Before you start
Learn what to expect from this tutorial, and how to get the most out of it.
About this series
This four-part tutorial series covers UNIX® basics from a user
perspective. This initial tutorial is a good brush-up for users who have been away from
UNIX-like operating systems for some time. It's also useful for brand-new UNIX users coming from a Windows®
background, because it uses references and comparisons to Windows.
Later tutorials in the series will cover specific applications
vi, for instance) in detail and discuss shell tricks and tips.
About this tutorial
Systems administrators can use a number of programs to maintain
files in a UNIX system from the command line. More exist than the basic
ones demonstrated here, but they're all fundamental parts of
your UNIX system. You can use the
to reposition and rename files and
cp command duplicates one or more
files or directories. An application called
tar can quickly group files together
into a single archive. This tutorial also discusses traversing the file system,
dealing with file permissions, and simple input/output so that new UNIX
users have a base to work from.
The objective of this tutorial is to make new UNIX users comfortable moving around on the command line and dealing with files. It focuses on common command-line utilities that manipulate files, but file permissions and input/output are also discussed to give you a complete picture of commands you need to use on a day-to-day basis.
You need a basic understanding of computers and files for this tutorial, but no experience in a UNIX-like operating system is expected. You should understand what directories (folders) and files are and be able to log in to your account on a UNIX-like operating system. If you're familiar with the DOS or Microsoft® Windows command line, you'll notice similarities, but users who have never used any sort of command line will do just fine.
Access to a user account on any computer running any UNIX-like operating system is all you need to complete this tutorial. UNIX-like operating systems include the IBM AIX® operating system, Linux®, Berkeley Software Distribution (BSD), and Mac OS® X (using Terminal to access the command line), among many others.
One quick caveat before you begin: Many different UNIX-like distributions are currently available, ranging from commercial distributions, such as AIX, to free distributions, such as BSD and Linux. This tutorial focuses on commands and command-line options that are available in just about every distribution; if you find differences in your distributions, check your man pages (as described later).
Log in to the UNIX-like operating system of your choice to
get started. When you log in, you should automatically start in your
user's home directory. The examples use the
Before you work with specific commands, it's important to learn about
man stands for manual; it's a critical tool for UNIX users who
pride themselves on being self-sufficient. Type
to see an explanation of anything you want to learn about. At each
page in this tutorial, you're encouraged to check the man page along
with the instructions.
Try typing the following (note that
$ in this tutorial refers to your command prompt;
in the examples, you should type what you see after the dollar sign):
$ man ls
The first thing you need to learn is how to traverse and examine your file system. In UNIX, directories are used to organize files in a hierarchical structure. Rather than clicking directories to enter them and seeing each file as an icon, you use a series of commands and lists to view and traverse a UNIX file system from the command line.
If you're using a
UNIX-like operating system for the first time but you've used a
DOS or Windows command line in the past,
ls is roughly equivalent to
dir. It's short for List Directory.
ls can be used with a variety of
command-line options to get detailed lists, show hidden files,
recursively list subdirectories, and so on. Enter the following examples:
$ ls $ ls -l $ ls -a $ ls -R
If you're sitting on the command line thinking about your next action, a quick
ls can help you visualize where you are and what
you're doing. Think of it almost like a screen refresh in a graphical user
interface, updating you on the system's current state.
To traverse the file system, use
cd to change directories. Type
cd and then the name of the directory you
want to go to. If you use a leading /, the directory you name is an
absolute directory path: It starts from the
root of your file system. If you don't use a leading /, the directory is a
relative path: You start from your current
working directory. For example, if you're in your home
directory, (/home/tuser), and you want to move to a
cd followed by the directory name you
want to go to. For instance, if you have a Documents directory, you can
type the following:
$ cd Documents
Because you start in /home/tuser/, this command puts you in /home/tuser/Documents/.
In contrast, you can use an absolute path name to specify a location. As an example, you can move into /tmp and then back to your home directory as follows:
$ cd /tmp $ cd /home/tuser
Special directory names
In UNIX, special directory names make traversing the file system easy.
