Linux remove old kernels

Ubuntu 18.04 remove all unused old kernels

H ow do I remove all unused old kernels in Ubuntu Linux 18.04/20.04 LTS using the command line?

Over time you get many Linux kernel packages installed on Ubuntu Linux 16.04 or 18.04 LTS server. All unused Linux kernel takes disk space on Ubuntu server, hence you must delete them from the system. This page shows you how to remove old kernels from an Ubuntu server install. The following commands works with Ubuntu 18.04 and 20.04 LTS server or desktop systems.

Tutorial details
Difficulty level	Easy
Root privileges	Yes
Requirements	Ubuntu 18.04 LTS
Est. reading time	4 minutes

Ubuntu 18.04 remove kernel that is not used

The procedure to delete all unused old kernels on Ubuntu Linux version 18.04 and 20.04 LTS is as follows:

1. First, boot into a new kernel
2. List all other older kernel using the dpkg command
3. Note down system disk space usage by running the df -H command
4. Delete all unused old kernels, run: sudo apt —purge autoremove
5. Verify it by running the df -H

How to list all installed Linux kernel

Simply type the following dpkg command along with egrp command:
dpkg —list | egrep -i —color ‘linux-image|linux-headers’
# only installed #
dpkg —list | grep -i -E —color ‘linux-image|linux-kernel’ | grep ‘^ii’

Listing old kernel for removal purpose

Also note down the amount of disk space available on the file system by typing the following df command:
df -H

Total 9.3G disk space used on my cloud server

How to remove old kernel versions on Ubuntu 18.04 or above

Dealing with Ubuntu 16.04 LTS and older

First, get list of currently running Linux kernel, run:
v=»$(uname -r | awk -F ‘-virtual’ ‘< print $1>‘)»
echo «$v»
Sample outputs:

Next, I am going to create a list that tells not to delete currently running Linux kernel as follows:
i=»linux-headers-virtual|linux-image-virtual|linux-headers-generic-hwe-|linux-image-generic-hwe-|linux-headers-$|linux-image-$(uname -r)|linux-image-generic|linux-headers-generic»
echo «$i»
Sample outputs (the following kernel I am going to keep on my systems):

Here is a list of the kernel that needs to be deleted:
dpkg —list | egrep -i ‘linux-image|linux-headers’ | awk ‘/ii/< print $2>‘ | egrep -v «$i»
Sample outputs:

How to delete old kernels in Ubuntu

Finally, I am going to delete all unwated and older Linux kernels on my Ubuntu Linux 16.04 LTS server, run:
$ sudo -i
# apt-get —purge remove $(dpkg —list | egrep -i ‘linux-image|linux-headers’ | awk ‘/ii/< print $2>‘ | egrep -v «$i»)

Result

Again run the df command:
df -H
Sample outputs:

• No ads and tracking
• In-depth guides for developers and sysadmins at Opensourceflare✨
• Join my Patreon to support independent content creators and start reading latest guides:
  • How to set up Redis sentinel cluster on Ubuntu or Debian Linux
  • How To Set Up SSH Keys With YubiKey as two-factor authentication (U2F/FIDO2)
  • How to set up Mariadb Galera cluster on Ubuntu or Debian Linux
  • A podman tutorial for beginners – part I (run Linux containers without Docker and in daemonless mode)
  • How to protect Linux against rogue USB devices using USBGuard

Join Patreon ➔

Now I have used 6.7G disk space as compared 9.3G before deleting kernels. In other words, I gained 2G disk free space.

Removing old kernels using GUI software updater tool

Ubuntu comes with graphical management of software packages updates called update-manager. It is a frontend for the apt command package management system. Start update manger from GUI or open the terminal app, then type:
update-manager
Make sure you select an option called “Unused kernel updates to be removed” > click on the Install Now button:

Naturally, the GUI method works on the Ubuntu desktop. Server users need to stick with the CLI method.

