Troubleshooting ESXi Host Unresponsiveness Due to OSDATA Exhaustion and APD/PDL Events
search cancel

Troubleshooting ESXi Host Unresponsiveness Due to OSDATA Exhaustion and APD/PDL Events

book

Article ID: 454304

calendar_today

Updated On:

Products

VMware vSphere ESXi

Issue/Introduction

ESXi hosts may enter a "Not Responding" state, rendering hosted VMs inaccessible. Symptoms include logs reporting "No space left on device" for the OSDATA partition and "APD Notify PERM LOSS" (All Paths Down) events in the vmkernel.log.

Environment

VMware vSphere ESXi 8.x

Cause

The OSDATA partition becoming full prevents the ESXi host from performing essential management tasks. Concurrent storage fabric or array instability triggers APD/PDL conditions, causing the host to stop responding as it attempts to retry I/O operations.

Resolution

  1. Check Space: Use vdf -h to check the OSDATA partition.
  2. Clear Space: Remove unnecessary log files or core dumps if the partition is full.
  3. Investigate Storage: If space is sufficient, examine the storage fabric for events leading to APD/PDL, such as switch configuration errors or array-side LUN issues.
  4. Patches: Ensure the ESXi host is running the latest recommended updates to benefit from stability improvements.
Powered by Wolken Software