The three most important ones refer to the current directory, the current directory's parent,
and the user's home directory. The current directory is specified by a dot (or period).
For instance, if you type
$ cd ., you remain in the current directory, which is still
/home/tuser/. This character becomes especially important
when you're running executables in the current working directory. By default, many UNIX
shells search application directories for applications, but it does not search the current working
directory. You can always explicitly refer to files and applications in your current working
directory by using ./ preceding a filename. Parent directories are referred
to with two dots (or two periods). To traverse to your current working
directory's parent, type the following:
$ cd ..
If you were in /home/tuser before, now you're in /home. To demonstrate the third special directory name, use the shortcut to get back to your home directory (the tilde character). Type the following command:
$ cd ~
To check your current directory, you can use
pwd, which stands for Print Working
Directory. It tells you where you are in the file system, which
helps you determine what to use when specifying relative path
names. Try experimenting with the three special directory names,
absolute paths, and relative paths to traverse your file system. At each step,
pwd to check your current location.
Check your current directory (you should be in your home directory if you followed the steps in the previous section):
rmdir are used to create and remove
rmdir works only if the directory is
empty (it doesn't remove files).
Try the following commands:
$ mkdir TUTORIAL $ cd TUTORIAL $ pwd $ ls
You're now in your newly created TUTORIAL directory, and it's empty.
You can also get to this directory by using the
~ (tilde) character. To get to
$ cd ~/TUTORIAL $ pwd
Now that you know how to move around directories and get listings, you're ready to look at the directory layout on a typical UNIX distribution. You can organize a UNIX file system several ways. This tutorial discusses a few root-level directories that are common to most UNIX-like distributions. There are other important root-level directories, but this is where you'll find yourself operating in most cases:
/home (or /users) /etc /bin /sbin /usr /car /tmp
/home is where user directories are found. For instance, the
is located in /home/tuser.
/etc is the directory used to store most system-wide settings, including startup scripts and network configuration files. You need root access to edit most files in this directory.
/bin and /sbin are the directories
used to store system executables (like the
commands you're learning about in this tutorial).
/sbin is used for system commands, such as
/bin is used for user commands.
Applications are usually installed in /usr. A /usr/local/ subdirectory often holds applications installed that aren't part of the base distribution.
/var is the directory where things like log files, which are continually updated, are stored.
Temporary files are stored in /tmp. This directory is usually writable by all users on the system, and old files are periodically removed on some systems.
Now that you know how to move around the file system on the command line, it's time to start working with files. This section teaches you how to create an example file, copy a file, remove a file, and view and change basic file permissions. In a multiuser operating system like UNIX, it's crucial to understand ownership and permission constructs.
To begin, create an empty file
that you can use for this tutorial. The
touch command can be used to
create an empty file (it's normally used to update the modified date and accessed date of a
file by touching it).
Go back to the TUTORIAL directory in your user's home, and create a file by typing the following command:
$ cd ~/TUTORIAL $ touch example.txt
cp command copies files.
cp followed by the name of the file you
want to copy, followed by the directory you want to copy the
file to (you have the option of specifying a new filename as well). For
instance, try copying the example.txt file to
$ cp example.txt /tmp/ $ ls /tmp/
You should see example.txt in /tmp/. Now, copy the file in /tmp/ back to your current directory, but give it a new name:
$ cp /tmp/example.txt ./example2.txt $ ls
Notice the use of a dot to specify that you want to put this new file in your
current directory. It isn't necessary to include
./ in this case (because the default path for a
copy is your current working directory), but it helps clearly
illustrate what you intend to do. The subsequent
ls command shows that you now have
two example files in your current working directory.
move command is completed with
mv. Most syntax and command-line options for move and copy are the same.
If you want to move your new file, example2.txt, out of the current
directory and into /tmp/, type the following:
$ mv example2.txt /tmp/.
Note again that a dot is used to explicitly call out what you're doing.
Remove the files created in /tmp/ to tidy up your system. The
rm command deletes files from your file
system. This isn't like moving a file into the
Recycle Bin or Trash; the command deletes the file pointer, so use the
rm command with caution. Type the following:
$ rm /tmp/example.txt $ rm /tmp/example2.txt $ ls /tmp/
Both example files in /tmp/ should be gone.
Depending on which
UNIX-like operating system you're using, you might have other delete
commands available, such as
can. Try typing
man srm and
man can to see if you have these commands.
srm is used as a secure version of
rm, which writes random data over deleted
files to keep them from being restored.
can is the opposite of
srm in some ways;
can retains the file but moves it into a
special trash directory similar to the Windows Recycle Bin.
Ownership and permissions
File ownership and permissions are important concepts in UNIX-like operating systems. UNIX is built as a multiuser system from the ground up. Windows is also capable of acting as a multiuser system in today's versions, but Windows has a single-user ancestry dating back to early personal computers and DOS. Even today, some Windows users don't think much about multiple accounts and the file permission issues that go along with a multiuser system. It's critical for any UNIX user to understand file ownership and permissions.
File ownership in UNIX can be examined using
ls -l and can be changed using
chgrp. Take a look by typing the following:
$ ls -l
-l specifies that you want to see the long
format of a file listing. The long format includes information about permissions, ownership,
modification date, and file size. You should see something like this:
tsystem:~/TUTORIAL tuser$ ls -l total 0 -rw-r--r-- 1 tuser admin 0 Aug 13 15:35 example.txt
When you do a detailed listing of a directory, you can see in the third and fourth
columns the user ownership and the group ownership of a file,
respectively. This file is owned by the
tuser user and the
admin group. For now, leave this file alone;
you can learn more by typing:
$ man chown $ man chgrp
The basic syntax for this command uses the name of the user or group as the first input followed by the file or list of files you want to operate on. For example:
$ chown tuser example.txt $ chgrp admin example.txt
Basic file permissions in UNIX are handled by a set of nine flags associated with each file. These flags are formed from three different user categories (user, group, and other) and three file operations (read, write, and execute). Take a look by typing the following command:
$ ls -l
You should see something like this:
-rw-r--r-- 1 tuser admin 0 Aug 13 15:35 example.txt
When you're reading a long listing of a directory, the first column shows
the file permissions. Notice that there are ten characters (not nine) in
this column. The first character denotes the type of file you're dealing with. A dash,
-, tells you it's a regular file, for
d denotes a directory, not a
plain file. For now, concentrate on the other nine characters, each of
which is one of the following:
x. A file that has all permissions turned
off reads like this:
---------- 1 tuser admin 0 Aug 13 15:35 example.txt
A file that has all permissions turned on reads this way:
-rwxrwxrwx 1 tuser admin 0 Aug 13 15:35 example.txt
You see three sets of
rwx. As mentioned, there are three
different user categories (user, group, and other). Each of these
rwx sets represents one of the
- The first
rwxrepresents the permissions the
userset. In other words, this tells you what the owner of the file can do to it.
- The second set represents the
groupset. This tells you what members of the group can do to the file.
- The third set represents the
otherset. This tells you what all users on the system can do to a file, regardless of who owns it.
Take a closer look at the current permissions of your example.txt file:
-rw-r--r-- 1 tuser admin 0 Aug 13 15:35 example.txt
The first set of three flags indicates that the user who owns this
file can read it and write to it, but the user can't execute it (there is
r and a
w, but not an
x). You can see from the second set of
three flags that the group that owns this file can read it, but can't
write to it or execute it (there is an
r, but not a
w or an
x). The third set of three
flags shows that all other users can read it, but they can't write to it or execute
it (there is an
r, but not a
w or an
x). Here's a good example of where
man pages are important. Pause to
type the following:
$ man chmod
Now, consider the possibility of having private information in this
file that you don't want any other users to read. You probably want to
remove read access for other groups and all other users. You can use
chmod to change the permissions. Like many things in UNIX,
there are multiple ways to use
chmod; this section focuses on one.
The three categories (user, group, and other) are represented by three letters (u, g, and o).
