Fix INACCESSIBLE_BOOT_DEVICE in Windows and Windows Server

In this blogpost, we will discuss “Fix INACCESSIBLE_BOOT_DEVICE in Windows and Windows Server”. Please see Active Directory Vulnerability Assessment with Purple Knight: Domain Controller Owner Is Not an Administrator, AGMP extended support ends April 2026: Find alternative solution, and Steps to customize Windows PE boot images. The INACCESSIBLE_BOOT_DEVICE stop error 0x0000007B occurs when Windows loses access to the system partition during startup.

Note: On a standard workstation, the cause is often a changed storage mode or motherboard issue as it is in my case. On a physical Windows Server, the same error may indicate a RAID controller, backplane, virtual disk, or physical-drive failure and not merely an operating-system problem.

Before proceeding to resolve this issue, changing RAID settings, removing disks, or running repair operations. Please confirm that a current backup exists. Do not initialize, format, delete, or recreate a virtual disk during troubleshooting.

Also, see Tape Backup Troubleshooting in Veeam: Real Cases, how to Fix unable to contact Server: This may be the server does not exist, and [World Backup Day] V13 Upgrade Failure: Veeam Configuration Backup Saved the Day.

Why Is This Error Happening?

At startup, Windows depends on storage drivers to communicate with the boot disk. If the storage controller configuration changes, storage drivers become corrupted, or the underlying storage hardware stops responding, Windows can no longer access the operating system volume and crashes with a blue screen. Common causes include:

• Changing BIOS storage settings from RAID/VMD to AHCI
• Missing or corrupted storage drivers
• Failed RAID controllers
• Degraded RAID arrays
• Disk media errors
• Failing SSDs or HDDs
• Corrupted boot configuration data (BCD)
• Firmware or hardware communication failures

Note: The same stop code therefore does not always have the same cause. A workstation may only need the correct storage mode restored, while a server may require a hardware and RAID investigation

Please see how to fix ADDS already contain a computer account, AZURE Kubernetes Service (AKS) upgrade minor version from 1.32.3 to 1.33.7, and how to reset or reinstall Windows Server without deleting the VM.

Windows Workstations (Storage Controller Mode mismatches)

Being abe to identify a recent change and a prior error “How to fix Windows Boot Error 0xc0000098 (winload.efi)” was very instrcumental for solving this issue.

On client PCs, this error stems from a controller mode mismatch in the BIOS after a BIOS reset I guess as this is the only valid assumption. If these torage protocols “RAID or the AHCI modes” changes, Windows attempts to load using its previous storage driver, panics when it cannot find the boot partition, and crashes.

Restore the original setting

Enter BIOS/UEFI setup using the manufacturer’s key and navigate through the storage, SATA, or advanced configuration section. Restore the previous setting:

• RAID On.
• Disabled
• AHCI/NVMe

Save the changes and restart. In many cases, restoring the original storage mode is sufficient to resolve the blue screen except there is a motherboard issue. No registry modification or Windows reinstallation is required.

Please see Build a New DC vs Swing Migration: Upgrade Server OS Correctly, how to build a New DC vs Swing Migration: Upgrade Server OS Correctly, and Azure File Sync : How to enable Self-Service restore.

Windows Server on Dell PowerEdge (RAID and Hardware faults)

The keyboard-layout screen indicates that the server has entered the recovery environment. It does not by itself identify the cause of the storage failure.

Note: As you can see from the logs, my issue pertains to a storage hardware problem and not preventing access to the Windows installation, is a corrupt or missing filter driver.

In enterprise environments with PowerEdge servers and PERC RAID controllers, this error is rarely a simple driver switch or Storage mode change. Instead, it often points to storage bus stalls, bad disk blocks, or RAID controller resets that temporarily drop access to the boot volume.

Note: When the OS starts loading system files and the PERC controller drops connection to the backplane, the system throws INACCESSIBLE_BOOT_DEVICE and automatically reboots into the Windows Recovery Environment (WinRE) frequently stopping at the Keyboard Layout Selection prompt as shown in the image above.

Step 1: Analyze iDRAC and System Event Logs

Log into your iDRAC web interface to inspect the Lifecycle Log and System Event Log, and look for Media Errors & Predictive Failures: Check if physical disks report media errors or status flags like Failure Predicted (For example, Physical Disk 0:1:5).

• Identify Controller Resets: Check if multiple drives log simultaneous insertion events (e.g., Disk 0 through 7 in Backplane 1 inserted at the exact same timestamp). This indicates the PERC controller suffered a full enclosure/backplane reinitialization.
• Inspect OS Event Logs: If you reach WinRE or command line, check for Event ID 7: The device \Device\HarddiskX\DRX has a bad block, indicating read/write errors on physical sectors.

