The terminal, the shell prompt and getting help

Log in to rocky01 with ssh -p 2211 student@127.0.0.1 and keep a second window open on ubu01 (ssh -p 2221 student@127.0.0.1). Nothing in this lab changes the system except one line you add to your own ~/.bashrc in step 7.

1Read your prompt

Run this on rocky01:

whoami
hostname
pwd
echo "$PS1"

You should see something like this:

student
rocky01
/home/student
[\u@\h \W]\$

Each part of the prompt has a command that tells you the same thing. The last line is the recipe for the prompt itself: \u user, \h host, \W the last part of the current folder and \$, which shows $ for normal users and # for root. On ubu01 the recipe is longer because it adds colours, but it ends in \u@\h:\w\$.

Now become root for a moment, and come back:

sudo -i
whoami
exit

You should see something like this:

[root@rocky01 ~]# whoami
root
[root@rocky01 ~]# exit
logout
[student@rocky01 ~]$

The $ turned into # and the user changed to root. That one character is your warning sign: as root, nothing stops you from deleting the wrong thing. Leave root as soon as the task is done.

✅ Checkpoint: You can point to the user, host, folder and $ or # in your prompt on both VMs.

2Find out what kind of command it is

Run this on rocky01:

type ls cd ll history ssh

You should see something like this:

ls is aliased to `ls --color=auto'
cd is a shell builtin
ll is aliased to `ls -l --color=auto'
history is a shell builtin
ssh is /usr/bin/ssh

Rocky sets the ls and ll aliases in /etc/profile.d, so when you type ls you get colours you didn’t ask for. cd and history are builtins inside bash, and ssh is an ordinary program file. On ubu01 the aliases come from ~/.bashrc instead, and ll means ls -alF.

✅ Checkpoint: You know which of the five names are aliases, builtins and program files on rocky01.

3Get quick help from the command itself

Run this on rocky01:

ls --help | head -n 4
help cd | head -n 2

You should see something like this:

Usage: ls [OPTION]... [FILE]...
List information about the FILEs (the current directory by default).
Sort entries alphabetically if none of -cftuvSUX nor --sort is specified.

Mandatory arguments to long options are mandatory for short options too.
cd: cd [-L|[-P [-e]] [-@]] [dir]
    Change the shell working directory.

Nearly every program accepts --help and prints a summary of its options. Piping it into head or less keeps it from scrolling away. For builtins, help is the built-in manual.

Now try the man page for cd on ubu01:

Run this on ubu01:

man cd

You’ll get this:

No manual entry for cd

cd is part of bash, so Ubuntu has no separate page for it (on rocky01 the same command opens the long page that covers all bash builtins). When man says “No manual entry”, run type first: if it says builtin, use help.

✅ Checkpoint: ls --help and help cd both print a usage line, and you know why man cd fails on ubu01.

4Read a man page and pick a section

Run this on ubu01:

whatis passwd crontab
man 5 passwd

You should see something like this (whatis part):

passwd (1)           - change user password
passwd (5)           - the password file
crontab (1)          - maintain crontab files for individual users (Vixie Cron)
crontab (5)          - tables for driving cron

whatis prints the one-line description of every page with that exact name, so you can see that passwd and crontab each have a command page (1) and a file format page (5). man 5 passwd opens the second one in a pager. Inside it, use these keys:

Every man page has the same layout: NAME, SYNOPSIS, DESCRIPTION, OPTIONS, then often FILES, EXAMPLES and SEE ALSO. When you’re in a hurry, search for EXAMPLES first, because many pages put ready-made commands there.

✅ Checkpoint: You found the seven fields of /etc/passwd in man 5 passwd and quit with q.

5Search when you don’t know the command

Say you need to work with disk partitions but can’t remember any tool names. apropos searches the NAME line of every man page, and -s 8 keeps only admin commands.

Run this on ubu01:

apropos -s 8 partition

You should see something like this:

addpart (8)          - tell the kernel about the existence of a partition
cfdisk (8)           - display or manipulate a disk partition table
delpart (8)          - tell the kernel to forget about a partition
fdisk (8)            - manipulate disk partition table
gdisk (8)            - Interactive GUID partition table (GPT) manipulator
parted (8)           - a partition manipulation program
partprobe (8)        - inform the OS of partition table changes
partx (8)            - tell the kernel about the presence and numbering of on...
resizepart (8)       - tell the kernel about the new size of a partition
sfdisk (8)           - display or manipulate a disk partition table

man -k partition does the same search. apropos reads a small database that the man-db package keeps up to date when packages are installed. If it ever says “nothing appropriate” for a word that should clearly match, rebuild that database with sudo mandb. Try the same search on rocky01 too: the list is shorter because a minimal Rocky install has fewer disk tools.

