• save the script below to a file on the root filesystem
• reboot into bash init=/bin/bash
• execute the script

The script will exit if any of the steps fail

#!/bin/bash
fail_notify() {
 echo "step "$@" on line $BASH_LINENO failed <><><><><><><><><><><><><><><"
 exit
}

grep 'vgsys-var_log[[:blank:]]' /etc/fstab
if [ $? == 0 ]
   then echo "/var/log filesystem already exists"
else
   start_udev || fail_notify "start_udev"
   mount -o rw,remount / || fail_notify "remount of root"
   vgchange -ay vgsys || fail_notify "activate vgsys"
   lvchange -ay /dev/vgsys/var || fail_notify " activate lv var"
   vgscan --mknodes -v || fail_notify "creating device nodes"
   lvcreate vgsys -L 2G -n var_log || fail_notify "lvcreate"
   mkfs.ext4 /dev/vgsys/var_log || fail_notify "mkfs"
   mkdir /tmp/var_log # no test as the directory may already exist if rerunning script
   mount /dev/vgsys/var_log /tmp/var_log || fail_notify "mount"
   mount /var || fail_notify "mount var"
   cd /var/log/ ||   fail_notify "cd /var/log "
   echo "copying files from /var/log/ to new file system /tmp/var_log/"
   find . -mount -print0 | cpio -0dump /tmp/var_log/ || fail_notofy "find"
   # need to put this line before  /var/log/audit mount point otherwise the  the file systems will overlap
   sed -i '/vgsys-var_log_audit/i \/dev\/mapper\/vgsys\-var\_log\ \/var\/log\             \ ext4  \ nodev\,nosuid\,noexec      \ 1\ 2' /etc/fstab  || fail notify " fstab update"
   df # display a df to the user so they can check what has happened
   echo "remounting filesystems as read only"
   mount -o ro,remount /tmp/var_log || fail_notify "remount of var_log"
   mount -o ro,remount /var || fail_notify "remount of var"
   mount -o ro,remount / || fail_notify "remount of /"
   echo "+--------------------------+"
   echo "| power cycle vm to finish |"
   echo "| or execute "exec init 6" |"
   echo "+--------------------------+"
fi