Last updated 14th November, 2017
Overview
If your VPS isn’t responding as you would like, the quickest way to diagnose the problem is to check what appears in the KVM. This guide explains how to modify the display so that everything is displayed in the console and the KVM.
Please note that for some environments, the KVM won’t provide any useful information because the boot sequence occurs in the serial console, or the GRUB is configured in silent mode.
Requirements
- You must have access to a VPS or a Public Cloud instance in rescue mode
Instructions
If your VPS is working normally, go directly to step 4.
These modifications will change the GRUB configuration. Be sure to perform backups before making any modifications. OVH cannot be held responsible for damage or loss of data following these operations.
Step 1: perform initial verification
After connecting, you should check the name of the disk with the command lsblk:
lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 3G 0 disk
└─sda1 8:1 0 3G 0 part /
sdb 8:16 0 10G 0 disk
└─sdb1 8:17 0 10G 0 part
Here, the primary disk is sdb and the primary partition is sdb1 (sdais the rescue disk and sda1 is the primary rescue partition mounted on /).
If the result is the following, your primary disk is vdb and your primary partition is vdb1 (vda is the rescue disk and vda1 is the primary rescue partition mounted on /):
lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
vda 8:0 0 3G 0 disk
└─vda1 8:1 0 3G 0 part /
vdb 8:16 0 10G 0 disk
└─vdb1 8:17 0 10G 0 part
For this guide, we will use sdb. If your disk is vdb, simply replace sdb with vdb for each command.
Step 1b: for VPS only
On a VPS in rescue mode, the primary disk is already mounted. Therefore, it first needs to be unmounted and then remounted with the correct settings:
umount /dev/sdb1
Step 2: mount the disk
Now that the name of the disk has been identified, you can mount it with the correct settings:
mount /dev/sdb1 /mnt
mount -t proc none /mnt/proc
mount -o bind /dev /mnt/dev
mount -t sysfs none /mnt/sys/
These commands will then allow you to use the chroot command and initiate the commands that require access to the sys, dev and proc directories.
Step 3: launch the CHROOT command
To apply it directly to the system, type the following command:
chroot /mnt
From now on, all commands will be applied to your VPS and not to your rescue mode.
When you are in rescue mode, you see the following screen:
sh
root@serveur-3:~#
After performing the chroot, you will have:
sh
[root@serveur-3 ~]#
Note the square brackets [ ], confirming that you are now in a CHROOT environment.
Step 4: make the modifications
In order to access the boot log in the KVM, make sure you have the following values in the /etc/default/grub file:
- For CentOS 6 and 7:
GRUB_TERMINAL_OUTPUT="console"
GRUB_CMDLINE_LINUX="crashkernel=auto rhgb"
GRUB_CMDLINE_LINUX_DEFAULT="console=tty0 console=ttyS0"
If you don’t have these values, edit and modify your file, then save it.
Then use the following command to regenerate the GRUB configuration file (the values will be saved for the next reboot):
grub2-mkconfig -o "$(readlink /etc/grub2.cfg)"
- For Debian and Ubuntu:
GRUB_CMDLINE_LINUX_DEFAULT="console=ttyS0 console=tty0"
If you don’t have these values, edit and modify your file, then save it.
Then use the following command to regenerate the GRUB configuration file (the values will be saved for the next reboot):
update-grub
Once you have made the modifications, reboot the VPS or the instance in normal mode and check the KVM: the boot log information should appear.
Go further
Join our community of users at https://community.ovh.com/en/.