Lab 08 - Disks & Filesystems

• Lab Procedure
• Part 1 - ssh config file
• Part 2 - What do we have?
• Part 3 - Something new
• Part 4 - File it away
• Part 5 - Take a fstab at this
• Part 6 - What is dead may still be read
• Part 7 - Citations & Resources
• Extra Credit - Create a Virtual Filesystem
• Submission
• Rubric

Lab Procedure

Make sure to return to the AWS Learner Lab page (link in Pilot if you forgot to bookmark it) and hit “Start Lab” to turn on your sandbox / AWS instance.

Use ssh to connect to your AWS Ubuntu instance.

Go to the folder that contains your repository (likely named ceg2350-yourgithubusername).

Create a new directory, Lab08.

Create a file named README.md in the Lab08 folder. The Lab 08 Template can be copied from this link:

• https://raw.githubusercontent.com/pattonsgirl/CEG2350/refs/heads/main/docs/Labs/Lab08/LabTemplate.md

You may refer to additional resources outside of the recommended resources provided.

Any resource that you use that contributes to your understanding of exercises in this lab should be cited in the Citations section of your lab answers. To add citations, provide the site and a summary of what it assisted you with. If generative AI was used, include which generative AI system was used and what prompt(s) you fed it.

If you make mistakes with commands in the lab, note them! Writing down what went wrong and what the correction was will significantly help your learning journey. If you cannot find a correction, it will help the TAs or myself understand what point you reached and determine some potential solutions.

BEFORE YOU BEGIN

Create a new AWS sandbox. Your current sandbox is likely in a working state. By creating a second sandbox, if you make any errors playing with the partition tables and filesystems you can destroy and rebuild without affecting your working instance where you have set up an SSH key to your GitHub profile and cloned your repository.

To set up your sandbox environment for this lab, return to Getting Started with AWS Academy and complete steps 8-13 - you’ll use your same private key to connect to the new sandbox environment. The new sandbox will also have the same default username - ubuntu. It will have a unique Public IP.

Once you sign in, I recommend changing the hostname of the system so that you can tell the two instances apart when logged in. On the instance you are making for this lab, run the following command to change the name:

sudo hostnamectl set-hostname disklab-sandbox

To see the new hostname, sign out of the system (disconnect from your current ssh session), then connect back to the system.

Part 1 - ssh config file

Did you read BEFORE YOU BEGIN? Got a second sandbox up? Good. Proceed.

Since you now have two sandboxes, it’s going to be a helpful time to learn a common strategy to configure ssh connection.

1. Verify and write into your lab template the ssh commands for connecting to each of your sandboxes.
2. Fill in the following table of options commonly used in a config file for ssh:

ssh config Option	Description
Host
HostName
User
Port
IdentityFile

1. Write two entries in your local system’s ~/.ssh/config - one for use for each of your sandboxes. Copy your config file contents to your lab template.
   • Notes:
     • you may not have a .ssh folder in your user’s home directory on your system - you’ll need to make the folder and the config file inside of it.
     • Windows users - place your config file in C:\Users\yourusername\.ssh
2. Confirm that when using ssh <host_block_name> that you can connect to each sandbox. Write the working commands into your lab template.

Resources

• SSH Essentials: Working with SSH Servers, Clients, and Keys - DigitalOcean
• How to Configure SSH Key Based Authentication - DigitalOcean
• Using the SSH Config File - Linuxize

Part 2 - What do we have?

Your AWS instances have one block device in use - xvda. In this section, you will explore commands view partition and filesystem information about xvda. Do not make any modification to xvda.

Hint, remember that disk devices are in the /dev/ folder

• Useful Commands: lsblk, parted, blkid, df, cat

For tasks that ask you to use a command, write the command used and include the output of the command.

1. Use lsblk to list only information about the xvda block device.
2. Use parted to print the partition table of the xvda block device.
3. For the xvda partition table:
   • Does it use MBR or GPT?
   • How many partitions are on the block device?
   • What is the largest partition?
4. Use blkid to view information of xvda and it’s partitions. Play with * to get all matches that start with \dev\xvda
5. For the partition with the root filesystem:
   • What is the device name?
   • What is the partition label?
   • What type of filesystem is on the partition?
6. Use df to view file system disk space usage in human readable format (meaning it prints MB/KB/GB)
7. For the root filesystem:
   • What is the total size?
   • How much space is used?
   • Where is it mounted to?
8. View the contents of the filesystem table in /etc/fstab
9. Explain fields & purpose of fields in the entry that mounts the root filesystem.

Resources

• IBM - Displaying partition information using parted
• linuxconfig - How fstab works – introduction to the /etc/fstab file on Linux

Part 3 - Something new

You have had an unformatted disk available on your AWS instance all along. The disk is xvdb - you can see it, but that it has no partitions, if you run lsblk. Time to create a partition table and a partition on /dev/xvdb so that in the next Part we can create a filesystem on the partition and mount it for use.

• Useful Commands: df, lsblk, blkid, gdisk

1. Using the gdisk GPT partition table manipulator, find out what the following main menu options do:
   • p
   • o
   • n
   • i
   • w
2. Edit the xvdb block device with gdisk. Using the main menu, configure the disk to use the GPT partition table type, have at least 1 partition, and have that partition use the Linux filesystem type. Save your changes to the disk.
   • This will be the only partition, so it can use the recommended sizes, which is to say, start at the end of the GPT partition table, and span to the last block of the disk.
3. View the partition table of xvdb
4. Answer the following about xvdb in its current state:
   • What is the device name of the only partition?
   • What is the size of the only partition?
   • What filesystem type will be used on the only partition?

