| Previous | Next |
| STATUS_SVHDX_UNIT_ATTENTION_CAPACITY_DATA_CHANGED | STATUS_SVHDX_UNIT_ATTENTION_RESERVATIONS_RELEASED |
STATUS_SVHDX_UNIT_ATTENTION_RESERVATIONS_PREEMPTED
Persistent-reservation ownership was taken away
STATUS_SVHDX_UNIT_ATTENTION_RESERVATIONS_PREEMPTED means a Shared VHDX initiator lost a reservation through a preempt operation. In a guest cluster, that can be an expected part of failover or fencing; it is not a signal for the displaced node to keep issuing writes until its next retry succeeds.
The critical question is whether the preemption matches the cluster manager’s recorded ownership change. A mismatch may indicate a split-brain risk, incorrect reservation keys, or an out-of-band storage operation. The reservation state must be handled by the workload or clustering control plane that owns it.
What to verify
- Stop write paths that depend on the lost reservation and identify the current reservation holder.
- Compare reservation keys, initiator identities, and cluster failover events around the time of preemption.
- Do not use manual raw-SCSI commands to take ownership unless that is an approved recovery procedure for the clustered workload.
References
- Microsoft: VHD Set files for guest clusters
- Microsoft: RawSCSIVirtualDisk
- T10: SCSI persistent reservations
Looking for a different code? Search another status or error code.