Redirecting Output and Errors (>, >>, 2>)

Control where output and errors go.

When working with command-line interfaces like Bash or other Unix shells, redirecting the output and errors of commands is a crucial skill. By mastering these redirection operators, you can capture command results, save logs, and troubleshoot problems more effectively. In this article, we'll peel back the layers of output redirection with >, >>, 2>, and related operators.

Understanding Standard Output and Error

When a command is executed in Unix-like systems, it communicates with you through file streams:

  • Standard Output (stdout): The normal output, typically sent to your screen.
  • Standard Error (stderr): Where error messages are sent. This is also usually shown on your screen, but it’s separate from the standard output.

Both of these are just numbered file descriptors:

  • 1 for stdout
  • 2 for stderr

Basic Output Redirection: >

The single greater-than sign > redirects the standard output (stdout) to a file, overwriting its contents if the file already exists.

echo "Hello, World!" > hello.txt
  • Effect: hello.txt will contain Hello, World!. If it existed before, its contents are replaced.

Examples

ls > files.txt
  • Outputs the result of ls into files.txt.
cat missing-file > out.txt
  • Since missing-file doesn't exist, nothing is written to out.txt (stderr is not redirected).

Appending Output: >>

To add output to the end of a file without overwriting its existing contents, use >>.

echo "Another Line" >> hello.txt
  • Effect: Adds "Another Line" to hello.txt, after existing contents.

Example

date >> log.txt
  • Appends the current date/time to log.txt every time you run the command.

Redirecting Errors: 2>

The 2> syntax redirects standard error (stderr) to a file, leaving standard output unaffected.

ls missing-file 2> error.log
  • Effect: The error message from trying to list a non-existent file is saved in error.log.

More Examples

cat missing-file 2> cat-error.log
  • Any error message produced will be stored in cat-error.log.

Combining Output and Errors

Redirecting Both to the Same File

Sometimes you want both standard output and errors together. You can achieve this using &> or by redirecting both separately:

command &> all-output.log

or

command > all-output.log 2>&1
  • &> redirects both stdout and stderr to the file (Bash).
  • > all-output.log 2>&1 does the same in a more POSIX-compliant way.

What does 2>&1 mean?

  • 2>&1 tells the shell to redirect stderr (2) to where stdout (1) is currently going.

Example

ls existing-file missing-file > results.log 2>&1
  • Both the output and any errors will be collected in results.log.

Capturing Output and Errors Separately

You can even capture output and errors to different files:

some-command > out.log 2> err.log
  • Standard output goes to out.log
  • Standard error goes to err.log

Practical Use Cases

  • Logging: Save command outputs and errors for later analysis.
  • Troubleshooting: Isolate error messages from normal output.
  • Automation: Capture what scripts do and what goes wrong during execution.

Recap Table

Symbol Function Example usage
> Overwrites file with stdout ls > file.txt
>> Appends stdout to file echo test >> file.txt
2> Redirects stderr to file ls /fake 2> err.txt
2>> Appends stderr to file ls /fake 2>> err.txt
&> Redirects both stdout and stderr ls /fake &> output.txt
> file 2>&1 Redirects both stdout and stderr ls /fake > output.txt 2>&1

Conclusion

Redirecting output and errors is fundamental for effective command-line usage. Whether for debugging, logging, or just keeping your terminal clean, knowing how to control where your command output and errors go empowers you to be a more efficient and effective user.

Now that you've unlocked the power of redirection, try combining these operators with your favorite commands and notice the difference!

Back to All Articles