[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On 26/05/16 21:14, Adrian Midgley wrote: > Got that. > Ok... > So, boot back into rescue system and as for a shell in the installer > environment rather than using a partition as root? > > Graphical rescue mode running... Exactly - you want to check at this point that you don't have anything from your disks mounted automatically. Check /etc/mtab to see what's currently mounted and ensure nothing from /dev/sda or /dev/sdb is mentioned. Ping 8.8.8.8 or whatever and you're ready. I'm going to cheat and simply use an excellent page that I always crib from (I'm not going to pretend that I can remember this off the top of my head)... http://linuxandfriends.com/how-to-reinstall-grub2-chroot-into-a-linux-partition This is your basic recipe, ignoring this guy's more complex single disk windows+linux layout: yours is more direct luckily. Create a new place to hold the chroot, I prefer something unmistakable such as: mkdir /TARGET Substitute this or whatever you prefer in the commands lifted from the webpage, but essentially: mount -t ext4 -o rw /dev/sda2 /TARGET mount -t proc none /TARGET/proc mount -o bind /dev /TARGET/dev mount -o bind /sys /TARGET/sys mount -t devpts devpts /TARGET/dev/pts In your case with a slightly unusual layout, you might want to mount your sdb partitions in the right place so if /home lives on sdb1 for example, you can mount it with: mount -t ext4 /dev/sdb1 /TARGET/home And the same for any other partitions from sdb. However, /home is a bad example as you don't NEED to mount it for this - all your chroot operations will be performed as root and home can remain unmounted. Same with any other unnecessary partitions from sdb. You also don't NEED to mount your EFI sda1 for this operation as it would be wise to just do the kernel first. When you're ready: chroot /TARGET /bin/bash You'll be dropped into a root shell on your resurrected system - at this point you are now genuinely root on your proper OS so be careful, it's "real" from now on in. Make sure you still have internet access - should be fine. apt-get update apt-get dist-upgrade Check your current kernel(s): dpkg -l | grep linux-image You may see some old kernels marked with (rc) - these are old kernels that were installed but have been removed and not yet fully purged. These are of particular interest as they presumably worked before and are good candidates for reinstallation. You will see I believe one kernel in your case marked (ii) - this is your currently faulty kernel. apt-get install linux-image- (hit tab here to see what's available). On Debian 8 the latest I think will be: apt-get install linux-image-3.16.0-4-amd64 Force it with --reinstall if necessary: apt-get install linux-image-3.16.0-4-amd64 --reinstall Also get the relevant headers: apt-get install linux-headers-3.16.0-4-amd64 --reinstall If you have the backports repo enabled (recommended) you can grab a much more up to date 4 series kernel. cat "deb http://ftp.debian.org/debian jessie-backports main" >> /etc/apt/sources.list apt-get update && apt-get dist-upgrade apt-get -t jessie-backports install linux-image-4.5.0-0.bpo.2-amd64 linux-headers-4.5.0-0.bpo.2-amd64 If there are more recent versions available, grab them instead. Force reinstall with --reinstall as necessary. Make sure to install at least two sets of linux-image + linux-headers and any dependencies dragged in. You will end up with fallback options to boot from to test afterwards. Obviously, any errors reported here are critical so keep an eye out for anything the shell spits out that doesn't look as expected and copy them out for us to look at. This should be it - drop out of your chroot shell with "exit", and then umount ALL the stuff you mounted earlier. When /etc/mtab in your installer shell is cleaned of your internal disks, issue "reboot" and cross your fingers! Interrupting GRUB on reboot should now list all the kernels you've just (re)installed - I'd default to booting whatever the newest is. If all has gone well, you'll be up and running again: if not, don't give up yet. Another chroot pass might be required, this time also pulling in the EFI partition and potentially reinstalling GRUB this time around. Let us know how you get on - good luck. Cheers -- The Mailing List for the Devon & Cornwall LUG https://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq