Long boot time for ESXi hosts even after marking RDM LUNs as perennially-reserved
search cancel

Long boot time for ESXi hosts even after marking RDM LUNs as perennially-reserved

book

Article ID: 307069

calendar_today

Updated On:

Products

VMware vSphere ESXi

Issue/Introduction

  • ESXi host has access to RDM LUN’s used for WSFC (Windows Server Failover Cluster).

  • ESXi host boot takes long time even though the perennially-reserved flag is enabled on RDM LUNs.

  • In the ESXi vmkernel.log you will see similar messages as below

     
    YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiDeviceIO: 9232: Get VPD 86 Inquiry for device "naa.600507680c####dca8000###000####" from Plugin "NMP" failed. Not supported
YYYY-MM-DDThh:mm:ss  cpu39:65984)WARNING: celerity16fc: Task management abort failed, status:fffffffd
YYYY-MM-DDThh:mm:ss  cpu39:65984)WARNING: celerity16fc: Task management abort failed, status:fffffffd
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiDeviceIO: 7916: Could not detect setting of QErr for device naa.600507680c####dca8000###000####. Error Timeout.
YYYY-MM-DDThh:mm:ss  cpu28:66342)WARNING: ScsiCore: 1806: Invalid sense buffer: error=0x0, valid=0x0
YYYY-MM-DDThh:mm:ss  cpu28:66342)NMP: nmp_ResetDeviceLogThrottling:3452: Error status H:0x0 D:0x18 P:0x0 Sense Data: 0x0 0x0 0x0 from dev "naa.600507680c####dca8000###000####" occurred 1090 times(of 1090 commands)
YYYY-MM-DDThh:mm:ss  cpu39:65984)WARNING: celerity16fc: Task management abort failed, status:fffffffd
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiDeviceIO: 8430: Could not detect setting of sitpua for device naa.600507680c####dca8000###000####. Error Timeout.
YYYY-MM-DDThh:mm:ss  cpu28:66342)WARNING: ScsiCore: 1806: Invalid sense buffer: error=0x0, valid=0x0
YYYY-MM-DDThh:mm:ss  cpu28:66342)NMP: nmp_ResetDeviceLogThrottling:3452: Error status H:0x0 D:0x18 P:0x0 Sense Data: 0x0 0x0 0x0 from dev "naa.600507680c####dca8000###000####" occurred 2218 times(of 2220 commands)
YYYY-MM-DDThh:mm:ss  cpu6:65607)ScsiDeviceIO: 3015: Cmd(0x4396017a7900) 0x1a, CmdSN 0x17c from world 0 to dev "naa.600507680c####dca8000###000####" failed H:0x3 D:0x0 P:0x0 Invalid sense data: 0x0 0x0 0x0.
YYYY-MM-DDThh:mm:ss  cpu21:65622)ScsiDeviceIO: 3015: Cmd(0x4396017a7900) 0x1a, CmdSN 0x17f from world 0 to dev "naa.600507680c####dca8000###000####" failed H:0x3 D:0x0 P:0x0 Invalid sense data: 0x0 0x0 0x0.
YYYY-MM-DDThh:mm:ss  cpu39:65984)WARNING: celerity16fc: Task management abort failed, status:fffffffd
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiEvents: 300: EventSubsystem: Device Events, Event Mask: 40, Parameter: 0x43072fde9800, Registered!
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiEvents: 300: EventSubsystem: Device Events, Event Mask: 200, Parameter: 0x43072fde9800, Registered!
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiEvents: 300: EventSubsystem: Device Events, Event Mask: 8, Parameter: 0x43072fde9800, Registered!
YYYY-MM-DDThh:mm:ss  cpu21:66065)ScsiDevice: 4450: Successfully registered device "naa.600507680c####dca8000###000####" from plugin "NMP" of type 0

Cause

The device identifier contains uppercase letters in the naa string, which can cause the perennially-reserved setting to not function as expected. For example:

esxcli storage core device setconfig -d naa.6005#####C8080DCA800####00A51 --perennially-reserved=true

Resolution

Ensure that the device identifier is entirely in lowercase when setting the perennially-reserved flag on the correct LUNs.

esxcli storage core device setconfig -d naa.6005######8080dca8000#####0000a51 --perennially-reserved=true

For more information on setting the RDM LUNs to perennialResvd = "true", refer to VMware KB: ESXi/ESX hosts with visibility to RDM LUNs being used by MSCS nodes with RDMs may take a long time to start or during LUN rescan
Powered by Wolken Software