0

When attempting to boot Debian, it stops at the following message. It seems to me the crucial line of failure is "Unable to load firmware patch rtl_nic/rtl8168d-1.fw (-2)". The lines under that seem like it's unable because it can't connect to the internet.

Phone Screenshot

This machine has started up before. This boot problem came into existence after I had to reset the computer because of an infinity loop in a subthread, and it didn't allow me to switch process anymore. It also used to be able to connect to the internet and install packages/use firefox just fine.

Possible related to Debian stretch failed to load firmware rtl_nic/rtl8168g-3.fw (-2), and I'd love to attempt the same fix. However, that failure was non-lethal and the user was able to boot into debian and fix it. My problem happens at a moment I cannot use Ctrl+Alt+FX to get into a terminal, and it does not progress beyond this screen.

Any other google results I got also had the user being able to boot into GUI/Terminal just fine.

My apologies for the screenshot being a bit out of focus. The screen also started flickering, which made it hard to get a good photo.

I don't have an installer USB, but I could make one.

I'm not that experienced with unix in general or debian, so please keep that in mind for answers.

Anybody got pointers how to recover this installation ?

Gloweye
  • 123

2 Answers2

0

Changing the definition of your interfaces in /etc/network/interfaces from auto to allow-hotplug. It will stop it hanging.

As in:

allow-hotplug eth0
iface eth0 inet dhcp

To modify the file, you have either to boot in single mode, or easier for a newbie, booting from a live image in a pen.

Rui F Ribeiro
  • 56,709
  • 26
  • 150
  • 232
  • That would be very nice, if I could access a terminal. Or can I type that in the grub-terminal ? – Gloweye Aug 20 '18 at 09:13
  • I've attempted to boot in single mode(after googling it), but that fails with the same problem. As in booting from a live image, you mean another copy installed on a USB ? I've tried that before, but that wouldn't boot on this (rather ancient) machine. – Gloweye Aug 20 '18 at 10:39
  • Burn a DVD then. Although I suspect you must be making some mistake jumping to single user mode. – Rui F Ribeiro Aug 20 '18 at 10:41
  • I'm afraid I might have not clarified enough. I got an installation USB now, but I should install debian itself on the USB so I can run a terminal from there ? – Gloweye Aug 20 '18 at 11:24
  • Since I've got a newer version of Debian, I need to use enp0s1, or some variant of that one. For the rest, this seems to have gotten me a bit closer, with some changes. I'll update the question with my results after I can boot without recovery mode again. – Gloweye Aug 20 '18 at 14:38
  • Yeah, new naming scheme for interfaces. it is just an example. – Rui F Ribeiro Aug 20 '18 at 15:23
  • Well, I didn't know there was a new naming scheme, so I was googling something like "why dont I have eth0 but it works under windows dual boot". Never underestimate how little new people know. I stubled across the new scheme by accident. – Gloweye Aug 21 '18 at 08:41
  • AntiX here in the desktop and CentOS/RH servers with the new naming scheme disabled – Rui F Ribeiro Aug 21 '18 at 09:10
0

Inability to connect to the internet should not cause the system to fail quite that drastically. I think you may have other problems, and their messages might have already scrolled off the screen. You could use Shift + PgUp/PgDown keys to scroll the messages backwards some amount; that might allow you to see the earliest abnormal messages.

At the top of your screenshot, there is a message:

You are in rescue mode. After logging in, type "journalctl -xb" to view 
system logs, "systemctl reboot" to reboot, "systemctl default" or ^D to
boot into default mode.

Normally, there should be a prompt to enter a root password after this text. But in your case, there is another message:

Cannot open access to console, the root account is locked.
See sulogin(8) man page for details.

Press Enter to continue.

This means the system was going to allow you rescue mode access, but there does not seem to be a valid root password available. You'll need to fix that first.

Reboot the system, interrupt GRUB when it displays the menu and press E to edit the boot options. Move to the kernel line and add init=/bin/bash to the end of it. Then press the appropriate key (indicated on the screen) to boot the system with the temporarily-modified options.

This should bring the system up to a point where only the root filesystem is mounted, and even that is in read-only mode, but you get a # root prompt straight away.

First, get the root filesystem into writeable state:

mount -o remount,rw /

Then, use the passwd command to set a new password for root. It will prompt you to enter the new password twice. If you plan to use characters that are not numbers or US-ASCII letters, use the command prompt to test the keyboard mapping before setting the password: it's possible that the keyboard mapping of your choice is not in effect and you're falling back to US English default layout in this mode.

Once you've successfully set a root password, it is time to restart the system again. But because of the special way used to start the system, some special steps are needed.

mount -o remount,ro /
sync
reboot -f

On next boot, the system will probably again drop into rescue mode, but now there should be a prompt to enter a root password. Enter the root password you previously set, and you should get the root prompt. Finally you can follow the suggestion to use the journalctl -xb command to view the system log.

The logs will normally contain a lot of informational (non-error) messages. You should find the earliest error message; that is probably the root cause of the boot process going off the rails.


An alternative method for gaining access to a system that has severe problems is to boot from an installation media (e.g. your installation USB stick) using the rescue boot parameter. The installer will present the standard language/keyboard layout questions, and then switch into a special rescue mode that attempts to detect an existing Debian installation, mount its disks and provide you with a command prompt + a set of tools to fix things.


Oh, and to fix your missing Realtek firmware: use another computer to grab the firmware-realtek .deb package, save it on some removable media, move it to your computer and use e.g. dpkg -i firmware-realtek*.deb to install it.

telcoM
  • 96,466
  • I got into terminal access again. Thanks for that, anyway. The real problem is still out there, but at least I can start googling for that again. – Gloweye Aug 20 '18 at 12:25
  • Added the instructions for downloading & installing your missing firmware-realtek package. But I'll be surprised if the lack of firmware is the only reason for your system crashing into rescue mode at boot. – telcoM Aug 20 '18 at 12:42
  • It's one of them. I managed to find that one through the linked question. I had to download another one through https://joshtronic.com/2017/11/06/fixed-radeon-kernel-modesetting-for-r600-or-later-requires-firmware-amd-graphics/ . I also had to reconfigure network connections before doing all that ( https://wiki.debian.org/NetworkConfiguration gave me most for that, but am still working on that one.) I'll update the question with the entire journey. – Gloweye Aug 20 '18 at 14:36
  • I made some progress and can attempt login through the GUI. However, it just returns me to the login screen again. Could this have anything to do with the fact that I installed with the root account disabled and with the user account as legal sudo'er, and stuff got mixed up re-enabling the root account ? – Gloweye Aug 20 '18 at 15:05
  • @JaccovanDorp Please do not update the question so much. Write an answer to your own question if you want to report what you have done. – Rui F Ribeiro Aug 20 '18 at 15:26