ESXi - High latency on disk and abort commands in logs

Products

• VMware vSAN
• VMware vSphere ESXi

Issue/Introduction

Datastore performance graph in vCenter Web Client UI indicates higher than expected latency for VMs and backend.

Vmkernel.log

YYYY-MM-DDTHH:MM:SS  cpu11:2098200)ScsiDeviceIO: 3414: Cmd(0x459b7c56c880) 0x2a, CmdSN 0x800e0046 from world 2102393 to dev "naa.################################" failed H:0x0 D:0x2 P:0x0 Valid sense data: 0xb 0x47 0x3.
YYYY-MM-DDTHH:MM:SS  cpu11:2098200)ScsiDeviceIO: 3414: Cmd(0x459b7c5025c0) 0x2a, CmdSN 0x800e0005 from world 2102393 to dev "naa.################################" failed H:0x0 D:0x2 P:0x0 Valid sense data: 0xb 0x47 0x3.
YYYY-MM-DDTHH:MM:SS  cpu12:2098200)ScsiDeviceIO: 3414: Cmd(0x45a2cc785300) 0x2a, CmdSN 0x800e0039 from world 2102393 to dev "naa.################################" failed H:0x0 D:0x2 P:0x0 Valid sense data: 0xb 0x47 0x3.


• SCSI Sense Key (0xb): Aborted Command
• Additional Sense Code (0x47): SCSI Bus Reset
• Additional Sense Code Qualifier (0x3): Communication failure due to corrupted data on the bus

Environment

VMware vSphere Esxi

Cause

This error suggests a parity error condition, where the disk has not received the FC packet correctly.
The potential causes of this issue could be any component on the FC loop, such as cables, interfaces, or any other related hardware.

Resolution

Contact storage vendor for assistance.

Additional Information

Interpreting SCSI Sense Data