What is the use of 2>&1 in shell scripting

Question

In shell scripting, 1 is file descriptor for standard output and 2 is for standard error. Now, since both of these point to console what is the use of 2>&1. Are not they already pointing to the same device?

Answer 1

The short answer is you are redirecting stderr to stdout so you get both error messages written to FD2 as well as normal output on FD1 written to FD1. (FD = File Descriptor). It generally allows you to capture the output of error messages you want to save in a log file, etc.. that would otherwise not be captured simply by redirecting stdout to the log.

By way of brief background, your shell has 3 well known file descriptors that cover basic reading and writing:

0 - stdin (your input buffer)

1 - stdout (your normal output descriptor)

2 - stderr (your normal error descriptor)

When you read or write, just about any programming language, including your shell, makes use of these common file descriptors to manage the input/output. In your shell, you have the ability to redirect or combine output from stdout and stderr. The general format is:

m > &n    ## where n & m are 1, 2

You can combine, for instance, both the stdout and stderr from any command to a file with:

./someprogram > myfile 2>&1

Which basically says, "take the output from someprogram on stdout redirect it to myfile while including stderr in stdout.

There are several good references, one being BASH Programming - Introduction HOW-TO: All about redirection

Note: Be sure you also understand that redirection can happen on a temporary (per command basis) or can be set for the scope of the script using exec.