ESXi hosts booting from USB or SD card media may enter an unresponsive state in vCenter Server. This occurs when the storage stack cannot communicate with the boot device, leading to a missing bootbank path.
Symptoms:
Not Responding' in vCenter.Operation not allowed" Alarm 'Host error' on ESXi_FQDN triggered by event. Issue detected on ESXi_FQDN in datacenter: Bootbank cannot be found at path '/bootbank'/var/run/log/vmkernel.log:YYYY-MM-DDTHH:MM:SS.Z cpuxx:####)VMW_SATP_LOCAL: satp_local_updatePath:856: Failed to update path "vmhba32:C0:T0:L0" state. Status=Transient storage condition, suggest retryYYYY-MM-DDTHH:MM:SS.Z cpuxx:####)ScsiPath: ####: Cancelled Cmd(0x####) 0x1a, cmdId.initiator=0x#### CmdSN 0x#### from world 0 to path "vmhba32:C0:T0:L0". Cmd count Active:0 Queued:2.YYYY-MM-DDTHH:MM:SS.Z cpuxx:####)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "mpx.vmhba32:C0:T0:L0" state in doubt; requested fast path state update...YYYY-MM-DDTHH:MM:SS.Z cpuxx:####)ScsiDeviceIO: ####: Cmd(0x####) 0x1a, cmdId.initiator=0x#### CmdSN 0x1407e8 from world 0 to dev "mpx.vmhba32:C0:T0:L0" failed H:0x5 D:0x0 P:0x0 Cancelled from path layer. Cmd count Active:1YYYY-MM-DDTHH:MM:SS.Z cpuxx:####)ScsiPath: ####: Cancelled Cmd(0x####) 0x0, cmdId.initiator=0x#### CmdSN 0x0 from world 0 to path "vmhba32:C0:T0:L0". Cmd count Active:0 Queued:2.
USB devices have a small queue depth. A race condition in the ESXi storage stack causes some I/O operations to be queued rather than sent to the device, eventually leading to a timeout and loss of access to the bootbank.
Fixed in release VMware vSphere ESXi 7.0 U2d and higher. See Download latest ISOs and patches for vSphere ESXi for steps to download this release.
Workaround
Note: It is recommended to upgrade to ESXi 8.x as vSphere 7.x has reached the End of General Support.