Conclusion

And there you have it. You successfully removed old kernel and header packages from the Ubuntu Linux system. On really older Ubuntu system, we had a command named purge-old-kernels but t is now deprecated. Hence, you must use the apt/apt-get.

🐧 Get the latest tutorials on Linux, Open Source & DevOps via

Источник

Debian / Ubuntu Linux Delete Old Kernel Images Command

Delete Old Kernel Images

The procedure is as follows for a Debian or Ubuntu Linux system. You must login as root user.

Step 1 – Boot into new kernel

First, boot into newly installed kernel. Verify this with the following command:
$ uname -mrs
$ uname -a
Sample outputs:

To list all installed Linux kernel images, type the following dpkg command:
# dpkg —list | egrep -i —color ‘linux-image|linux-headers’
Sample outputs:

Fig.01: Check what kernel image(s) are installed on your system (click to enlarge)

Step 2 – Delete unwanted and unused kernel images

You can remove kernel images one by one using the following apt-get command or apt command syntax:
# apt-get —purge remove linux-image-3.13.0-67-generic
OR
$ sudo apt-get —purge remove linux-image-3.13.0-67-generic
The purge removes packages and any configuration files are deleted too.

A note about newer Ubuntu and Debian based system

On newer system all obsolete kernels and headers should automatically be flagged as no more needed, and thus can be purged with the following single command:
$ sudo apt —purge autoremove
OR
$ sudo apt-get —purge autoremove
Sample outputs:

Please note that the autoremove is used to remove packages that were automatically installed to satisfy dependencies for other packages and are now no longer needed. This may remove rest of unwanted packages too.

Understanding package states in Ubuntu and Debian Linux

Consider the following example:
# dpkg —list | grep linux-image
Sample outputs:

The first column indicates package flags like rc, ii. So, what do the various dpkg flags like ‘ii’ ‘rc’ mean?

• rc : It means package is in remove/deinstall state and only config file exists.
• ii : It means package is in install state and it is 100% installed on the system.

You can remove all linux-image packages in rc state using the following command:
# x=$(dpkg —list | grep -i linux-image | grep ^rc| awk ‘< print $2>‘)
# echo «$x»
# apt-get —purge remove $x
Sample outputs:

Type the following command again to see the results:
# dpkg —list | egrep -i —color ‘linux-image|linux-headers’
Sample outputs:

Источник

How to Delete Old Unused Kernels in Debian and Ubuntu

In our last article, we’ve explained how to delete old unused kernels in CentOS/RHEL/Fedora. In this article, we will explain how to delete old unused kernels in Debian and Ubuntu systems, but before moving further, you may want to install the latest version in order to take advantage of: security fixes, new kernel functions, updated drivers and so much more.

To upgrade your kernel to latest version in Ubuntu and Debian, follow this guide:

Important: It is advisable to keep at least one or two old kernels to fall back to in case there is a problem with an update.

To find out the current version of Linux kernel running on your system, use the following command.

To list all installed kernels on your system, issue this command.

Remove Old Unused Kernels on Debian and Ubuntu

Run the commands below to remove a particular linux-image along with its configuration files, then update grub2 configuration, and lastly reboot the system.

Although this method works just fine, it is more reliable and efficient to use a handy script called “byobu” that combines all the commands above into a single program with useful options such as specifying number of kernels to keep on the system.

Install byobu script package which provides a program called purge-old-kernels used for removing old kernels and header packages from the system.

Then remove old kernels like so (the command below allows 2 kernels to be kept on the system).

You may also like to read these following related articles on Linux kernel.

In this article, we have described how to remove old unused kernel images on Ubuntu and Debian systems. You can share any thoughts via the feedback from below.

If You Appreciate What We Do Here On TecMint, You Should Consider:

TecMint is the fastest growing and most trusted community site for any kind of Linux Articles, Guides and Books on the web. Millions of people visit TecMint! to search or browse the thousands of published articles available FREELY to all.

If you like what you are reading, please consider buying us a coffee ( or 2 ) as a token of appreciation.

