Handling I/O failures in virtual machines due to degraded SW RAID-1
search cancel

Handling I/O failures in virtual machines due to degraded SW RAID-1

book

Article ID: 301186

calendar_today

Updated On:

Products

VMware vSphere ESXi

Issue/Introduction

This article provides steps to handle the IO failures seen in guest VM in case of degraded SW RAID-1.

Symptoms:
As part of setting up ESX for supporting disk mirroring at VM, it is recommended to enable the vmx options using the below parameters for releases below 6.7:

scsi0.returnBusyOnNoConnectStatus=FALSE
scsi0.returnNoConnectDuringAPD=TRUE


For more information, see Set Up Dynamic Disk Mirroring.

In 6.7 release, when the SW RAID-1 in VM is put into degraded mode by failing a disk, the windows GOS may still see IO failures based on the HBA and/or the driver for the HBA.

Environment

VMware vSphere ESXi 6.7

Resolution

To prevent the IO failures seen at GOS, an additional vmx option needs to be enabled in case of 6.7 version for each of the vmdks participating in SW RAID-1

scsi0:1.passthruTransientErrors=TRUE
scsi0:2.passthruTransientErrors=TRUE
.
.
scsi0:x.passthruTransientErrors=TRUE
scsi0:y.passthruTransientErrors=TRUE