| Previous | Next |
| HNS_SWITCH_EXTENSION_NOT_FOUND | ERROR_SVHDX_ERROR_NOT_AVAILABLE |
ERROR_SVHDX_ERROR_STORED
What the stored-error result represents
ERROR_SVHDX_ERROR_STORED is an indirection result from the Shared VHDX path. The server has already captured a more specific failure together with SCSI sense data, so this HRESULT is not the root cause of the failed read, write, reserve, or management request. Treat it as a signal to retrieve and preserve the server-side diagnostic record before retrying an operation that might change the state of shared storage.
Remote Shared Virtual Disk implementations can forward SCSI operations for a virtual disk that is accessed by multiple consumers. In that model, the command result, sense key, additional sense code, and additional sense code qualifier can carry the actual reason, such as a Unit Attention condition, a reservation conflict, a transport failure, or a device-state change. The same top-level HRESULT can therefore accompany very different recovery actions.
What to collect before acting
- Record the original operation, virtual disk identity, initiator identity, request time, and any server correlation or activity identifier.
- Retrieve the extended error and sense data using the mechanism exposed by the caller or protocol. Do not replace the record with a later retry result.
- Classify the sense result before choosing recovery. A Unit Attention commonly requires state revalidation; a reservation conflict requires ownership analysis; a medium or hardware error points to a different layer.
- For a clustered workload, correlate the result with cluster ownership transitions, disk attach or detach events, failover, and registration changes. Avoid direct reservation-changing commands on a production shared disk unless they are part of the cluster recovery procedure.
Why a blind retry can be unsafe
A retry may be harmless for a read that was interrupted by a transient condition, but it can be unsafe for an operation that changes reservation state or writes metadata. The retained error is valuable precisely because it describes the first failure in context. If the server has already released the record, the later request can instead report ERROR_SVHDX_ERROR_NOT_AVAILABLE; that does not mean the original operation succeeded.
References
- Microsoft Open Specifications: Remote Shared Virtual Disk Protocol
- T10: SCSI persistent reservations
Looking for a different code? Search another status or error code.