ReFS kernel dump truncation (12%) on Hyper-V CSV failure – How to configure DedicatedDumpFile to a raw/local disk tier

Zephyr Vance 20 Reputation points
2026-08-04T15:07:39.54+00:00

1. Environment:

  • OS: Windows Server 2022 Datacenter (Hyper-V Cluster)
  • Storage: ReFS on Cluster Shared Volumes (CSV)

2. Issue Summary:

During an unhandled ReFS metadata exception on a Hyper-V host node, the system fails to write a complete kernel dump (MEMORY.DMP). The crash dump generation consistently truncates at ~12%.

3. Root Cause Analysis:

Because the crash is triggered by an unhandled exception inside the ReFS storage stack itself, access to the ReFS-formatted target partition is lost midway through the dump writing process (dumpfve.sys / disk.sys).

4. Question:

What is the recommended workflow to redirect the early crash dump driver to write directly to a raw dedicated local disk tier (or non-ReFS partition) to bypass the failing file system stack during a BSOD?

Windows for business | Windows Server | Storage high availability | Virtualization and Hyper-V
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Daphne Huynh (WICLOUD CORPORATION) 985 Reputation points Microsoft External Staff Moderator
    2026-08-05T02:58:49.05+00:00

    Welcome to Microsoft Q&A!

    Thank you for providing the detailed information.

    Based on the scenario described, the recommended approach is not to store crash dumps on a ReFS Cluster Shared Volume (CSV) when the suspected failure occurs within the ReFS or CSV storage path itself. Instead, configure a Dedicated Dump File on a local, non-clustered volume that remains accessible to the crash dump stack during bugcheck processing.

    A dump that consistently stops at approximately 12% is consistent with the dump writer losing access to the target storage while the dump is being written. If the bugcheck originates within the ReFS metadata path, CSV infrastructure, or related storage stack components, the system may no longer be able to reliably access the same ReFS-backed volume that is being used as the dump target.

    The DedicatedDumpFile feature specifically to decouple crash dump generation from pagefile and system volume requirements. A dedicated dump file is reserved exclusively for dump capture and can be located on a separate local volume.

    For a Hyper-V cluster, the recommended workflow is:

    1. Provision a local NTFS volume (or another supported local fixed-disk volume) that is not hosted on CSV and does not depend on ReFS.
    2. Configure the following registry values:

    HKLM\SYSTEM\CurrentControlSet\Control\CrashControl

    DedicatedDumpFile = D:\DedicatedDumpFile.sys

    DumpFile = D:\MEMORY.DMP

    DumpFileSize = <appropriate size in MB>

    1. Configure the system for a Kernel Memory Dump or Complete Memory Dump, depending on the data required for analysis.
    2. Reboot the node so the crash dump configuration is fully initialized.
    3. Validate dump generation using a controlled test method, such as an NMI-triggered crash or NotMyFault, during a maintenance window.

    Important considerations:

    • A DedicatedDumpFile must reside on a local fixed disk and should not be placed on clustered storage. Please notes that clustered storage is not an appropriate location for a dedicated dump file.
    • There is no supported method to write a Windows kernel dump directly to a raw, unformatted disk device. The crash dump mechanism still requires a supported local volume and a preallocated DedicatedDumpFile.
    • When the suspected fault domain includes ReFS, CSV, storage filter drivers, or SAN connectivity, use a dump target that is completely independent of that storage path, such as a local SSD or a local NTFS partition.

    Therefore, the recommended mitigation is to place both DedicatedDumpFile.sys and the resulting MEMORY.DMP on a non-clustered local NTFS volume, rather than on the affected ReFS CSV. This provides the highest likelihood of obtaining a complete crash dump even when the ReFS or CSV stack is the component that triggered the bugcheck.
    For additional information, please visit:

    How to use the DedicatedDumpFile registry value to overcome space limitations on the system drive when capturing a system memory dump | Microsoft Learn
    Configure memory dump files for Server Core installation | Microsoft Learn
    Memory dump file options - Windows Server | Microsoft Learn

    If you find this answer helpful, please click Accept Answer.

    Thank you for using Microsoft Q&A!

    Was this answer helpful?

    0 comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.