DDPDoIO; DDP_WriteBlkSync and DDPValidBlkReformatHandler PSOD on vSAN

Products

• VMware vSAN

Issue/Introduction

Symptoms:

• vSAN Host enters a PSOD loop with a below stack:

Panic_vPanic@vmkernel
vmk_PanicWithModuleID@vmkernel
[email protected]
[email protected]
[email protected]
[email protected]
HelperQueueFunc@vmkernel
CpuSched_StartWorld@vmkernel

Environment

VMware vSAN 6.7

Cause

PSOD is due to a bitmap block on the vSAN disk(s) which has gone bad or corrupted.

Resolution

Notice: Please verify if all the objects in the vSAN Datastore are healthy prior proceeding with the below steps.

• Engage Hardware vendor to perform a sanity check and/or if any drives requires a pro-active replacement before recreating the disk-groups

• Identify the disks with errors by reviewing the logs.

• If the disk is identified, please destroy the affected disk-group(s) on the crashed host and recreate Ref: How to manually remove and recreate a vSAN disk group using esxcli.

• If disk-group deletion fails, you may go ahead and destroy the partition tables on all the vSAN Drives from the affected Host.

If the above symptoms and issue matches and the hardware is healthy, please contact Broadcom Support to investigate the issue.

Additional Information

Impact/Risks:

• Affected vSAN node will be in a PSOD Loop