Configuring sudo

sudo stands for “super-user do”. Sudo allows any non-root user to run applications as root. When sudo is configured correctly, it greatly increases the security of the environment.

  • Slowing hackers down. Since the root login will most likely be disabled and your users are properly granted sudo, any attacker will not know which account to go after, thus slowing them down. If they are slowed down enough, they may stop the attack and give up.

  • Allow non-privileged users to perform privileged tasks by entering their own passwords.

  • Keeps in line with the principle of least privilege by allowing administrators to assign certain users full privileges, while assigning other users only the privileges they need to complete their daily tasks.

Adding users to the sudo group

On Ubuntu 18.04, unless otherwise specified upon account creation, the user is automatically added to the sudo group.

To add a user who is not part of the sudo group already, add him/her with:

usermod -aG sudo <username>

Using the -a option helps username retain any previously existing groups.

To directly add a user to the sudo group upon username account creation:

useradd -G sudo <username>

By default, Ubuntu allows sudo users to execute any program as root with their password. There are a few ways we can check this information. The first way is as user with sudo -l.

Another way to view this information is with visudo. This opens the sudo policy file. The sudo policy file is stored in /etc/sudoers.

This gives the same information as sudo -l but with one difference; the %sudo indicates that it is for the group, sudo. There can be other groups in this file such as admin. This is where administrators can set what programs a user in a certain group can perform and whether they need a password.

In the %sudo ALL=(ALL:ALL) ALL NOPASSWD: ALL, the NOPASSWD part says that the user that is part of the sudo group does not need to enter their local password to use sudo privileges. Generally, this is not recommended - not even for home use.

Users can also be added to the sudoers using the sudo policy file. Always edit sudoers with the sudo visudo command.

If you are managing users in a network across multiple flavours of Linux (CentOS, Red Hat, etc.), where the sudo group may be called something different, this method may be more preferable: add a User Alias to the policy file and add users to that alias, or add lines for individual users.

User_Alias ADMINS = lela, barzh



or for an individual user:

username ALL=(ALL) ALL

It is not recommended to use individual user aliases in a large network since this can become unwieldy very quickly. The first option is going to be your best bet as you can simply add users to this alias and control which commands they have access to with sudo easily.

Setting up sudo for only certain delegated privileges

To ensure that users are assigned to the groups they belong to and only are allowed access to the programs they need to complete their daily tasks, set Command Aliases in the sudo policy file: Use sudo visudo, and edit the line under the comment # Cmnd alias specification.

Cmnd_Alias SOFTWARE = /bin/rpm, /usr/bin/up2date, /usr/bin/yum

To assign the SOFTWARE command alias to the SOFTWAREADMINS user alias:


You can also assign Command Aliases to individual users, specific commands to individual users, and Command Aliases to groups.

Host aliases

Host Aliases are a way to trickle down a sudo policy across the network and different servers. For example, you may have a MAILSERVERS Host Alias which contains servers mail1 and mail2. This Host Alias has certain users or groups assigned to it, and that Host Alias has a Command Alias assigned to it stating which commands those users are able to run.

When those users run a command on mail1 or mail2, the server will check the sudo policy file to see if they can do what they are trying to do.

In a home environment and small-medium business environments, it probably is just easier to copy the sudo policy file to each server in the network. This will really only come into play with large enterprise networks and even then they will probably be using a centralised Ansible or other automation in effect.

Preventing users from using shell escapes

Programs like text editors and pagers have a handy shell escape feature. This allows a user to run a shell command without having to exit the program first. For example, from the command mode of the vi and vim editors, someone could run the ls command by running :!ls.

To allow a user to edit the sshd_config file and only that file, add a line to the sudo configuration:

username ALL=(ALL) sudoedit /etc/ssh/sshd_config

sudoedit has no shell escape feature.

Other programs that have a shell escape feature: emacs, less, view, and more.

Preventing users from using other dangerous programs

Some programs that don’t have shell escapes can still be dangerous if you give users unrestricted privileges to use them, like cat, cut, awk, and sed.

If you must give someone sudo privileges to use one of these programs, limit their use to only specific files.

Preventing abuse via user’s shell scripts

A rule for Fritz so he can create scripts running with elevated privileges:

fritz ALL=(ALL) /home/fritz/

And his script reads:

echo "This script belongs to Fritz the Cat."
sudo -i

What can happen next:

fritz@server:~$ sudo ./
This script belongs to Fritz the Cat.

sudo -i logs a person in to the root user’s shell, the same way that sudo su - does.

To remedy this, move Fritz’ script to the /usr/local/sbin directory and change the ownership to the root user so that Fritz will not be able to edit it. And delete that sudo -i line in it.