After some time WebDevelopers local machine (as well as server) gets full of stuff not needed, here are some commands that I use to erase unnecessary stuff when needed.

I’m planing to update this article each time I find something new so if you like this article keep the link somewhere ;-)

Warning as this Article is about destructive actions, be really careful what you do and don’t execute destructive commands mentioned here if you don’t understand what they are doing. I’m not taking any responsibility if you remove/destroy some data just because you didn’t understood what the command is doing.

Don’t just copy-paste it !!!

Standard disk analyzing commands:

Here are some commands to analyze disk space usage, and other stuff.

Overal disk space usage in human readable format

df -h

Disk usage of Inodes

If you are writing small files on your server (e.g.: file storage cache server) it may happen you will run out of Inode space.

df -ih

Disk usage of particular folder

du -sh /tmp/
sudo du -sh  /etc/

Memory usage

free -m

Removing Docker images

Get rid of all untagged images.

# non sudo version
docker rmi -f $(docker images | grep "<none>" | awk "{print \$3}")

# sudo version
sudo docker rmi -f $(sudo docker images | grep "<none>" | awk "{print \$3}")

Idea stolen from Damien Coraboeuf

Get rid of old live releases

I’m tagging my every docker image release with live-yyyymmdd, for example: quay.io/equivalent/myproject:20151129_0001

…so in order to remove Images of year 2015 I can do:

# non sudo version
docker rmi -f $(docker images | grep live-2015 | awk "{print \$3}")

# sudo version
sudo docker rmi -f $(sudo docker images | grep live-2015 | awk "{print \$3}")

…in order to remove January images of 2015 I can do:

# non sudo version
docker rmi -f $(docker images | grep live-201601 | awk "{print \$3}")

# sudo version
sudo docker rmi -f $(sudo docker images | grep live-201601 | awk "{print \$3}")

… or if you use format like live-yyyymmdd_xxxx where xxxx is release number of a day (live-20160130_0002) you can do

# non sudo version
docker rmi -f $(docker images | grep live-201602.._ | awk "{print \$3}")

# sudo version
sudo docker rmi -f $(sudo docker images | grep live-201602.._ | awk "{print \$3}")


# two dots in this context represent regular expression "any two char" before underscore

…or if you are a Ruby developer and you prefer Ruby syntax or you have no idea what awk does:

# non sudo version
docker rmi -f $(docker images | ruby -ne 'puts $_.split[2] if $_.match(/live-201602\d\d_/)')

# sudo version
sudo docker rmi -f $(sudo docker images | ruby -ne 'puts $_.split[2] if $_.match(/live-201602\d\d_/)')

Quick tip: similar way you can quickly run Rails console on running docker container running Rails (assuming your docker image name contains word rails) rails_container_id=$(sudo docker ps | ruby -ne 'puts $_.split.first if $_ =~ /rails/') && sudo docker exec -it $rails_container_id rails c

Related article: How to lunch bash specific docker image name or Docker container name

Docker containers cleanup

To see list of all docker containers

docker ps -a

# sudo version
sudo docker ps -a

To see overal size of containers

docker ps -as

# sudo version
sudo docker ps -as

#   NAMES        #...                 SIZE
# b455d4dc8320   #...  2 B (virtual 1.512 GB)
# 9aaf4133edd6   #...  0 B (virtual 132.3 MB)
# ed37f797b6f4   #...  0 B (virtual 132.3 MB)
# 96db55573ae8   #...  2 B (virtual 1.512 GB)

But this is not actually a full picture !

As described by Maciej Łebkowski in his excelent article:

docker run leaves the container by default. This is convenient if you’d like to review the process later – look at the logs or exit status. This also stores the aufs filesystem changes, so you can commit the container as a new image. This can be expensive in terms of disk space usage, especially during testing. Remember to use docker run –rm flag if you don’t need to inspect the container later. This flag doesn’t work with background containers (-d), so you’ll be left with finished containers anyway.

