Privilege separation is one of the fundamental security paradigms implemented in Linux and Unix-like operating systems. Regular users operate with limited privileges in order to reduce the scope of their influence to their own environment, and not the wider operating system.
A special user, called
root, has “super-user” privileges. This is an administrative account without the restrictions that are present on normal users. Users can execute commands with “super-user” or “root” privileges in a number of different ways.
In this article, we will discuss how to correctly and securely obtain
root privileges, with a special focus on editing the
We will be completing these steps on an Ubuntu 16.04 server, but most modern Linux distributions should operate in a similar manner.
This guide assumes that you have already completed the initial server setup discussed here. Log into your server as regular, non-root user and continue below.
There are three basic ways to obtain
root privileges, which vary in their level of sophistication.
The simplest and most straight forward method of obtaining
root privileges is simply to log into your server as the
root user from the onset.
If you are logging into a local machine (or using out-of-band Console Access if you are using DigitalOcean), simply enter “root” as your username at the login prompt and enter the
root password when asked.
If you are logging in through SSH, specify the
root user prior to the IP address or domain name in your SSH connection string:
If you have not set up SSH keys for the
root user, enter the
root password when prompted.
Logging in as
root is usually not recommended, because it is easy to begin using the system for non-administrative tasks, which is dangerous.
The next way to gain super-user privileges allows you to become the
root user at any time, as you need it.
We can do this by invoking the
su command, which stands for “substitute user”. To gain
root privileges, simply type:
You will be prompted for the
root user’s password, after which, you will be dropped into a
root shell session.
When you have finished the tasks which require
root privileges, return to your normal shell by typing:
The final, and most complex, way of obtaining
root privileges that we will discuss is with the
sudo command allows you to execute one-off commands with
root privileges, without the need to spawn a new shell. It is executed like this:
sudo command will request the password of the user calling the command, not the
Because of its security implications,
sudo access is not granted to users by default, and must be set up before it functions correctly. If you followed the initial server setup guide, you already completed a bare-bones configuration.
In the following section, we will discuss how to modify the configuration in greater detail.
sudo command is configured through a file located at
Warning:Never edit this file with a normal text editor! Always use the
visudo command instead!
Because improper syntax in the
/etc/sudoers file can leave you with a system where it is impossible to obtain elevated privileges, it is important to use the
visudo command to edit the file.
visudo command opens a text editor like normal, but it validates the syntax of the file upon saving. This prevents configuration errors from blocking
sudo operations, which may be your only way of obtaining
visudo opens the
/etc/sudoers file with the
vi text editor. Ubuntu, however, has configured
visudo to use the
nano text editor instead.
If you would like to change it back to
vi, issue the following command:
sudo update-alternatives --config editor
OutputThere are 4 choices for the alternative editor (providing /usr/bin/editor). Selection Path Priority Status ------------------------------------------------------------ * 0 /bin/nano 40 auto mode 1 /bin/ed -100 manual mode 2 /bin/nano 40 manual mode 3 /usr/bin/vim.basic 30 manual mode 4 /usr/bin/vim.tiny 10 manual mode Press <enter> to keep the current choice[*], or type selection number:
Select the number that corresponds with the choice you would like to make.
On CentOS, you can change this value by adding the following line to your
export EDITOR=`which name_of_editor`
Source the file to implement the changes:
After you have configured
visudo, execute the command to access the
You will be presented with the
/etc/sudoers file in your selected text editor.
I have copied and pasted the file from Ubuntu 16.04, with comments removed. The CentOS
/etc/sudoers file has many more lines, some of which we will not discuss in this guide./etc/sudoers
Defaults env_reset Defaults mail_badpass Defaults secure_path="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin" root ALL=(ALL:ALL) ALL %admin ALL=(ALL) ALL %sudo ALL=(ALL:ALL) ALL #includedir /etc/sudoers.d
Let’s take a look at what these lines do.
The first line, “Defaults env_reset”, resets the terminal environment to remove any user variables. This is a safety measure used to clear potentially harmful environmental variables from the
The second line,
Defaults mail_badpass, tells the system to mail notices of bad
sudo password attempts to the configured
mailto user. By default, this is the
The third line, which begins with “Defaults secure_path=…”, specifies the
PATH (the places in the filesystem the operating system will look for applications) that will be used for
sudo operations. This prevents using user paths which may be harmful.
The fourth line, , which dictates the
sudo privileges, is different from the preceding lines. Let’s take a look at what the different fields mean:
root ALL=(ALL:ALL) ALL
demo ALL=(ALL:ALL) ALL
demo ALL=(ALL:ALL) ALL
rootuser can run commands as all users.
demo ALL=(ALL:ALL) ALL
rootuser can run commands as all groups.
demo ALL=(ALL:ALL) ALL
This means that our
root user can run any command using
sudo, as long as they provide their password.
The next two lines are similar to the user privilege lines, but they specify
sudo rules for groups.
Names beginning with a “%” indicate group names.
Here, we see the “admin” group can execute any command as any user on any host. Similarly, the
sudogroup can has the same privileges, but can execute as any group as well.
The last line might look like a comment at first glance:/etc/sudoers
. . . #includedir /etc/sudoers.d
It does begin with a
#, which usually indicates a comment. However, this line actually indicates that files within the
/etc/sudoers.d directory will be sourced and applied as well.
Files within that directory follow the same rules as the
/etc/sudoers file itself. Any file that does not end in
~ and that does not have a
. in it will be read and applied to the
This is mainly meant for applications to alter
sudo privileges upon installation. Putting all of the associated rules within a single file in the
/etc/sudoers.d directory can make it easy to see which privileges are associated with which accounts and to reverse credentials easily without having to try to manipulate the
/etc/sudoers file directly.
