Replace the power supply unit

Why it fails: The PSU fan is integrated, but the logs point to CPU temperature, not power supply failure. Replacing PSU would not address the root cause.

The switch port is configured for auto-negotiation while the NIC is set to 1000/full

Why it fails: A speed/duplex mismatch leads to CRC errors and poor performance but does not cause the physical link to drop repeatedly.

The NIC driver is outdated

Why it fails: An outdated driver may cause sporadic issues but is less likely to produce such a clear pattern of repeated link state changes.

Spanning Tree Protocol is blocking the port

Why it fails: STP blocking prevents data forwarding but the physical link remains up; it would not generate link up/down messages.

Disable SELinux and restart the httpd service

Why it fails: SELinux would typically produce a permission denied error, not a syntax error about an invalid command.

Run `yum update httpd` to update the application

Why it fails: Updating httpd would not install the mod_ssl module, which is required for the SSL directive.

Edit the configuration file to correct the typo on line 15

Why it fails: There is no typo; the directive is valid but the module providing it is not loaded.

Increase the number of physical cores in the host by adding another CPU.

Why it fails: Adding physical cores is a hardware change that may not be immediately feasible and does not address the root cause; the host may have sufficient capacity if the VM's vCPU count is appropriate.

Enable CPU affinity to pin the VM to specific physical cores.

Why it fails: CPU pinning can reduce overhead but may not resolve contention if the host is oversubscribed; it can also lead to other issues.

Reduce the number of vCPUs assigned to the VM to 4.

Why it fails: Reducing vCPUs might help if the VM is not using all vCPUs efficiently, but it could hurt performance if the workload needs 8 vCPUs; migration is a less intrusive first step.

Manually force start the cluster service on Node B with the /forcequorum switch.

Why it fails: Using /forcequorum can cause a split-brain scenario; it is intended for emergency recovery and not a standard fix.

Restore the system state backup on Node B.

Why it fails: System state restore may revert cluster configuration but is more time-consuming and may not be necessary compared to eviction.

Run the cluster validation wizard to check configuration.

Why it fails: The validation wizard only reports configuration issues but does not make changes; it won't fix the quorum problem.

Initiate a full array consistency check to find and repair any parity mismatches before rebuilding.

Why it fails: A consistency check verifies data integrity but does not fix a bad drive. It may stress the array further without solving the rebuild problem.

Back up all data immediately, then break and recreate the RAID 5 array from scratch, restoring the backup.

Why it fails: This extreme measure should be a last resort. It involves full downtime and risk of backup integrity, and does not address why the new drive isn't working.

Replace the RAID controller card and its cache module, then attempt the rebuild again.

Why it fails: RAID controller failure is possible but less likely when other drives function normally and the error is localized to the hot spare. It's an expensive and disruptive step.

The system drive has insufficient free space

Why it fails: Low disk space would cause a specific error related to disk, not a service timeout.

A recent printer driver update is corrupted

Why it fails: A corrupt driver typically causes the spooler to crash after starting, not to time out during startup.

The password for the service account has expired

Why it fails: The Print Spooler uses the local system account, which does not have an expiring password.

Reset the BIOS to factory settings.

Why it fails: UEFI firmware reset does not restore a missing boot manager file on the EFI system partition.

Replace the hard drive and restore from backup.

Why it fails: This is excessive; the drive is not necessarily failed, only the boot loader needs repair.

Boot from the installation media and run chkdsk.

Why it fails: chkdsk repairs file system errors, but 'Bootmgr is missing' points to boot loader issues, not disk corruption.

Run the cluster validation wizard

Why it fails: Validation wizard identifies configuration issues but is not a first step for emergency service restoration.

Force a quorum on Node B using the /forcequorum switch

Why it fails: Forcing quorum may allow the cluster to start but does not resolve the underlying storage connectivity problem.

Reboot Node A

Why it fails: Rebooting the failed node is time-consuming and does not guarantee storage connectivity from Node B.

The new drive has a higher rotational speed.

Why it fails: Differences in rotational speed do not typically prevent a RAID rebuild; the array can function with mixed speeds.

The new drive has a different firmware version than the other drives.

Why it fails: Firmware mismatch can cause operational issues but seldom leads to a rebuild that begins successfully and then aborts.

The new drive is not hot-swappable and was inserted while the server was on.

Why it fails: Most server-class drives are hot-swappable; inserting a non-hot-swappable drive usually causes immediate damage or recognition failure, not a rebuild that starts and then fails.

Replace the Broadcom NIC with a different brand network adapter and disable the onboard NIC in the BIOS.

Why it fails: Replacing hardware is unnecessary if rolling back the driver resolves the issue, and would incur additional cost and downtime.

Reinstall Windows Server 2016 and restore data from backups to eliminate any corrupted system files.

Why it fails: Reinstalling the OS is excessive, causes significant downtime, and does not guarantee that the same driver issue would not recur if the same update is applied again.

Run Windows Memory Diagnostic overnight to check for hardware memory errors that might corrupt driver loading.

Why it fails: While memory errors can cause random crashes, the consistent pattern tied to a specific driver update and the crash dump identifying the network driver make RAM unlikely. Moreover, the server uses ECC memory which would log correctable errors before these crashes.