0

I did the root zfs install and everything worked great. After shutting down I was unable to reboot and got stuck at initrd shell prompt.

Reinstalled from scratch on same SSD. Rebooted. Worked great. Applied Ubuntu updates and then ran tasksel to add some more desktop gui environments. (Same thing I did with earlier failure...) Hmm. Seems repeatable. Anyone have a good set of notes on getting this bootable. Tried to mount the stuff under the 19.10 install -- but ran into issues. The zpool's are there, look good and can be brought on line. I need to figure how to change the mountpoint, and get the stuff fixed so it will boot again.

I know it is experimental -- anyone else see this?

Bill

Improve this question
1
  • I just found out if I go to one of the recovery single-user options for any kernel in grub it works fine. Is this possibly some kind of wait delay time issue that needs to be in grub.
    – Bill Pechter
    Oct 21, 2019 at 21:25

2 Answers 2

Reset to default
1

Obvious thing in that error that I hadn't given much thought: it's line 180 of the NEW grub.cfg (/boot/grub/grub.cfg.new) that is in error.

That line has an extra }. So I deleted that line, ran 

grub-script-check grub.cfg.new

and - no error.

So I copied grub.cfg.new to grub.cfg and rebooted; and it's good. However, update-grub still seems to add the extra curly brace whenever it's run.

Improve this answer
1
0

Yes, I've run into a similar problem on two separate installs. The first time I edited /etc/default/grub to remove "quiet" and "splash" from boot, it worked fine. At some point, some process rebuilt the grub config again and I had your exact problem. Subsequent attempts to run update-grub always fail with a syntax error warning, even if I put back the default /etc/default/grub:

Adding boot menu entry for EFI firmware configuration
error: syntax error.
error: Incorrect command.
error: syntax error.
Syntax error at line 180
Syntax errors are detected in generated GRUB config file.
Ensure that there are no errors in /etc/default/grub
and /etc/grub.d/* files or please file a bug report with
/boot/grub/grub.cfg.new file attached.

I'm a long time Solaris guy, including the latest versions, so I'm quite familiar with working with ZFS. I also run it (not on root) on production CentOS systems. It's been stable there, but again--no root pool.

As far as fixing it, as I recall here's how to get it back if there is an old or new grub.cfg from a failed or previous attempt at running update-grub.

Let's just say your pool is on sda. /boot/grub is on sda2 by default on my systems.

Boot the live CD/USB and go into the "try it" option, however it's worded, but don't install.

Open a command prompt, then type "mount /dev/sda2 /mnt".

cd /mnt, then hopefully you'll find another grub.cfg that worked.

For me, it's the largest file--the new ones are usually about half the original (original 7.1K) and they simply don't work. Copy it to grub.cfg and reboot.

This sounds similar to the bug below, but installing gawk didn't fix it for me. It was supposedly fixed, and may not be the real issue anyway:

https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1834095

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .