VMware vSphere

 View Only
  • 1.  Get Around Custom initrd that VMware Tools breaks

    Posted Nov 27, 2013 02:20 PM

    I've got an application vendor who uses a custom initrd to install a custom CentOS image and their software when updates are applied.

    Installing VMware Tools breaks this and ends in a kernel panic on the next reboot. I can provide more details if that is helpful.

    The vendor isn't being terribly flexible about their install process. I wonder if anyone has run into something similar that precludes installing VMware tools. It is putting a bit of a hitch in our plan to virtualize some physical boxes.

    Are there any mechanism for manually injecting the VMware tools components in a fashion that won't break an existing boot process?

    -

    Tim



  • 2.  RE: Get Around Custom initrd that VMware Tools breaks

    Posted Nov 27, 2013 03:03 PM

    Well, if you are going by the traditional method of installing the vmware tools, then the modules are compiled at runtime and hence the need for re-initrd imaging happens. You can work around by

    1. Giving -p while giving vmware-config-tools.pl

    -p, --prebuilt          Force the use of pre-built kernel modules.

      --preserve          Always preserve user-modified configuration
                              files.

      --overwrite         Always overwrite user-modified configuration
                              files.

    2. Installing OSP http://packages.vmware.com/tools/docs/manuals/osp-esxi-51-install-guide.pdf



  • 3.  RE: Get Around Custom initrd that VMware Tools breaks

    Posted Dec 02, 2013 06:01 PM

    Unfortunately all the mechanisms seem to insist on rewriting the initrd and changing the boot sequence

    Message appears during execution of vmware-config-tools.pl with --preserve option: Creating a new initrd boot image for the kernel.

    The problem comes in the init script. (I'm leaving out the VMware-related device driver modules which seem to load without a problem)

    The original is:

    echo Scanning and configuring dmraid supported devices

    echo Scanning logical volumes

    lvm vgscan --ignorelockingfailure

    echo Activating logical volumes

    lvm vgchange -ay --ignorelockingfailure  LearningSpace0

    resume LABEL=Swap00

    echo Creating root device.

    mkrootdev -t ext3 -o defaults,ro /dev/LearningSpace0/RootB

    echo Mounting root filesystem.

    mount /sysroot

    echo Setting up other filesystems.

    setuproot

    echo Switching to new root and running init.

    switchroot

    This gets changed by the VMware tool script to:

    echo Scanning and configuring dmraid supported devices

    resume LABEL=Swap00

    echo Creating root device.

    mkrootdev -t ext3 -o defaults,ro /dev/root

    echo Mounting root filesystem.

    mount /sysroot

    echo Setting up other filesystems.

    setuproot

    echo Switching to new root and running init.

    switchroot

    and, on next restart of the system, I end up with a failed boot  and a kernel panic.

    In case it is helfpul/relevant, grub.conf is:

    title CentOS (2.6.18-371.1.2.el5)

            root (hd0,0)

            kernel /vmlinuz-2.6.18-371.1.2.el5 ro root=/dev/LearningSpace0/RootB

            initrd /initrd-2.6.18-371.1.2.el5.img

    -

    Tim