We are thankful for your never ending support.

Источник

Ubuntu Documentation

LVM installs and encrypted installs use a separate /boot partition. The partition by default is capable of holding only four or five kernels, and can fill to capacity quickly. To prevent your /boot partition from getting full, you need to configure automatic removal of old kernels, or manually remove old kernels regularly.

Changing the kernel providing packages on your system requires commands with root access, so please read RootSudo.

Regular Maintenance

Removing old kernels is easy to do on a system with sufficient free space in your root parition or separate boot partition. You can remove them manually, or configure unattended-upgrades to do it automatically. If you receive or have received an error from a package management tool, manual removal may not work until the problem is fixed. See Problems section.

Manual Maintenance

Using Apt

You can remove old kernels with a simple autoremove command in a terminal:

Note: In Ubuntu 14.04, because of Bug #1439769 this works only, if you have installed security updates automatically, and not manually by e.g. Software Updater, see in this section on how to configure it.

This just adds unnecessary complexity, replace instructions after 14.04 is EoL. In 16.04 and newer you could alternatively use

The packages to remove are determined based in part on whether the package is marked as manually or automatically installed. You can check if a kernel providing package is marked as automatically installed using this command in a terminal:

or to see if it is marked as manually installed:

In the event some kernel providing packages are marked as manual (possibly because they were installed via ‘apt-get’) and you want them autoremoved, you can change the status of the package using apt-mark auto and the package name. For example, to mark kernel 4.12.0-12-generic as autoremovable:

Note: apt-get autoremove will not remove all automatically installed old kernel providing packages as fallback versions are kept; the list of kept kernels is maintained and automatically updated in the file /etc/apt/apt.conf.d/01autoremove-kernels as a list of matching regular expressions.

Other Methods

If you want to purge one specific kernel providing package you can do so via the following command in a terminal:

This will also purge the corresponding linux-image-extra package if it is installed e.g. linux-image-extra-4.12.0-12-generic. It will not purge linux-headers-4.12.0-12, if there is another linux-headers-4.12.0-12 flavor installed besides -generic.

If you just need to purge kernels selectively, you may benefit from this answer with unofficial code, if the system is not already broken.

There is an unofficial script for purging kernels; it is called linux-purge. By it you can purge kernels selectively, or choose to keep given number of older kernels and remove other depending on options given. It can do do even some fixing (with —fix option), if system is already broken; that is a kind of scripted version of what is told in chapter Safely Removing Old Kernels.

Automatic Maintenance

The unattended-upgrades package, included with the default install of all Ubuntu flavors, includes a feature to remove unused packages automatically. Enabling this feature is a two-step process in Ubuntu 14.04.

Enable Unattended Upgrades (Ubuntu 14.04)

In Ubuntu 16.04 and later unattended-upgrades is enabled by default, however in Ubuntu 14.04 you will need to enable it either via the GUI or a command-line. Note: There is a separate documentation about setting up automatic security updates.

GUI Way

Enable unattended upgrades using Software & Updates application’s «Updates» tab:

• Check the box for *-security (and/or any other repositories you wish)
• Automatically check for updates: Set to any frequency (except ‘Never’)
• When there are security updates: Set to Download and Install Automatically

Shell Way

By default this installs available security updates every day. If you want to change which sort of packages it will upgrade (if any), or how often, see here for details.

Configure Unattended Upgrades to Remove Unneeded Kernels Automatically

Note: The following methods will only remove kernels that are marked as being automatically installed as described above. In Ubuntu 16.04 kernels installed by Software Updater are marked as being automatically installed. In Ubuntu 14.04 only kernels installed by unattended-upgrades are marked as being automatically installed. See bug #1439769 for details.

Note: This way will not remove all automatically installed old kernel providing packages as fallback versions are kept; the list of kept kernels is maintained and automatically updated in the file /etc/apt/apt.conf.d/01autoremove-kernels as a list of matching regular expressions.

