Introduction

In this lab we will review the mount commands, the /etc/fstab file, the /etc/ttys file and the dmesg utility. Using an editor of choice we will also modify the startup configuration file /etc/rc.conf. FreeBSD accesses several files during the start-up procedure and we need to know which ones we should edit for our custom configurations. Determine when to boot into single user mode (someone changed the root password?). Review the dmesg utility and various files in the /var/log directory. We will also configure the network using the ifconfig and /etc/rc.d/netif commands and review dhclient. Review the find and tar commands to locate and save files that have changed. (This you will need to complete your assignment)

Lab Objectives

• Configure a network using both DHCP and static IP information
• Install the X server software
• Install a full desktop environment (KDE)
• Configure your system for the X environment and KDE desktop (on boot)
• Finding your way around virtual terminals (F1-F7)
• Describe the FreeBSD startup process
• Describe a "service" and determine which services are configured
• Configure new services

On the host machine:

• Start VMware from the host workstation
• Create a new guest OS and use the ISO images found under D:/courses/ISO (7.2-RELEASE)
• Follow the instructions for your guest virtual environment for creating a logical drive and bridged network
• Make sure that the CDROM is available to the guest OS and that it is pointing to ISO images on the file system
• Sysinstall will take you through the installation and prompt for your custom configurations
• The Disk partition editor let's you take a default configuration by choosing the entire drive by choosing "A" for "All"
• Once you have the partitioning done you can choose "Q" to quit
• You will now be prompted to select a boot loader, choose the default
• The Disk Label Editor then asks you for a partioning scheme for your file systems. press "A" for "Automatic", then "Q" to quit

sysinstall - choose a packaged installation:

• choose a standard installation and the following distribution set: "Average User With X - Binaries, Docs and X Windows"
• when prompted answer no to the following configuration options:
• nfs, inetd, dhcp, creating new user accounts

• Complete the installation and post-configuration tasks, just like we did in Lab 1
• Reboot into the freshly installed system

Step 1

Most of you are familiar with the bash shell from unix admin 1, (the bash shell is used more in the linux world rather than the unix world but is available if you prefer). It is not included in FreeBSD by default (Due to it being under the restrictive GPL license, rather than the liberal BSD license) so we will have to install it. We can install packages from our distribution media using sysinstall but we can also install from the command line using the pkg_add utility. See the BSD Handbook and the man pages for review of the mount commands and BSD Handbook on pkg_add. We will cover these topics in a later lab in more detail.

• Mount the disk and find the bash package (use the find command)
• You can also install using the remote switch with pkg_add
• pkg_add –r firefox will look at the FreeBSD site for the latest version of firefox.

  How did you mount the ISO images, what was your mount point?


  _________________________________________________________________________________________

  What file system type do CDs use?


  _________________________________________________________________________________________

  How did you find the bash package on the ISO images?  What commands did you use?


  _________________________________________________________________________________________

  What remote location does the command pkg_add –r  install from ?


  _________________________________________________________________________________________

Step 2

Change your account to use the bash shell - Change the default shell for your account with the chsh command. There is a program called chsh which allows ordinary users to change their shell. The chsh command will use the Environment variable setting in your account to find out what editor to use. By default it is set to the vi editor. You can change this environment variable to point to another editor eg: ee (easy editor). To change your environment variables if you are currently using the csh or tcsh shell you can use:

	setenv EDITOR ee
	setenv VISUAL ee
        

      What command can you use to change the environment variable for your default
      editor if you are using the bash shell? 

  _________________________________________________________________________________________

      How would you change your default editor permanently (subsequent logins)?.

  _________________________________________________________________________________________

Step 3

        mount -uw /     (this command would upgrade the root (/) partition to write access).
      

    Now try to run the passwd command to change the password.
    What happens and why?

  _________________________________________________________________________________________

  _________________________________________________________________________________________

    You will also have to mount any other partitions that hold the utilities
    that you want to use. What utilities would you need to use to modify a file?
    Do you have access to an editor and if not where would you find the editor?

  _________________________________________________________________________________________

  _________________________________________________________________________________________

    mount -at ufs  (this command will mount everything that is listed in
    the /etc/fstab file that has a file type of ufs)  There is a very good
    reason why you wouldn't want to mount all file system types. An example
    is nfs or (network file systems) and we will discuss this in the nfs lab.
    Mount the partitions that you will need to modify the root password.

  _________________________________________________________________________________________

  _________________________________________________________________________________________

    Once you are in single user mode then you can change the root password
    (without knowing the original password) !!So can everyone else!!
    This of course suggests that there should be some protection here.
    What could we do to change that?  There is an entry in the /etc/ttys
    configuration file that controls whether a password must be entered at
    single user mode. Setting this will make your system more secure
    (from a physical access) but it also means that if you have forgotten
    the root password you would not be able to recover using single user.   

    Edit the /etc/ttys and find the entry for this setting.  What entry is it?

  _________________________________________________________________________________________

Step 4 - Configure a single Network device with DHCP

As we saw during the installation we can configure the network with sysinstall. For those who really like menus this is fairly quick and easy. It is also very important to know which files are being modified and also how to make changes "on the fly" which is what system administrators often. These changes will not persist once the machine is rebooted, these are useful for testing, as if the new configuration is broken, you can just reboot the machine and it will return to the original configuration. This is especially recommend when building firewalls, incase you lock your self out of the system. Running sysinstall really doesn't require much discussion, it is quite straight forward and the fields will be very familiar to you.

