Free up space on /boot disk (ubuntu)

free-space-on-boot-disk.md

Free disk space when /boot is full (Ubuntu)

TL;DR

dpkg -l linux-image*
uname -r
sudo apt-get remove linux-image-2.6.32-{21,37,38,39,40,41,42,43,44}-server
sudo apt-get autoremove

This morning, I noticed New Relic that the /boot disk was almost full on several of our servers. It turns out that when the old kernel images are left on the disk after a kernel update. The solution is to delete old kernels that we don't need anymore.

1. Find packages to delete.

You can safely delete all but the latest package that matches the pattern linux-image-2.6.32-[0-9][0-9]-server. Leave linux-image-2.6.32-45-server on your system because you will need it to reboot!

$ dpkg -l linux-image*
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name                                          Version                                       Description
+++-=============================================-=============================================-==========================================================================================================
un  linux-image                                   <none>                                        (no description available)
un  linux-image-2.6                               <none>                                        (no description available)
ii  linux-image-2.6.32-21-server                  2.6.32-21.32                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-37-server                  2.6.32-37.81                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-38-server                  2.6.32-38.83                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-39-server                  2.6.32-39.86                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-40-server                  2.6.32-40.87                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-41-server                  2.6.32-41.91                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-42-server                  2.6.32-42.96                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-43-server                  2.6.32-43.97                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-44-server                  2.6.32-44.98                                  Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-2.6.32-45-server                  2.6.32-45.104                                 Linux kernel image for version 2.6.32 on x86_64
ii  linux-image-server                            2.6.32.45.52                                  Linux kernel image on Server Equipment.

2. Identify the kernel version you are currenlty running

DON'T delete this kernel image! And if it's not the latest, don't delete the latest one either.

$ uname -r
2.6.32-45-server

3. Delete the old kernels.

Removing the old kernels is the same as removing any other package. I'm using shell expansion for the version numbers to save typing. It will prompt you with a list of packages that will be removed, so you can double check the list before continuing.

sudo apt-get remove linux-image-2.6.32-{21,37,38,39,40,41,42,43,44}-server

4. Remove dependencies.

There are some dependencies left on the system after removing the old kernels. Fortunately, you can easily clean these up, too.

sudo apt-get autoremove

Thank you. This is the what I needed and it is explained so everybody can learn not just copy/paste.

I found this link with a one line solution:
http://ubuntuhandbook.org/index.php/2016/05/remove-old-kernels-ubuntu-16-04/

Thank you for solving the issue. It helped me and I'm looking for exactly what described here.

When I try this, it is giving me a message "Package 'linux-image-3.8.0-31-generic' is not installed, so not removed"
This happens for every kernel image that is listed from the dpkg -l output

@keenlogic, I had the same problem "Package 'linux-image-x.y.z-j-generic' is not installed, so not removed"

I believe what is happening is that when you run the "dpkg -l linux-image*" command, it's still listing those "already removed" packages because there are still things depending on them. I guess this means that they haven't actually been removed and are still taking up disk space because I would still get disk space related errors when running other commands.

What I had to do was remove the dependencies manually first, so for each "x.y.z-j" image listed, I would run the following commands in this order:
sudo dpkg --purge linux-signed-image-x.y.z-j-generic sudo dpkg --purge linux-image-extra-x.y.z-j-generic sudo dpkg --purge linux-image-x.y.z-j-generic

Some of those commands would fail for certain packages claiming that they were already removed, usually the third command.

Then another:
dpkg -l linux-image*
would show that the image was no longer listed.

I was careful not to remove my current version and the latest version as suggested in the post.

No doubt you can batch up the commands for multiple images in one go, I just went through them one by one manually to see what was happening at each point, didn't take long.
sudo apt-get autoremove
command now runs happily

jbgo your solution worked perfectly for me. Thanks man.

Filesystem      Size  Used Avail Use% Mounted on
/dev/root        28G   25G  1.7G  94% /
devtmpfs        7.7G     0  7.7G   0% /dev
tmpfs           7.7G   15M  7.7G   1% /dev/shm
tmpfs           7.7G   14M  7.7G   1% /run
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
tmpfs           7.7G     0  7.7G   0% /sys/fs/cgroup
tmpfs           786M   68K  786M   1% /run/user/1001
/dev/mmcblk1p1   31G  9.4G   21G  32% /media/nvidia/9016-4EF8

does this look normal for a nvidia jetson tx2?
i ran out of disk space after a failed installaiton, now i am purging a lot of packages to get disk space.
How can i delete the tmp files created through the failed installation

I just wrote this. My bash is a little rusty though

#!/bin/bash
# BETLOG - 2018-03-31--19-48-34
# based on https://gist.github.com/jbgo/5016064

echo -ne "CAUTION::\\nThis script apt-get removes all but the currently operational kernel"

read -p "Continue? (y/n)?" answer
case ${answer:0:1} in
    y|Y )
        sudo -v
        current=`uname -r`
        available=`dpkg -l | grep -Po "linux-image.*generic"`
        remove=()
        for name in $available; do
            if [[ ${name/$current/} == ${name} ]]; then
                remove+=("$name")
            fi
        done
        sudo apt-get remove ${remove[@]}
        sudo apt-get autoremove
        echo FINISHED
    ;;
    * )
        echo ABORTING;
    ;;
esac

Your very first code line:

dpkg -l linux-image*

Try this instead:

dpkg -l linux-image* | grep "ii"

otherwise the list will be too long on many systems. You may want to update that.

I fixed @BETLOG 's script cause It contains some redundant strings like arch, kernel version (not kernel name) , etc.
Also, I added filtering only 'ii'

#!/bin/bash
# ryul99 - 2023-04-13
# Origin: BETLOG - 2018-03-31--19-48-34
# based on https://gist.github.com/jbgo/5016064

echo -ne "CAUTION::\\nThis script apt-get removes all but the currently operational kernel"

read -p "Continue? (y/n)?" answer
case ${answer:0:1} in
    y|Y )
        sudo -v
        current=`uname -r`
        available=`dpkg -l | grep -Po "linux-image.* | grep ii" | cut -d ' ' -f 1`
        remove=()
        for name in $available; do
            if [[ ${name/$current/} == ${name} ]]; then
                remove+=("$name")
            fi
        done
        sudo apt-get purge ${remove[@]}
        sudo apt-get autoremove
        echo FINISHED
    ;;
    * )
        echo ABORTING;
    ;;
esac