The second step is to edit old configuration file to enable automatic removal, or preferably create an overriding new configuration file where you need to write only options that you change from the default or from the values defined in the shipped one, because updates to shipped configuration file may conflict with the local changes blocking updating unattended-upgrades itself.

The default configuration file is

/etc/apt/apt.conf.d/50unattended-upgrades

You could name the new one as /etc/apt/apt.conf.d/52unattended-upgrades-local

See here for reference.

Option for All Ubuntu Releases

The following setting configures unattended-upgrade to remove unused dependencies after an unattended upgrade.

Make sure the configuration file contains line

(and that it is not commented out; comments start with ‘//’.)

Thereafter unattended-upgrades will automatically remove packages providing old kernels as part of unattended upgrade. (It does not purge them, however.) It also removes other unneeded packages, as well, which is or is not what you want.

Options for Ubuntu 16.04 and Newer

Unattended-upgrades version 0.90 supports a new configuration variable called Unattended-Upgrade::Remove-New-Unused-Dependencies that makes possible to automatically remove only packages that become excessive during a run of unattended-upgrades. It is enabled i.e. «true» by default. The way this is designed, it is important that you let unattended-upgrades handle installation of security updates. Otherwise unattended-uprades will not remove old kernels and you may have to do some manual removing of kernels.

Version 1.1 that is also currently available for Ubuntu 16.04 adds yet another option called Unattended-Upgrade::Remove-Unused-Kernel-Packages. It removes unused kernel packages before installing new one (or even if nothing is installed). It is enabled by default, too.

Problems

If your package management is broken, or if regular maintenance above is not working, any of several problems may have occurred. You may be out of storage space, or you may have a package version mismatch, or you may have another problem.

Safely Removing Old Kernels

For users of LVM systems, encrypted systems or limited-storage systems, the most frequent problem is that the /boot partition is simply full. The package manager cannot install a pending upgrade due to lack of space. Addiitionally, apt-get can not remove a package due to broken dependency.

This problem can be fixed from the shell: by manually remvoing one or two old kernels, we can provide the package manager enough space to install the queued upgrade.

First remove any leftover temporary files from previous kernel updates.

Determine the version number of the currently running kernel, which you DO NOT want to remove.

List all the kernels, including the booted one (4.2.0-21-generic in this example), in the package database and their statuses.

To free space in /boot we’ll remove an initrd.img file for a suitable old kernel manually, this is necessary due to a kenel packaging bug.

Now we’ll use dpkg in order to TRY to purge the kernel package for the same old kernel:

NOTE: The previous command will probably fail, as there probably is a depending linux-image-extra package installed together with a ‘generic’ kernel package. In general, the output of the previous command will tell which package you need to remove first. In this example case you would run

Finally, we will fix the package installation process that previously failed.

If the last command fails due to insufficient disk space in /boot, you have to purge another kernel (e.g. linux-image-4.2.0-16-generic) same way. Otherwise you could continue to chapter Manual Maintenance to remove more kernels. Make sure the linux-headers packages respective to the kernels purged above will also be purged.

Unmet Dependency Errors

If you have ignored apt and dpkg errors for a while, then some packages may have upgraded while others did not. This is an expected follow-on effect. You will see mysterious, persistent version errors upon normal upgrades.

The simple way to fix most version mismatch errors is to update the package database, clean out the package cache, and download-and-reinstall the newer version of the offending package.

This is easier than it sounds. For example, if the ‘hello’ package is one of those unmet dependencies:

Oops, Removed All Kernels!

If you got a bit carried away and deleted all the kernels, you fall into the «You did WHAT. » class. You will be pleased to know that you are not the first, nor will be the last, person to do this. Get your LiveCD and head over to GRUB reports no operating system.

RemoveOldKernels (последним исправлял пользователь jarnos 2020-09-26 09:58:55)

The material on this wiki is available under a free license, see Copyright / License for details
You can contribute to this wiki, see Wiki Guide for details

Источник