VMs freeze and datastore becomes inaccessible due to Cisco storage controller firmware bug
Article ID: 409401
Updated On:
Products
Issue/Introduction
Virtual machines freeze, halt, or become unresponsive due to read/write timeouts prior to any hard disk drive (HDD) replacement activity. Datastore inaccessibility occurs when the primary virtual device (VD-0, hosting the ESXi root filesystem) is affected by a failing disk. Furthermore, failed read/write operations may persist even after the failed disk is replaced, preventing normal operations until the ESXi host is rebooted. These symptoms are often observed after upgrading from ESXi 6.x to ESXi 7.x.
Environment
VMware vSphere ESXi 7.x
Cause
The issue is caused by a hardware firmware-level defect in the Cisco storage controller (Cisco Bug ID: CSCwm47183). When an HDD begins to degrade, the controller firmware fails to correctly isolate the disk, causing the Virtual Drive (VD) to enter an inconsistent or failed state. The upgrade to ESXi 7.x alters how the operating system handles I/O latency and failure paths, which exposes this pre-existing latent hardware bug more frequently.
Resolution
Reach out to Cisco to review and upgrade the Cisco storage controller firmware to version 4.3(2.250016) or a later supported release to resolve Cisco defect CSCwm47183.
If a degraded disk has already been replaced but the ESXi storage stack has not properly detected or refreshed the virtual device state, reboot the affected ESXi host. This will flush the device state, reload the storage stack, and allow normal operations to resume.
Additional Information
WARNING:
This procedure requires a host reboot and firmware maintenance. Ensure all virtual machines are migrated or powered off and follow your organization's maintenance window protocols.
Feedback
