Intro to System Administration

Becoming super user

In order to run admin tools or commands, you need to be a super user, or root. If you don't feel like logging in as root, you can always become one by running the su command. Being executed without any arguments, this command prompts you for the root password and and gives you the root prompt. If you see the root prompt (usually the pound sign), you are the super user. There are several options of the su command:

• option -l or just - allows you to specify the user you would like to become. If the user is not specified

  $ su -

  the root user is the default one. It seems like su and su -l are to be the identical commands, but it's not quite true. In fact, there is a little difference: simple su command leaves you in the current directory, while su -l or just su - brings you in the root home directory:

  [daniel@localhost ~]$ pwd
  /home/daniel
  [daniel@localhost ~]$ su
  Password: 
  [root@localhost daniel]# pwd
  /home/daniel
  [root@localhost daniel]# exit
  exit
  [daniel@localhost ~]$ su -
  Password: 
  [root@localhost ~]# pwd
  /root
    

• You also can become a different user simply specifying the user login as an argument of the su command:

  # su -l daniel

In case you need to run only one command and do not want to switch to super user mode, you can use the sudo command, that executes only one command by elevating your privileges. Please note that you should have these privileges in order to be able to run the command you want. That is, this command doesn't switch you to the super user mode for this single command, but allows a user to use his/her privileges that normally are not available. In order to grant a user these extra privileges, we need to modify settings in the /etc/sudoers file. This is a simple text file with good amount of comments, but it cannot be edited directly. Instead, you should use special command visudo. This command will open the file in vi editor and lets you edit and save it. If you would like to use another text editor, you need to define the environment variable DISPLAY and assign the name of the text editor of your choice as a value :

# export DISPLAY=gedit
# visudo

%users  ALL=/sbin/mount /mnt/cdrom, /sbin/umount /mnt/cdrom

%users ALL=/sbin/reboot

The next step would be to add the users to whom you would like to grant these privileges to the group users. In order to do this, open the file /etc/group in a text editor of your choice, locate the line

users:x:100:

users:x:100:daniel,mary

To test the changes, exit from the super user mode and run the following commands:

$ /sbin/reboot
$ sudo /sbin/reboot

Graphical admin programs

Fedora and Red Hat Enterprise Linux offer a list of graphical tools to perform individual administrative tasks. These tools are available through the main menu or could be executed from the shell.

Admin commands

Administrative commands mostly located in two directories: /sbin and /usr/sbin. The first directory is intended to contain tools that only root can use, while the second one contains mostly programs that can be executed by normal users too, but normal users cannot use the full functionality. For example, the mount command can be run by anyone to see the list of mounted volumes, but only the super user (by default) can mount new volumes.

Configuring modules

Fedora system comes with tools for configuring the drivers that stand between the programs you run and the hardware they use. The intention is to have only the most critical drivers your system needs built into the kernel; these are called resident drivers. Other drivers that are added dynamically as needed are referred to as loadable modules. The trend is to keep basic kernel as lean as possible, so that each running system has only a few drivers and it can dynamically add what it needs.

The commands to deal with modules are:
• lsmod – shows the loaded modules, their size, and the "parent" modules
• modinfo – shows information about the module provided as the argument of the command
• modprobe – allows to load the module given as an argument of the command. Additional arguments let you specify the base I/O address (io) and desired IRQ (irq). This command loads modules temporarily. At the next reboot, the modules your started with modprobe disappear.
• insmod – also starts the given module. The difference with modprobe is that insmod starts only the given module while modprobe runs this module and all related modules.
• rmmod – removes the given module.
• modprobe -r – removes the given module and all related modules.

Managing file system and disk space

• fdisk – is a menu driven program for creation and manipulation of partition tables. It understands DOS type partition tables and BSD or SUN type disklabels. Examples: lists the current partitions of the disk

  $ fdisk -l

• mount – mounts a new volume to a directory. For example, the following command mounts the disk in the device dev/cdrom to the directory /mnt/mycd. The type of the file system on the disk will be detected automatically (-t auto).

  $ mount -t auto /dev/cdrom /mnt/mycd

  Being used with no arguments, this command prints the list of currently mounted devices. File /etc/fstab stores the list of definitions how some devices which should be mounted. This file can look like

  /dev/hdb8 / ext3 defaults 1 1
  /dev/hda1 /boot ext3 defaults 1 2
  /dev/md1 /data ext3 defaults 1 2
  none /dev/pts devpts mode=0620 0 0
  /dev/md0 /home ext3 defaults 1 2
  none /mnt/cdrom supermount dev=/dev/hdc,fs=auto,ro,--,iocharset=iso8859-1,codepage=850 0 0
  none /mnt/floppy supermount dev=/dev/fd0,fs=auto,--,iocharset=iso8859-1,sync,codepage=850 0 0
  none /proc proc defaults 0 0
  /dev/hda6 /tmp ext3 defaults 1 2
  /dev/hdb7 /usr ext3 defaults 1 2
  /dev/hda5 swap swap defaults 0 0
  		

  Where,
  1. field #1 (e.g., /dev/hda6) – the name of the device reresenting the file system. The word none is often placed in this field for file systems that are not associated with special devices.
  2. field #2 (e.g., /tmp) – the mount point in the file system. The file system contains all data from the mount point down the directory tree structure, unless another file system is mounted at some point beneath it.
  3. field #3 (e.g., ext3) – the file system type
  4. field #4 (e.g., defaults) – options to the mount command
  5. field #5 (e.g., 0) – the number in this field indictaes whether or not the indicated file system needs to be dumped (1 means yes, 0 means no).
  6. field #6 (e.g., 1) – the number in this field indictaes wheather or not the file system needs to be checked with fsck (0 is no, 1 means that the file system should be checked first, 2 means that the file system can be checked at any point after the root system is checked)
  Example, create a CD image and mounting it as a new volume

  $ dd if=/media/mycd of=mycd_image.iso 
  $ mkdir /mnt/mycd
  $ mount -o loop mycd_image.iso /mnt/mycd
   

• umount – unmounts the device. The argument is the directory the device is mounted to.
• dmesg – get information about the kernel ring buffer including information about hard disks
• mkfs – creates new file sstem on a given partition
• df – reports the amount of unused spces on the system
• du – reports the amount of used space on the system

Homework assignment

• Read chapter 10 from the course textbook.
• Read about Administrative configuration files (Understanding Linux configuration files)
• Read about Security Enhanced Linux
• Read The UnOfficial SELinux FAQ
• Read about the find command and figure out how to use it.