The three types of permissions (read, write, and execute) are also represented by three
letters (r, w, and x). To change permissions, use
by the letters of the categories you want to change, followed by a plus or a minus
(for turn on or turn off), followed by the letters of the permissions you want to change. After this,
write the name of the file (or files) you want to make the change to. This is best illustrated
with an example:
$ chmod og-r example.txt $ ls -l
You should see this result:
-rw------- 1 tuser admin 0 Aug 13 15:35 example.txt
In this example, you specify other and group (o and g) and use a minus sign to indicate that you want to turn off certain permissions for these categories. Then, you used read (r) to indicate that you're specifically turning off read access. Now, the owner (tuser) can still read and write the file, but all other users on the system (except the superuser) can't access the file. Note: The superuser (root) can override all file permissions.
Dealing with multiple files
You've now learned to traverse directories and handle individual files. The next step in this tutorial is learning to deal with files in groups. Almost all UNIX commands can handle file lists as opposed to individual files. You can enter file lists by explicitly typing the name of each file you want to use or by using a wildcard to indicate that you want to use all files with a common name trait.
The most common method of dealing with multiple files is to use
*, the wildcard character.
You can select a list of files by using
to represent any character or any number of characters. To create a few
more files for a demonstration, type these commands:
$ cp example.txt example2.txt $ cp example.txt script.sh
Now, type the following:
$ ls *.txt
You should just see the files with the .txt extension. Next, type:
$ ls exa*
Again, you should see the two example files, but you should not see script.sh.
The wildcard character can be used with any command-line application that can deal with multiple files.
Many command-line applications that deal with files have the
-R option. When
-R is used, the application
recursively enters a directory and any subdirectories, and it performs the
desired command on each file. For instance, you can go back in your
home directory and copy the entire TUTORIAL directory:
$ cd ~ $ cp -R TUTORIAL /tmp/. $ ls /tmp/TUTORIAL/
Now, remove that directory to tidy up:
$ rm -R /tmp/TUTORIAL/ $ ls /tmp/
The entire directory is removed, including all the files contained
in it. Be careful: You could easily delete much more data than
you plan to, especially when combining a wildcard with
Archives and compression
Many day-to-day file operations can be completed using single files, directories, and wildcards; but learning how to use archives and compression is important for any user interested in backups, transmitting file sets to other users, or simply saving space. A number of UNIX utilities are built into the operating system for archiving and compressing data.
The most common method to group multiple files into a single file (an
archive) is to use the
tar is short for Tape Archiver, due to its
roots for use with backup tapes; but today, it's commonly used for
disk-to-disk operations. Try archiving what you've done so far in the
$ cd ~ $ tar cvf /tmp/tutorial.tar TUTORIAL $ ls /tmp/
You should now see a file called tutorial.tar in /tmp/. This file
contains both the directory TUTORIAL and the files contained in it.
This example uses the command-line options
create, which tells
tarto create a new archive.
vstands for verbose, telling
tarto show a listing of each file that goes into the archive.
fstands for file, telling
tarto write the archive to a file rather than a device.
Next, go into /tmp and extract the contents of the archive:
$ cd /tmp/ $ ls $ tar xvf tutorial.tar $ ls
Notice that the first
ls command shows
tutorial.tar, but it doesn't show a TUTORIAL directory. The second
ls command, after the
tar command, demonstrates that you now
have a TUTORIAL directory in /tmp/. Remove the
/tmp/TUTORIAL directory to tidy up:
$ rm -R /tmp/TUTORIAL
Once you have a tar file, you might want to compress it. A number of
compression options are available in most UNIX distributions, but this tutorial focuses on
gzip to create something commonly
called a tarball. A tarball is a tar file that has been
compressed. Go ahead and create a compressed version of
tutorial.tar so that it takes up less space:
$ gzip tutorial.tar $ ls
Your tutorial.tar archives are now compressed and renamed tutorial.tar.gz. This is a tarball. To uncompress it, type:
$ gzip -d tutorial.tar.gz
The result is your original uncompressed tar file, tutorial.tar.
The file system and file sizes
It's easy to learn how to deal with individual files and examine their size and content. You can use the same techniques to examine the contents of entire directories and the file system. Many new versions of UNIX can display this information in simple numbering formats, using letters to denote units.
df stands for Display Free disk space.