As with the
/etc/sudoers file itself, you should always edit files within the
/etc/sudoers.d directory with
visudo. The syntax for editing these files would be:
sudo visudo -f /etc/sudoers.d/file_to_edit
The most common operation that users want to accomplish when managing
sudo permissions is to grant a new user general
sudo access. This is useful if you want to give an account full administrative access to the system.
The easiest way of doing this on a system set up with a general purpose administration group, like the Ubuntu system in this guide, is actually to just add the user in question to that group.
For example, on Ubuntu 16.04, the
sudo group has full admin privileges. We can grant a user these same privileges by adding them to the group like this:
sudo usermod -aG sudo username
gpasswd command can also be used:
sudo gpasswd -a username sudo
These will both accomplish the same thing.
On CentOS, this is usually the
wheel group instead of the
sudo usermod -aG wheel username
sudo gpasswd -a username wheel
On CentOS, if adding the user to the group does not work immediately, you may have to edit the
/etc/sudoers file to uncomment the group name:
. . . %wheel ALL=(ALL) ALL . . .
Now that we have gotten familiar with the general syntax of the file, let’s create some new rules.
sudoers file can be organized more easily by grouping things with various kinds of “aliases”.
For instance, we can create three different groups of users, with overlapping membership:/etc/sudoers
. . . User_Alias GROUPONE = abby, brent, carl User_Alias GROUPTWO = brent, doris, eric, User_Alias GROUPTHREE = doris, felicia, grant . . .
Group names must start with a capital letter. We can then allow members of
GROUPTWO to update the
aptdatabase by creating a rule like this:/etc/sudoers
. . . GROUPTWO ALL = /usr/bin/apt-get update . . .
If we do not specify a user/group to run as, as above,
sudo defaults to the
We can allow members of
GROUPTHREE to shutdown and reboot the machine by creating a “command alias” and using that in a rule for
. . . Cmnd_Alias POWER = /sbin/shutdown, /sbin/halt, /sbin/reboot, /sbin/restart GROUPTHREE ALL = POWER . . .
We create a command alias called
POWER that contains commands to power off and reboot the machine. We then allow the members of
GROUPTHREE to execute these commands.
We can also create “Run as” aliases, which can replace the portion of the rule that specifies the user to execute the command as:/etc/sudoers
. . . Runas_Alias WEB = www-data, apache GROUPONE ALL = (WEB) ALL . . .
This will allow anyone who is a member of
GROUPONE to execute commands as the
www-data user or the
Just keep in mind that later rules will override earlier rules when there is a conflict between the two.
There are a number of ways that you can achieve more control over how
sudo reacts to a call.
updatedb command associated with the
mlocate package is relatively harmless on a single-user system. If we want to allow users to execute it with
root privileges without having to type a password, we can make a rule like this:/etc/sudoers
. . . GROUPONE ALL = NOPASSWD: /usr/bin/updatedb . . .
NOPASSWD is a “tag” that means no password will be requested. It has a companion command called
PASSWD, which is the default behavior. A tag is relevant for the rest of the rule unless overruled by its “twin” tag later down the line.
For instance, we can have a line like this:/etc/sudoers
. . . GROUPTWO ALL = NOPASSWD: /usr/bin/updatedb, PASSWD: /bin/kill . . .
Another helpful tag is
NOEXEC, which can be used to prevent some dangerous behavior in certain programs.
For example, some programs, like “less”, can spawn other commands by typing this from within their interface:
This basically executes any command the user gives it with the same permissions that “less” is running under, which can be quite dangerous.
To restrict this, we could use a line like this:/etc/sudoers
. . . username ALL = NOEXEC: /usr/bin/less . . .
There are a few more pieces of information that may be useful when dealing with
If you specified a user or group to “run as” in the configuration file, you can execute commands as those users by using the “-u” and “-g” flags, respectively:
sudo -u run_as_user command sudo -g run_as_group command
For convenience, by default,
sudo will save your authentication details for a certain amount of time in one terminal. This means you won’t have to type your password in again until that timer runs out.
For security purposes, if you wish to clear this timer when you are done running administrative commands, you can run:
If, on the other hand, you want to “prime” the
sudo command so that you won’t be prompted later, or to renew your
sudo lease, you can always type:
You will be prompted for your password, which will be cached for later
sudo uses until the
sudo time frame expires.
If you are simply wondering what kind of privileges are defined for your username, you can type:
This will list all of the rules in the
/etc/sudoers file that apply to your user. This gives you a good idea of what you will or will not be allowed to do with
sudo as any user.
There are many times when you will execute a command and it will fail because you forgot to preface it with
sudo. To avoid having to re-type the command, you can take advantage of a bash functionality that means “repeat last command”:
The double exclamation point will repeat the last command. We preceded it with
sudo to quickly change the unprivileged command to a privileged command.
For some fun, you can add the following line to your
/etc/sudoers file with
. . . Defaults insults . . .
This will cause
sudo to return a silly insult when a user types in an incorrect password for
sudo. We can use
sudo -k to clear the previous
sudo cached password to try it out:
sudo -k sudo ls
Output[sudo] password for demo: # enter an incorrect password here to see the results Your mind just hasn't been the same since the electro-shock, has it?
password for demo: My mind is going. I can feel it.
You should now have a basic understanding of how to read and modify the
sudoers file, and a grasp on the various methods that you can use to obtain
Remember, super-user privileges are not given to regular users for a reason. It is essential that you understand what each command does that you execute with
root privileges. Do not take the responsibility lightly. Learn the best way to use these tools for your use-case, and lock down any functionality that is not needed.