• run sysinstall
• choose post install tasks => networking => DHCP

        test your configuration with ifconfig (what switches would you use?)
        Check the man pages for switches and list them here.

  _________________________________________________________________________________________

        Running the configuration from sysinstall will modify all of the startup
        scripts for you so that on the next boot dhclient will run. This is
        good but as system administrators you will be making many changes
        especially as you are testing. You could run the dhclient command
        interactively as well as we saw in last weeks lab. 
              dhclient (interface name)
        What is the name of our network card and how did you find out? 

  _________________________________________________________________________________________

Step 5 - Configure a single Network device with static IP

Of course you could also configure your static network environment from sysinstall. Again all the changes will be made permanent for you. But first we want to test out our network configurations so let's take a look at ifconfig. Remember ifconfig -a will give you information on the devices and on the status of any networks that are configured. ifconfig can also be used to configure your networks

            The long format:
    ifconfig le0 inet 10.87.0.21 netmask 255.255.255.0 up
            Normal:
    ifconfig le0 10.87.0.21 netmask 255.255.255.0
            Short:
    ifconfig le0 10.87.0.21/255.255.255.0
            Shortest: (/24 means '24 bit subnet mask' 255.255.255.0)
    ifconfig le0 10.87.0.21/24
        

Simple and quick. In fact we can make it even simpler and quicker. The inet switch is to tell ifconfig that we are using IPv4 (the default), so we can leave that out. We can use /24 notation to describe the netmask and if it was a "class address" we could eliminate the netmask altogether, ifconfig would take the class default. We don't need the "up" keyword because again it defaults to bringing the interface up. So the command to start your network interface configured for the Internet just got a lot shorter. We will need more than the device configured for our network to work properly though. What about DNS and default gateways? Where do we put the DNS server entries? We will not be able to access any other networks without a default gateway.

    ifconfig 
    netstat -rn
                (gives you information on configured routes)
    route add default 192.168.0.1
                (change the ip to your default gateway)
    hostname
                (no parameters will show your hostname, or you can provide a new hostname)
      

    Make a note of the commands that you used to configure your network using the 10.100.row0.seat
  
  _________________________________________________________________________________________
  
  
  _________________________________________________________________________________________

    See if you can find them in /etc/defaults/rc.conf and make a note of them here.
  
  _________________________________________________________________________________________
  
  _________________________________________________________________________________________
  
  _________________________________________________________________________________________

Step 6 - Configuring the X windowing system and the kdm session manager

Once you have installed the basic X environment you could start X from the prompt using your regular account with the command startx. This will start the X server and using default settings will give us a Window Manager called TWM.  It is an old window manager but still quite useful. Window managers provide extra functionality than a text based shell only. Having multiple windows can be very useful for multiple tasks but there is a cost in overhead and if you are managing a server you may not want to even install "X". There are lots of window managers available for the unix environments, examples of which are: Window Maker, and AfterStep. There are also full blown Desktop environments like GNOME or KDE. The advantage of a simple window manager over a full blown desk top environment is "GUI capability" without the overhead. You could also configure your system to have the X server started automatically on boot. If X gets started at boot then we will need a login manager like "kdm" or "xdm" to handle the authentication. To have X started at boot time, and to select which login manager to use, you must modify the terminal configuration file /etc/ttys.

The virtual terminal entry for virtual terminal ttyv8 (ALT F9, since terminals are numbered from 0, and ttyv0 is F1) is set to "off". You will need to set this entry to "on" and make sure that it points to the "kdm" login manager.

    How are the other virtual terminals used (remember alt F1- F8)?

  _________________________________________________________________________________________

    How does Alt-F1 differ from the other virtual terminals and how would you use it?

  _________________________________________________________________________________________

Step 7 - Saving our configurations or customizations for the following weeks

We will be making lots of changes and will want to save our changes as we customize our systems. An easy way to do that is to copy any files that you have changed to your CSUnix account. You will need to set up networking and then you can scp the files.

      tar -czf myfilename.tar /etc/rc.conf /home/username/somefile /etc/morefiles /boot/filename
      

Or you could put a list of all the files that you have changed into a file and pass that file name as an argument to tar (called an include list):

• start by touching a file under the root (/) directory - call it my_changes - (This you should have already done at the beginning of the lab)
• at the end of the lab run the find command looking for files that are newer than the file that you created with touch (see the man page if you need help)
• redirect the output from the previous command to a file (the my_changes file)
• now you can tar all of the files that you have modified by providing the tar command with the include list
• now we can copy the tarball using scp

      scp myfilename.tar 000123456@csu.mohawkcollege.ca:filename.tar
            (Note: after the colon you can provide a path, if you do not,
            the file will be saved in the root of your home directory)
            
            To copy the file back next class, you would simply reverse the command:
            
      scp 000123456@csu.mohawkcollege.ca:filename.tar myfilename.tar
      
            Notice how the scp command is very similar to the cp (copy) command,
            the only difference is that it allows you to provide a username and
            a remote host, and copy the file to or from that remote machine.
      

Verifying the Lab

• Test the network configuration, after a reboot as well
• X starts up on boot with a login manager
• Demonstrate booting into single user and changing the root password