OPS235 Lab 5 - Fedora17

This is a draft only!
It is still under construction and content may change. Do not rely on this information.

OPS235 Lab 5 - Archiving Files, Compiling Software Packages from Source

Introduction

• In this lab, you are going to help conserve disk space by learning how to compress and decompress files that are stored your computer server. In addition, you will learn a traditional method of how to install programs by decompressing source code archives and install (compile) them on your computer.

• In addition, you will learn about how certain processes (services) work, and how the system administrator can manage these services (i.e. turn "on" and "off").

Objectives

1. To create and use archive files (tar and tar.gz)
2. Compiling software packages from source code
3. Customising file-system start-up

References

• manpages for mount, fstab, tar, gzip, make, chkconfig, dd, service
• Resources on the web:
  • Services in Fedora 13
  • How to use the dd command

Required Material

• SATA Hard Disk with Fedora 13 (the same one used for Lab 3 and 4)
• Lab log book

Prerequisites

• Completion of Labs 1-4

Introduction

There are two techniques used to place multiple files within one file: the first is to create a filesystem within a file, and the second is to use an archive program to create an archive file. Archive files are often used to contain source code for software; in this lab you will also be compiling software from a source code archive. Finally, you will modify your system startup to eliminate unnecessary services and configure your system to automatically mount a partition.

Instructions

Investigation 1: How do you create an archive file?

Use fedora3
Perform these steps in the fedora3 virtual machine.

Warning!
Don't miss the . at the end of the tar commands below! It specifies what should go into the archive: the contents of the current directory.

1. Change your working directory to /usr/share/doc/sudo*
2. Use the tar (tape archiver) command to create an archive file named /tmp/archive1.tar
   • tar cvf /tmp/archive1.tar .
3. What do the options c, v, and f mean?
4. Record the archive file size.
5. Compress the file using gzip:
   • gzip /tmp/archive1.tar
6. Record the archive file size after compression.
7. Make sure you're still in /usr/share/doc/sudo* and then create a compressed archive:
   • tar cvzf /tmp/archive2.tgz .
8. What does the z option do?
9. Compare the sizes of /tmp/archive1.tar.gz and /tmp/archive2.tgz. Why are they so close in size?

Investigation 2: How do you restore files from an archive?

Use fedora3
Perform these steps in the fedora3 virtual machine.

1. Create the directory /tmp/extract1 and make it your current working directory (change into that directory).
2. Move the file archive1.tar.gz to your current directory.
3. Unzip the first archive you created:
   • gunzip archive1.tar.gz
4. Extract the files from the first archive:
   • tar xvf archive1.tar
5. Are all the files there?
6. Compare /tmp/extract1/README and /usr/share/doc/sudo*/README. Are they exactly the same? Why?
7. Create the directory /tmp/extract2 and make it your current working directory.
8. Move the file archive2.tgz to your current directory.
9. Extract the files from the second archive:
   • tar xvzf archive2.tgz
10. Note that this time a separate gunzip command was not needed. Why?
11. Repeat the previous command, leaving out the option "z". Does it work? Why?
12. Compare the README file in this directory with the original file. Are they exactly the same?

Investigation 3: How do you build software from source code?

Use fedora2
Perform these steps in the fedora2 virtual machine.

Compiling requires the correct tools to be installed.
In order to build software from source code, you must have the appropriate software development tools (such as make and gcc) and libraries (such as GTK) installed. The required tools will vary depending on the computer languages used in the software being built. In Fedora, the following command will install a basic set of development tools and libraries:

yum groupinstall "Development Tools" "Development Libraries"

You may need to install additional tools or libraries in order to compile a particular software package.

1. Go to the directory /tmp
2. Use the wget command to download the "tar ball" that contains the source code for the NLED text editor. wget is a command-line tool to download files from the web using the http or ftp protocols.
   • wget http://cdot.senecac.on.ca/software/nled/nled_2_52_src.tgz
3. Extract the files. Change to the newly-extracted directory (/tmp/nled-2.52)
4. Check to see if there is a file named configure. If so, run it; if not, skip this step. (Most but not all source code archives contain this file)
5. Check to see if there is a file named Makefile or makefile. If so, type the command:
   • make
6. What does make do?
7. Some software distributed as source code can automatically install itself. Try this command:
   • make install
8. Most but not all source code archives include the capability of installing themselves this way.
9. If the command make install does not work (how can you tell?), copy the nled program manually:
   • cp nled /usr/local/bin
10. Test nled to make sure it works.