✅ Checkpoint: apropos gives you at least five partition tools on ubu01.

6Look in /usr/share/doc

Run this on ubu01:

ls /usr/share/doc | wc -l
ls /usr/share/doc/bash

You should see something like this (second command):

COMPAT.gz  INTRO.gz  NEWS.gz  POSIX.gz  RBASH  README.Debian.gz  README.gz  changelog.Debian.gz  copyright  inputrc.arrows

Most packages drop extra files here: README notes, changelogs, licences and, for many servers, sample config files. On Ubuntu many are gzip-compressed, so read them with zless /usr/share/doc/bash/README.gz. On rocky01 the files are usually plain text, for example ls /usr/share/doc/openssh.

When you set up a service you’ve never used, the sample config here is often the best place to start.

✅ Checkpoint: You opened one README from /usr/share/doc with less or zless and quit with q.

7Reuse commands with history

Run this on rocky01:

history | tail -n 5

You should see something like this:

  118  cat /etc/os-release
  119  df -h
  120  sudo dnf -y makecache
  121  systemctl status sshd
  122  history | tail -n 5

Each line has a number you can rerun: !120 runs line 120 again and !! repeats the last command. The famous one is sudo !!, which reruns the last command with sudo after you forgot it. Even faster is Ctrl+R: press it, type makec and bash shows the last matching command, then Enter runs it or the right arrow lets you edit it first.

By default, history doesn’t record when you ran each command. Add a time format to your own ~/.bashrc so it does:

Paste this whole block:

echo "export HISTTIMEFORMAT='%F %T '" >> ~/.bashrc
source ~/.bashrc
hostname
history | tail -n 2

You should see something like this (last lines):

  123  2026-10-27 11:42:18 hostname
  124  2026-10-27 11:42:25 history | tail -n 2

Rocky keeps 1000 commands in memory (HISTSIZE in /etc/profile) and Ubuntu sets 1000 in memory and 2000 in the file in ~/.bashrc. The list is saved to ~/.bash_history when the shell exits.

✅ Checkpoint: history shows a date and time next to your newest commands, and sudo !! or Ctrl+R has worked for you once.

8Let Tab do the typing

On rocky01, type this and press Tab twice (don’t press Enter):

systemctl st

You should see something like this:

start   status  stop

Completion knows the subcommands of systemctl because of the bash-completion package (installed on both VMs). Now type systemctl status ss and press Tab twice: sshd.service is among the matches on rocky01 (ssh.service on ubu01), and typing a few more letters plus Tab finishes the name. Tab also completes file names, so cat /etc/os-r plus Tab gives you /etc/os-release without a typo.

Practise the editing keys on a long line: type echo one two three four, press Ctrl+A (cursor to the start), Ctrl+E (to the end), Ctrl+W twice (deletes four and three), Ctrl+Y (pastes them back), then Ctrl+U to clear the line.

✅ Checkpoint: Tab shows you the sshd service name, and you used Ctrl+A, Ctrl+E, Ctrl+W and Ctrl+U at least once.

9Read exit codes

Paste this whole block on rocky01:

ls /etc/hostname; echo "exit=$?"
ls /nothere; echo "exit=$?"
grep -q priya /etc/passwd; echo "exit=$?"
nosuchcmd; echo "exit=$?"

You should see this:

/etc/hostname
exit=0
ls: cannot access '/nothere': No such file or directory
exit=2
exit=1
bash: nosuchcmd: command not found
exit=127

Zero means success and anything else means something went wrong, with the number saying roughly what. grep returned 1 because user priya doesn’t exist yet (you’ll create her in Lesson 15), which for grep means “not found” and is not an error. Now run sleep 60, press Ctrl+C and run echo $?: you’ll get 130, which is 128 plus signal 2.

✅ Checkpoint: You got 0, 2, 1, 127 and 130 and can say what each one means.