Errors vs Output: 2>

Linux separates command output into two streams:

• stdout (stream 1): normal output (what you usually see).
• stderr (stream 2): error messages.

When you use >, only stdout is redirected. Errors still appear on screen:

ls real.txt fake.txt > output.txt

ls: cannot access 'fake.txt': No such file or directory

The listing of real.txt went into output.txt, but the error stayed on screen.

Redirecting errors with 2>

ls real.txt fake.txt 2> errors.txt

Now errors go into errors.txt, and normal output prints to screen.

Discarding errors with /dev/null

/dev/null is a black hole: anything sent there disappears:

ls real.txt fake.txt 2> /dev/null

Only the success output remains; errors vanish silently.

Combining both streams with 2>&1

ls real.txt fake.txt > all.txt 2>&1

This sends both stdout and stderr into all.txt. The 2>&1 means "send stream 2 wherever stream 1 is going."

Mission: Ghost File Investigation

Station sensors flagged a reference to a file called ghost.txt that may no longer exist. You need to check both files, but the error output is cluttering the main console and you want it filed away.

1. Run ls report.txt ghost.txt to check both files, but capture only the error into a file called errors.log. The normal output should stay on screen.
2. Display errors.log to read the captured error.
3. Create a folder called logs and move errors.log into it.
4. Search the captured log for the words No such to confirm the failure was recorded.

Expected result

logs/errors.log contains the "No such file" error for ghost.txt.