ESXi 8.x Host Unresponsive due to DCBD Memory Failures and Dell BOSS Device Degradation

Products

• VMware vSphere ESXi

Issue/Introduction

An ESXi 8.x host becomes unresponsive in vCenter ("Not Responding") or experiences a kernel hang/panic. Upon reboot, the host may recover, but logs indicate critical memory allocation failures in the networking stack and hardware health warnings for the local boot media.

Environment

• VMware ESXi 8.0 Update 3 (Build 24859861 and similar)
• Dell PowerEdge servers using Dell BOSS2DN1 boot devices
• High-speed NICs (e.g., nmlx5_core or bnxtnet drivers)

Cause

1. Management Deadlock: The Data Center Bridging Daemon (DCBD) fails to allocate VMkernel memory pages (set_hw_pg: Failed status). This is typically triggered by a mismatch between the NIC driver and the physical firmware, causing hostd to hang while waiting for hardware I/O.
2. Storage Degradation: SMART monitoring detects that the Dell BOSS boot device has exceeded its reallocated sector threshold, leading to OSData corruption or I/O hangs that contribute to the host unresponsiveness.

Resolution

1. Hardware Replacement: Replace the failing Dell BOSS2DN1 boot device through the hardware vendor.
2. Driver/Firmware Alignment: Ensure the NIC driver and firmware versions are perfectly aligned according to the VMware Compatibility Guide (VCG).
3. Workaround: To restore management connectivity without an immediate reboot (if the host is still reachable via SSH), stop the DCBD service:
   bash

   /etc/init.d/dcbd stop

Additional Information

• KB 433257: ESXi host shows "Not Responding" due to DCBD errors