After upgrading an host with a SATADOM boot device from ESXi 6.7 to ESXi 7.0, bootbank is mounted in ramdisk /tmp
search cancel

After upgrading an host with a SATADOM boot device from ESXi 6.7 to ESXi 7.0, bootbank is mounted in ramdisk /tmp

book

Article ID: 326201

calendar_today

Updated On:

Products

VMware vSphere ESXi

Issue/Introduction

Symptoms:
  • Your ESXi host is configured to boot from a SATADOM as boot device
  • After upgrade from ESXi 6.7 to EXi 7.0 the boot partition /bootbank is no longer backed by a physical device, but points to /tmp in the ramdisk:
    $vmkfstools -P -v 10 bootbank
    (...
    )UUID: 00000000-00000000-0000-000000000000
    Logical device:
    Partitions spanned (on "notDCS"):
            ramdisk:tmp
    Is Native Snapshot Capable: NO
    OBJLIB-LIB: ObjLib cleanup done.
    WORKER: asyncOps=0 maxActiveOps=0 maxPending=0 maxCompleted=0
    
 
  • The jumpstart log jumpstart-native-stdout.log contains the following errors:
    2022-08-12 11:01:53,519 SystemStorage Waiting up to 120s for boot device to be ready
    2022-08-12 11:03:53,811 SystemStorage Failed to find boot device after 120 seconds
    2022-08-12 11:03:53,812 SystemStorage no elligible disk found for ESX SystemStorage
    2022-08-12 11:03:53,814 SystemStorage Finding OSDATA on non-boot device...
    2022-08-12 11:03:53,815 SystemStorage No persistent storage available for system logs and data. ESX is operating with limited system storage space, logs and system data will be lost on reboot.
    2022-08-12 11:03:53,816 SystemStorage /locker: initializing storage endpoint on non-persistent (RAM-backed) OSDATA (no persistent OSDATA or LOCKER partition found)
    2022-08-12 11:03:53,826 SystemStorage /vmfs/volumes/5addcfe5-d91ac028-6ca2-00e0ed789337/vmkdump/00000000-0000-0000-0000-AC1F6B603E86-6200229888.dumpfile: no such coredump file location, unconfiguring....
    VmkCtl: unable to create diagnostic file: No usable VMFS Volume found for Diagnostic File
    2022-08-12 11:03:53,842 SystemStorage failed to smart assign coredump file: Errors:
    Unable to smart activate a dump file.  Error was: No suitable diagnostic file found
 
  • after another reboot, the system starts with ESXi 6.7 again


Environment

VMware vSphere ESXi 7.x

Cause

In ESXi 6.7 SATADOM devices are using the vmklinux ahci driver. However, with ESXi 7.0 vmklinux drivers have been depricated, therefore the vmw_ahci driver is required for these devices.
When the vmw_ahci driver was disabled in ESXi 6.x though, the SATADOM device won't function, resulting in the issue.

Resolution

To resolve this problem, enable the vmw_ahci driver, before attempting to upgrade to ESXi 7.0:
  • Backup the file /etc/vmware/esx.conf
    # cp /etc/vmware/esx.conf /etc/vmware/esx.bak
 
  • Edit the file /etc/vmware/esx.conf
  • Find the following line:
    /vmkernel/module/vmw_ahci/enabled = "false"
 
  • Change this line to:
    /vmkernel/module/vmw_ahci/enabled = "true"
 
  • Save the file and run the following command to push the change down to the physical boot device:
    # /sbin/auto-backup.sh
 
  • Upgrade the Host to ESXi 7.0.