Resources

• DigitalOcean - click “Interactive partitioning with gdisk”
• IBM - Partitioning an GPT disk using gdisk

Part 4 - File it away

Now that you have a partition, you can create a filesystem on it in order to interact with it to store and organize files and create permissions for the files.

• Useful Commands: mkfs, mount

1. Make an ext4 filesystem on the partition on xvdb
2. Use blkid to view information of the partition on xvdb
3. Make a directory in /mnt/ named expanse
4. Mount the partition on xvdb to expanse
5. In expanse create some files and directories
6. umount the partition on xvdb
7. When can I interact with files on the filesystem on the partition in xvdb?

Resources

• IBM - Creating Filesystems - Information on ext4 is folded into ext3 since it makes slight improvements over ext3
• TechMint - How to Change Linux Partition Label Names on EXT4 / EXT3 / EXT2

Part 5 - Take a fstab at this

Right now, every time you want to access your new filesystem on xvdb1 after a system reboot you need to mount it. It would be handy to have it auto-mount. The filesystem table file - fstab is a file that stores information about what to mount when the system boots. Your task in this part is to append a new entry to fstab to automount the filesystem on xvdb1.

1. Make a backup of the current version of /etc/fstab to /etc/fstab.bak
2. Add a line to /etc/fstab to mount the partition on xvdb to the mount point (/mnt/expanse)
3. Test your changes using mount -a to mount / remount records entries in etc/fstab and then check that your additional entry worked (make sure xvdb1 is unmounted first, then test and verify)
4. If you do not think your changes are correct restore /etc/fstab from /etc/fstab.bak. If you think they are correct, you may leave your changes in place.

Resources

• HowToGeek - How to write an fstab file on Linux
• linuxconfig - fstab
• ubuntu - fstab

Part 6 - What is dead may still be read

When you delete a file, you are used to it no longer being accessible, or to it still being temporarily available / recoverable via the Recycle Bin. But once you can’t open it anymore, it should be gone, including from the disk, right? Right?!?

This part will have you acknowledge that to truly make data gone and no longer readable, there are extra steps involved. The general recommendation is to trust nothing, and take disks that have had important data on it, like tax returns, credit card info, passwords, etc, taken to a shredding center and properly ripped to computer-illegible pieces.

• Useful Commands: mount, strings

1. On the filesystem you created on the xvdb partition, create two files, each with a different FAKE secret about you.
2. Find out information on the strings command. If you referred to an internet resource, make sure you cite it by including the URL.
3. Run strings on the filesystem partition on xvdb - read through the output and make an analysis about what output you are viewing.
4. Delete one of the files with a secret.
5. Run strings on the filesystem partition on xvdb - read through the output and determine if the secret, while no longer accessible via the filesystem, is still readable on the partition.
6. Find out information on the shred command. If you referred to an internet resource, make sure you cite it by including the URL.
7. Use shred to overwrite the contents of your second secret file on the disk. Write a short report of steps and provide proof that the file is no longer readable on the disk or accessible in the filesystem. Include an explanation of flags used (if any).

Resources

• freeCodeCamp - How to Securely Erase a Disk and File using the Linux shred Command

Part 7 - Citations / Resources

Any resource that you use that contributes to your understanding of exercises in this lab should be cited in the Citations / Resources section of your lab answers. You may refer to your own notes (if they are digitized into GitHub), course notes, or external resources / websites.

To add citations / resources, provide the reference / link and a summary of what it assisted you with.

If generative AI was used, include which generative AI system was used, what prompt(s) you fed it, and a summary of what it assisted you with.

We expect a minimum of one citation / useful resource per part of your lab. So this lab would require a resource for:

• the ssh config file
• manipulating partition tables gdisk
• creating filesystems with mkfs
• mounting filesystems
• changing the fstab file
• using strings

Extra Credit - Create a Virtual Filesystem

Now that you have made a virtual machine, a thought might be how does it use the disk space on your host. In the context of a virtual machine, the reserved space on the host has a partition table, partitions, and a filesystem mounted to /

Use guides to create a guide on how to complete this tasking.

1. Check for used / available loop device labels
2. Use either dd or fallocate to reserve 2 GB of space (a “file”) from /dev/xvda.
3. Create the loop device and have it use the 2 GB space reserved from the command above.
4. Make a filesystem on the loop device.
5. Mount the filesystem to /mnt/vfs (vfs for virtual filesystem).

Resources

• LinuxOPsys - How to use fallocate
• LinuxOPsys - Creating virtual filesystem with dd
• DZone - Loop Device in Linux

Submission

1. Verify that your GitHub repo has a Lab08 folder with at minimum:

   • README.md

2. In the Pilot Dropbox, paste the URL to the Lab08 folder in your GitHub repo

   • URL should look like: https://github.com/WSU-kduncan/ceg2350s24-YOURGITHUBUSERNAME/tree/main/Lab08

Rubric

Your files should be cleanly presented in your GitHub repository. Citations should be included as needed. Include which generative AI system was used and what prompts were used if generative AI was used.

Rubric