Note: A Dell status such as Failure Predicted means that the disk or RAID controller has detected error patterns suggesting that the disk may fail. The disk may still remain online, but it should not be treated as reliable. Dell recommends backing up the data, collecting controller diagnostics, checking firmware, and replacing the disk if the warning persists.

Please see Unicode Manipulation: CVE-2026-25177 Privilege Escalation in AD, how to replace Veeam Recovery Orchestrator License, and how to Fix failed to connect to the backup server: Make sure it is online.

Step 2: Perform a Complete Flea Power Drain (Fupp Power Drain)

A stalled RAID controller or hanging backplane can retain corrupted state across soft reboots. Clearing residual power resets the storage bus hardware:

1. Power off the server cleanly via iDRAC or the power button.
2. Disconnect all redundant Power Supply Unit (PSU) cables.
3. Press and hold the physical power button for 15 seconds to drain all residual electricity from the motherboard and PERC controller.
4. Wait 2 minutes, reconnect the power cables, allow the iDRAC/BMC to fully initialize (approx. 2 minutes), and power the server back on. See the video below for more information.

Play

Please see How to fix Domain Join Error during Windows Deployment, SCVMM setup Error 10421: Fix VMM Service Account conflict, and SCVMM setup Fails: Fix Missing Windows ADK Deployment Files.

Step 3: Isolate Faulty Drives and Verify RAID Volumes

Since the drive has a predictive failure causing bus resets during boot:

1. Verify RAID Isolation: Ensure your virtual disks separate OS and Data (e.g., RAID 1 across Disks 0–1 for OS; RAID 5 across Disks 2–5 for Data).
2. Reseat / Unseat Degraded Disks: If bad sectors on a secondary array (such as a RAID 5 data volume) cause the storage bus to reset during boot as it is in my case. Temporarily unseat or replace the failing drive (e.g., Disk 5) to restore controller stability.
3. Run Hardware Diagnostics: Run Dell Embedded System Diagnostics via Lifecycle Controller (F10 during boot) to ensure the backplane and PERC controller maintain reliable communication.

Note: For a RAID 5 volume, one failed disk may be tolerated, but the array is no longer fully redundant. A second disk failure or an unrecoverable read error during rebuild can make the volume unavailable. Replace the affected disk promptly and monitor the rebuild.

Do not remove a disk merely to test whether it is causing the problem unless you have confirmed the disk identity, backup status, RAID state, and vendor procedure.

Please see how to Integrate Hyper-V [SCVMM] with Veeam Recovery Orchestrator, how to Integrate Hyper-V: Install System Center Virtual Machine Manager, and how to reset Microsoft 365 User Password.

Server recovery after storage is stable

Dell’s documentation notes that a single failed physical disk can place a RAID 5 virtual disk into a degraded state, while a second physical-disk failure can cause the virtual disk to become failed as we saw from the log, disk 4 also failed as shown below. A failed boot virtual disk prevents Windows from booting. This is because, a controller, backplane, cable, firmware issue, or multiple disk failures can interrupt access to the virtual disk.

A disk media error on Disk 4 in Backplane 1 of Integrated RAID Controller 1 was corrected during recovery.
A disk media error on Disk 5 in Backplane 1 of Integrated RAID Controller 1 was corrected during recovery.

Since we are now sure that the OS volume has no issues which is RAID 1, and Windows still cannot boot normally. We will have to use Windows Recovery Environment (WinRE) for recovery. There are various apporaches you can employ to resolve this issue. Once the RAID controller consistently exposes the system virtual disk:

1. Boot into WinRE.
2. Select the correct keyboard layout.
3. Run Startup Repair.
4. Try Safe Mode.
5. Check the file system.
6. Run offline System File Checker.
7. Review the BCD only after confirming the correct system and Windows partitions.
8. Restart and monitor the server event logs.

After the server returns to operation:

• Confirm that every virtual disk is optimal.
• Confirm that any rebuild has completed.
• Review controller and Windows logs.
• Run a new backup, and if possible, test the backup or recovery procedure.
• Lastly, monitor the affected disks for additional errors and do not forget to replace the drive in RAID 5 as soon as possible.

The most important takeaway is that INACCESSIBLE_BOOT_DEVICE does not automatically indicate operating system corruption. On workstations, a storage-mode mismatch can prevent Windows from locating the boot partition. On Dell PowerEdge servers, the same stop code may point to RAID controller resets, backplane communication issues, degraded virtual disks, or failing physical drives. Always investigate storage accessibility first before considering operating system repair.

I hope you found this blogpost on how to fix INACCESSIBLE_BOOT_DEVICE in Windows and Windows Server very useful. Please feel free to leave a comment below.