Using it is as simple as typing
df; you get information about the
amount of disk space in each file system on your machine, the amount
of space used, and the amount of space available. Most systems default
to display the values in 512KB blocks, which is hard to read.
-g to display information in gigabytes
-m to display information in megabytes.
Some systems have a
-h option, which stands for
human-readable. This makes
df use suffixes like G, M, and K, and it
displays each number in three or fewer digits. Type this command:
$ df -h
This is an example of output you might see on a simple server:
$ df -h Filesystem Size Used Avail Use% Mounted on /dev/sda1 7.9G 3.7G 3.9G 50% / none 3.9G 0 3.9G 0% /dev/shm /dev/sda3 24G 20G 1.9G 92% /export
If your system has
df, you can also use it with
ls. Type this command to see a detailed
listing with easy to read file sizes:
$ ls -lh
du is a third way to check file sizes, but it
has the added advantage of summing up directory sizes. It can also be
-h on some systems; otherwise, try using
-k, which gives results in 1024-byte
blocks. You can also use
-s and a filename or wildcard to specify
which directories and files you want to examine. Try this:
$ cd ~ $ du -sk * $ du -sh *
This is an example of output you might see in a home directory:
$ du -sk * 316468 OLD 637940 MyData1 571788 Code 12356364 Build 3224480 Hardened $ du -sh * 310M OLD 623M MyData1 559M Code 12G Build 3.1G Hardened
/dev directory holds special files called device files,
which, among other things, are used to access disk drives on your system.
To learn about the /dev directory, take another look
at the output of
df. This is different on every machine,
but it's important to note that
df shows results for each
file system on your computer. Unlike Windows-based computers, each mounted file system
is addressed from the system's root, which is denoted by a forward slash:
/. This is different from systems that separate disks
with drive letters, such as C, D, E, and so on.
In UNIX, it's common for SCSI (and SATA) disks to use device names, such as /dev/sda, /dev/sdb, /dev/sdc, and so on. A common device name for a CD-ROM drive is /dev/cdrom. These devices are mounted to directories so that they can be accessed without using the device name. Please consult the documentation for your flavor of UNIX to find out how devices on your system are labeled.
Any device can be mounted at any location
(any directory). For instance, it's common to mount a CD-ROM at
/mnt/cdrom. Some UNIX-like operating systems (such as
many versions of Linux and Mac OS) mount CD-ROMs automatically, but it's good
to learn how to use the
mount command either way.
Insert a CD-ROM, and type the following commands:
$ mount -t iso9660 /dev/cdrom /mnt/cdrom $ df $ ls /mnt/cdrom
Note: This will work only if /dev/cdrom and /mnt/cdrom exist on your system. If so, you see
in the output of the
df command that the CD-ROM is now
part of the file system. The
ls command should show
the contents of the CD-ROM drive you just mounted.
To unmount a device, most UNIX-like operating systems use
umount. The syntax is
followed by the mount point. If your previous
succeeded, type the following:
$ umount /mnt/cdrom $ df $ ls /mnt/cdrom
Note: You must not be in the mounted file system for the device to unmount
properly; otherwise, the system complains that the file system is busy. After
a proper execution of
command no longer shows the CD-ROM drive in the file system, and the
ls command shows that /mnt/cdrom
is now empty (because nothing is mounted there -- it's a normal directory).
Input and output
Input and output are often thought of in simple default terms:
keyboard/mouse and monitor/speakers. In UNIX, access to a system of input
and output streams (and an error stream) gives users and developers a way
to efficiently move input and output to and from applications, which
can streamline complex processes by minimizing human interaction.
stderr are significant parts of what make UNIX
an ideal platform for scripting.
Most command-line applications can take input from
stdin and direct output to
stdin stands for Standard Input.
stdout stands for Standard Output. By
default, input from
stdin comes from the keyboard (what
you type in your terminal), and output to
stdout goes to your display (what
you see on your screen). An additional type of output,
stderr, prints errors and
debugs messages, but this tutorial focuses on
Redirection allows users to send output that would normally go to
stdout to another destination -- a file,
for instance. Create a text file of your TUTORIAL directory listing, as follows:
$ cd ~/TUTORIAL $ ls > listing.txt $ ls
Another form of redirection is
>>, which appends a file as
opposed to creating a new file. You can also redirect
stderr to a file with
2> or to redirect all output (
stderr) to a file with
&>. You can combine forms of
redirection; for instance, use
2>> to append
stderr to a file.