So in order to remove this dead containers run this command:

docker ps --filter status=dead --filter status=exited -aq | xargs docker rm -v

# sudo version

sudo docker ps --filter status=dead --filter status=exited -aq | xargs sudo docker rm -v

I had a situation where I runned every command possible but still my /var/lib/docker/containers had several GB. This command dropped 100% usage to 30%

Related article: How to lunch Rails console in specific Docker image name or Docker container name

Removing old release Git branches

After some time release branches piles up and we may want to clean up our Github from old live-* branches

Given we name our release branches live-20150821 (live-yearmmdd) here is an example how to remove all live branches from previous year (Given it’s 2016)

  • cd to the repo of your project you want to cleanup
  • create the cleanup.rb with content bellow
  • lunch ruby cleanup.rb
# cleanup.rb
old_live_branches = `git fetch origin && git branch -r | grep live-2015`  # all branches `live-2015*`
old_live_branches
  .split("\n")
  .map(&:strip)
  .map { |i| i.gsub("/", ' :') }
  .each do |destroy|
    # e.g.: git push origin :live-20151129
    puts `git push #{destroy}`
  end

you can do the same with your local git (on your computer)

old_live_branches = ` git branch | grep live-2015`  # all local branches
`live-2015*`
old_live_branches
  .split("\n")
  .map(&:strip)
  .each do |destroy|
    # e.g.: git branch -D live-20151129
    puts `git branch -D #{destroy}`
  end

source of info:

Free up space on your Linux server

Delete downloaded packages (.deb)

E.g.: already installed (and no longer needed)

sudo apt-get clean

Remove stored archives in your cache

E.g.: packages that can not be downloaded anymore, packages are no longer in the repository or that have a newer version in the repository

sudo apt-get autoclean

Remove packages after uninstalling an application

sudo apt-get autoremove

Remove old unused kernels

list all your kernels (installed and deinstalled) :

dpkg --get-selections | grep linux-image

your currently used kernel

uname -r

to remove particular kernel:

sudo apt-get remove --purge linux-image-X.X.XX-XX-generic

You can also run this script that will remove all unnecessary kernels, Be really really carefull with this !! server.

#!/bin/sh
dpkg -l linux-*  | \
awk '/^ii/{ print $2}' | \
grep -v -e `uname -r | cut -f1,2 -d"-"` | \
grep  -e '[0-9]' | xargs sudo apt-get -y purge
  • dpkg -l linux-* list all kernels
  • uname -r will tell you current kernel

Source of information:

When stuff goes wrong

clean up /boot partition

when you install kernel and you get error similar to this one:

update-initramfs: Generating /boot/initrd.img-3.13.0-62-generic

gzip: stdout: No space left on device
E: mkinitramfs failure cpio 141 gzip 1
update-initramfs: failed for /boot/initrd.img-3.13.0-62-generic with 1.
run-parts: /etc/kernel/postinst.d/initramfs-tools exited with return
code 1
dpkg: error processing package linux-image-extra-3.13.0-62-generic
(--configure):
 subprocess installed post-installation script returned error exit
status 1
No apport report written because MaxReports has already been reached
                                                                    Processing
triggers for initramfs-tools (0.103ubuntu4.2) ...
update-initramfs: Generating /boot/initrd.img-3.13.0-62-generic



gzip: stdout: No space left on device

it may be you run out of space on boot partition

df /boot      # 100%
ls /boot

-rw-r--r--  1 root root  1169023 May 26 20:18 abi-3.13.0-54-generic
-rw-r--r--  1 root root  1169023 Jun 18 01:14 abi-3.13.0-55-generic
-rw-r--r--  1 root root  1169201 Jun 19 10:30 abi-3.13.0-57-generic
-rw-r--r--  1 root root  1169346 Jul  8 04:00 abi-3.13.0-58-generic
-rw-r--r--  1 root root  1169346 Jul 24 23:11 abi-3.13.0-59-generic
-rw-r--r--  1 root root  1169346 Jul 29 12:40 abi-3.13.0-61-generic
-rw-r--r--  1 root root  1169478 Aug 11 15:51 abi-3.13.0-62-generic
-rw-r--r--  1 root root  1169421 Aug 14 22:58 abi-3.13.0-63-generic
-rw-r--r--  1 root root   169832 May 26 20:18 config-3.13.0-54-generic
-rw-r--r--  1 root root   169832 Jun 18 01:14 config-3.13.0-55-generic
-rw-r--r--  1 root root   169832 Jun 19 10:30 config-3.13.0-57-generic
-rw-r--r--  1 root root   169832 Jul  8 04:00 config-3.13.0-58-generic
-rw-r--r--  1 root root   169832 Jul 24 23:11 config-3.13.0-59-generic
-rw-r--r--  1 root root   169833 Jul 29 12:40 config-3.13.0-61-generic
-rw-r--r--  1 root root   169833 Aug 11 15:51 config-3.13.0-62-generic
-rw-r--r--  1 root root   169833 Aug 14 22:58 config-3.13.0-63-generic

Remove the old kernels like this:

sudo dpkg --purge linux-image-3.13.0-53-generic
sudo dpkg --purge linux-image-3.13.0-54-generic
# ...
sudo apt-get -f install # tell to continue installing the latest kernel
sudo apt-get autoremove

Discussion