ESXi Host Unresponsive with PDL Errors and Cessation of Logging

Products

• VMware vSphere ESXi

Issue/Introduction

ESXi hosts become unresponsive and fail to respond to network pings, leading to VMware High Availability (HA) triggering and restarting production virtual machines.

Review of the logs reveals the following signatures:

• Cessation of Logging: A complete gap in vmkernel.log and vobd.log entries exists for several minutes or hours prior to the host reboot.
• PDL Reported: Permanent Device Loss (PDL) events are logged during or immediately following the host boot sequence.
• SCSI Sense Code: Sense Code 0x5 0x25 0x0 (Illegal Request / Logical Unit Not Supported).
• Affected Devices: The PDL typically impacts discovery LUNs or RAID controllers (LUN 0) rather than active VMFS datastores.

Environment

• VMware ESXI 8.0.3
• Dell PowerStore
• Dell Unity
• Boot from SAN

Cause

Simultaneous loss of connectivity to the management network and the storage fabric hosting the Boot-from-SAN LUN (OSDATA partition). The reported PDL errors on LUN 0 are a symptom of the connectivity restoration process rather than the cause of the initial unresponsiveness.

Resolution

Since the issue is rooted in the physical infrastructure or storage fabric rather than the ESXi software layer, follow these steps:

1. Contact the server or blade chassis vendor (e.g., Cisco, Dell, HPE) to review chassis backplane logs and interconnect health.
2. Inspect physical network switches and Fabric Interconnects for port flaps or hardware failures corresponding to the log gap timestamps.
3. Verify path stability to the Boot-from-SAN LUNs on the SAN array.
4. Ensure network/storage adapter (HBA/VIC) drivers and firmware align with the hardware vendor's compatibility matrix.

If further assistance is required, see Contact Broadcom Support.

Additional Information

Related Articles

• Permanent Device Loss (PDL) and All-Paths-Down (APD) on host.
• ESXi host becomes unresponsive following repeated PDL errors on paths.