Now that you have a file with some text in it, look at what's inside.
The quickest way to examine the contents of a file is by using the
cat command, which stands for
Concatenate; it can be used in combination with redirection to
concatenate text. Type the following command:
$ cat listing.txt
If you've executed each step in the tutorial, you should see something like this:
example.txt example2.txt listing.txt script.sh
cat command was executed on a
relatively small file, so you can easily see all the data on a single
page. If you're viewing a file that doesn't fit on a single page, the
more command is commonly used;
it pauses the output each time a full page of data is displayed to
stdout. Pressing the
Spacebar advances the output to the
next page. Try creating a longer file to use for an example:
$ ls /etc/ > listing2.txt $ cat listing2.txt $ more listing2.txt
If you use
cat, the file scrolls much too quickly to
read; but if you use
more, you can press
the Spacebar to advance through the output step by step.
head and tail
If you quickly want to view the first or last few lines of a file, you can use
tail. These commands are
commonly used to view the top of script or the bottom of a log file. It's also handy
to use them as a quick sanity check on an output file when you're debugging code. Try typing the following:
$ head listing2.txt
To view the last few lines of a file instead of the first few lines, try testing
$ tail listing2.txt
Both commands commonly default to displaying 10 lines, but you can use the
-n option to display any
number of lines. For instance, type this command:
$ head -n 2 listing2.txt
Now that you're starting to create files with more data in them, you might
want to search for specific text.
grep is a powerful search utility
used here in a bare-bones example:
$ grep host listing2.txt
This command outputs all the lines in listing2.txt that contain the
To skip the file-creation step, you can use the
|) to use the output of one command as the
input for another command. This is another form of redirection and is
incredibly powerful for linking long lists of commands to
efficiently manage input and output. Try this simple example:
$ ls /etc/ | grep host
This command gives the same output as the two-step process listed previously. The
single line takes the output of the
ls command and uses it as the search input
grep command. Here's another example:
$ du -sh /etc/* | more
In this case, you check the disk usage (
du) for each file and directory in
/etc/. The output is more than one page, so it's useful to
pipe the result to
more, which makes the output pause
each time one page of data is displayed.
After running through this tutorial, you should have the basic knowledge to navigate a UNIX file system from the command line and work with files and directories. You should experiment with the commands listed throughout this text and closely examine the man page for any item you wish to learn more about.
Future tutorials in the series will cover specific applications (
vi, for instance) in detail and cover
shell tricks and tips. In the meantime, feel free to experiment with a
test user; in short order, you'll likely find yourself whizzing
through command-line options faster than you could imagine pointing
and clicking with your mouse!
- UNIX tips and tricks for a new user: Check out other parts in this series.
- The Open Group: This site can teach you more about UNIX and UNIX certifications.
- Linux.org: Learn more about Linux, an open source UNIX-like operating system.
- Make UNIX work with Windows XP and Mac OS X (developerWorks, April 2006): Integrate a UNIX system into a cross-platform environment.
- FreeBSD.org: Visit this site to learn more about BSD operating systems.
- Linux Journal: Keep track of Linux developments.
- AIX and UNIX: Visit the developerWorks AIX and UNIX zone to expand your UNIX skills.
- New to AIX and UNIX: Visit the New to AIX and UNIX page to learn more about AIX and UNIX.
- developerWorks technical events and webcasts: Stay current with developerWorks technical events and webcasts.
- AIX 5L Wiki: A collaborative environment for technical information related to AIX.
- Podcasts: Tune in and catch up with IBM technical experts.
Get products and technologies
- IBM trial software: Build your next development project with software for download directly from developerWorks.
- Participate in the AIX and UNIX forums:
- Participate in the developerWorks blogs and get involved in the developerWorks community.