Investigation 4: How do we turn system services on and off?

Use fedora2
Perform these steps in the fedora2 virtual machine.

Security Advice
We have seen that maintaining unneeded packages can be a security risk due to the unnecessary increase in the complexity of your system. Similarly, it is also unnecessarily hazardous, and even more so, to leave unneeded services running. In this investigation, we will learn how to control services, and turn off those that we think are not necessary.

1. Determine your current runlevel using the runlevel command.
2. Change to the appropriate startup directory in a terminal window. For example, if your runlevel is 4, change to /etc/rc.d/rc4.d
3. Observe the names of the symbolic links in that directory. What is the target (destination) of the links to?
4. If you are not using the graphical user interface (GUI), use the startx command to start it.
5. Run the menu option System>Administration>Services. You will then see the Service Configuration screen.
6. Find the bluetooth service and select customize from the tool bar. and deselect the checkbox for your runlevel from step 1. Quit the app.
7. Return to the terminal window you used in Step 2. Get a listing of the symbolic links in the startup directory. Is it any different from what you observed previously?
8. Using the Service Configuration screen, re-enable the bluetooth service.
9. Check the symbolic links again. What has changed this time?
10. Run this command:
    • chkconfig bluetooth off
11. Now what has changed in the directory?
12. Run this command:
    • chkconfig bluetooth on
13. Now what has changed?
14. Run the command:
    • chkconfig --list
15. What does the output show?
16. What is the relationship between the Service Configuration tool, the symbolic links in the startup directory, and the chkconfig command?
17. You can also use chkconfig to set services to be on or off for certain runlevels. For example, to turn httpd on for runlevel 4, we issue the command:
    • chkconfig --level 4 httpd on
18. To turn it off, we type:
    • chkconfig --level 4 httpd off
19. To increase your computer's security, make sure these services are disabled:
    • btseed, bttrack, cups, irda, mdmonitor, netfs, nfs, nfslock, pcscd, rpcgssd, rpcidmapd, rpcsvcgssd, sendmail, ypbind.

Determining Required Services
You should understand that while most services in this list are often not needed, or are needed under circumstances different than those which occur here, there may be situations in which some of them are required, and it may be up to you as an administrator to determine which services are needed, and which should be turned off.

You can find out some basic information about a particular service by querying rpm for information about the package that installed the service file in /etc/rc.d/init.d. For example, to find out about the cups service, execute: rpm -qif /etc/rc.d/init.d/cups

Investigation 5: How do we automatically mount a new partition in Fedora?

Use your Fedora host
Perform these steps on the Fedora Host (f13host).

1. Create the empty directory /lab5
2. Verify that the /dev/sda5 partition is not in use, then reformat /dev/sda5 as ext3
3. Check the man page for /etc/fstab by entering the command man fstab
4. Edit your /etc/fstab file to mount the partition /dev/sda5 to /lab5
5. Issue this command:
   • mount -a
6. What does that command do?
7. Confirm that the partition mounted and copy some files to it.

Completing the Lab

Arrange evidence for each of these items on your screen, then ask your instructor to check each item:

1. [ ] /tmp/fstest filesystem counted with files copied to it.
2. [ ] nled is installed
3. [ ] Correct services are disabled
4. [ ] /lab5 has an entry in /etc/fstab

Preparing for the Quizzes

1. What is the advantage of disabling services such as bluetooth?
2. What is the difference between a .tgz file and a .tar.gz file? What do these stand for?
3. How do you ensure that a filesystem is mounted every time the system is started?
4. How is creating and using a filesystem in a file different from creating and using a filesystem in a block device (such as a partition)?
5. What is source code?
6. How do you build software from source code?
7. Which is preferred: installing from an RPM file, or installing from source code? Why?
8. How do you use chkconfig?
9. The startup directory for each runlevel contains symbolic links. Where are the targets of the symbolic links?
10. When you use the menu option System>Administration>Services, what program is run?

Backup your work
Backup your disk images for fedora2 and fedora3 Virtual Machines.