104.2 Lesson 1

Certificate:

LPIC-1

Version:

5.0

Topic:

104 Devices, Linux Filesystems, Filesystem Hierarchy Standard

Objective:

104.2 Maintain the integrity of filesystems

Lesson:

1 of 1

Introduction

Modern Linux filesystems are journaled. This means that every operation is registered in an internal log (the journal) before it is executed. If the operation is interrupted due to a system error (like a kernel panic, power failure, etc.) it can be reconstructed by checking the journal, avoiding filesystem corruption and loss of data.

This greatly reduces the need for manual filesystem checks, but they may still be needed. Knowing the tools used for this (and the corresponding parameters) may represent the difference between dinner at home with your family or an all-nighter in the server room at work.

In this lesson, we will discuss the tools available to monitor filesystem usage, optimize its operation and how to check and repair damage.

Checking Disk Usage

There are two commands that can be used to check how much space is being used and how much is left on a filesystem. The first one is du, which stands for “disk usage”.

du is recursive in nature. In its most basic form, the command will simply show how many 1 Kilobyte blocks are being used by the current directory and all its subdirectories:

$ du
4816	.

This is not very helpful, so we can request more “human readable” output by adding the -h parameter:

$ du -h
4.8M	.

By default, du only shows the usage count for directories (considering all files and subdirectories inside it). To show an individual count for all files in the directory, use the -a parameter:

$ du -ah
432K	./geminoid.jpg
508K	./Linear_B_Hero.jpg
468K	./LG-G8S-ThinQ-Mirror-White.jpg
656K	./LG-G8S-ThinQ-Range.jpg
60K	./Stranger3_Titulo.png
108K	./Baidu_Banho.jpg
324K	./Xiaomi_Mimoji.png
284K	./Mi_CC_9e.jpg
96K	./Mimoji_Comparativo.jpg
32K	./Xiaomi FCC.jpg
484K	./geminoid2.jpg
108K	./Mimoji_Abre.jpg
88K	./Mi8_Hero.jpg
832K	./Tablet_Linear_B.jpg
332K	./Mimoji_Comparativo.png
4.8M	.

The default behaviour is to show the usage of every subdirectory, then the total usage of the current directory, including subdirectories:

$ du -h
4.8M	./Temp
6.0M	.

In the example above, we can see that the subdirectory Temp occupies 4.8 MB and the current directory, including Temp, occupies 6.0 MB. But how much space do the files in the current directory occupy, excluding the subdirectories? For that we have the -S parameter:

$ du -Sh
4.8M	./Temp
1.3M	.

Tip	Keep in mind that command line parameters are case-sensitive: `-s` is different from `-S`.

If you want to keep this distinction between the space used by the files in the current directory and the space used by subdirectories, but also want a grand total at the end, you can add the -c parameter:

$ du -Shc
4.8M	./Temp
1.3M	.
6.0M	total

You can control how “deep” the output of du should go with the -d N parameter, where N describes the levels. For example, if you use the -d 1 parameter, it will show the current directory and its subdirectories, but not the subdirectories of those.

See the difference below. Without -d:

$ du -h
216K	./somedir/anotherdir
224K	./somedir
232K	.

And limiting the depth to one level with -d 1:

$ du -h -d1
224K	./somedir
232K	.

Please note that even if anotherdir is not being shown, its size is still being taken into account.

You may wish to exclude some types of files from the count, which you can do with --exclude="PATTERN", where PATTERN is the pattern against which you wish to match. Consider this directory:

$ du -ah
124K	./ASM68K.EXE
2.0M	./Contra.bin
36K	./fixheadr.exe
4.0K	./README.txt
2.1M	./Contra_NEW.bin
4.0K	./Built.bat
8.0K	./Contra_Main.asm
4.2M	.

Now, we will use --exclude to filter out every file with the .bin extension:

$ du -ah --exclude="*.bin"
124K	./ASM68K.EXE
36K	./fixheadr.exe
4.0K	./README.txt
4.0K	./Built.bat
8.0K	./Contra_Main.asm
180K	.

Note that the total no longer reflects the size of the excluded files.

Checking for Free Space

du works at the files level. There is another command that can show you disk usage, and how much space is available, at the filesystem level. This command is df.

The command df will provide a list of all of the available (already mounted) filesystems on your system, including their total size, how much space has been used, how much space is available, the usage percentage and where it is mounted:

$ df
Filesystem     1K-blocks      Used Available Use% Mounted on
udev             2943068         0   2943068   0% /dev
tmpfs             595892      2496    593396   1% /run
/dev/sda1      110722904  25600600  79454800  25% /
tmpfs            2979440    951208   2028232  32% /dev/shm
tmpfs               5120         0      5120   0% /run/lock
tmpfs            2979440         0   2979440   0% /sys/fs/cgroup
tmpfs             595888        24    595864   1% /run/user/119
tmpfs             595888       116    595772   1% /run/user/1000
/dev/sdb1          89111      1550     80824   2% /media/carol/part1
/dev/sdb3          83187      1550     75330   3% /media/carol/part3
/dev/sdb2          90827      1921     82045   3% /media/carol/part2
/dev/sdc1      312570036 233740356  78829680  75% /media/carol/Samsung Externo

However, showing the size in 1 KB blocks is not very user-friendly. Like on du, you can add the -h parameters to get a more “human readable” output:

$ df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            2.9G     0  2.9G   0% /dev
tmpfs           582M  2.5M  580M   1% /run
/dev/sda1       106G   25G   76G  25% /
tmpfs           2.9G  930M  2.0G  32% /dev/shm
tmpfs           5.0M     0  5.0M   0% /run/lock
tmpfs           2.9G     0  2.9G   0% /sys/fs/cgroup
tmpfs           582M   24K  582M   1% /run/user/119
tmpfs           582M  116K  582M   1% /run/user/1000
/dev/sdb1        88M  1.6M   79M   2% /media/carol/part1
/dev/sdb3        82M  1.6M   74M   3% /media/carol/part3
/dev/sdb2        89M  1.9M   81M   3% /media/carol/part2
/dev/sdc1       299G  223G   76G  75% /media/carol/Samsung Externo

You can also use the -i parameter to show used/available inodes instead of blocks:

$ df -i
Filesystem      Inodes  IUsed   IFree IUse% Mounted on
udev            737142    547  736595    1% /dev
tmpfs           745218    908  744310    1% /run
/dev/sda6      6766592 307153 6459439    5% /
tmpfs           745218    215  745003    1% /dev/shm
tmpfs           745218      4  745214    1% /run/lock
tmpfs           745218     18  745200    1% /sys/fs/cgroup
/dev/sda1        62464    355   62109    1% /boot
tmpfs           745218     43  745175    1% /run/user/1000

One useful parameter is -T, which will also print the type of each filesystem:

$ df -hT
Filesystem     Type      Size  Used Avail Use% Mounted on
udev           devtmpfs  2.9G     0  2.9G   0% /dev
tmpfs          tmpfs     582M  2.5M  580M   1% /run
/dev/sda1      ext4      106G   25G   76G  25% /
tmpfs          tmpfs     2.9G  930M  2.0G  32% /dev/shm
tmpfs          tmpfs     5.0M     0  5.0M   0% /run/lock
tmpfs          tmpfs     2.9G     0  2.9G   0% /sys/fs/cgroup
tmpfs          tmpfs     582M   24K  582M   1% /run/user/119
tmpfs          tmpfs     582M  116K  582M   1% /run/user/1000
/dev/sdb1      ext4       88M  1.6M   79M   2% /media/carol/part1
/dev/sdb3      ext4       82M  1.6M   74M   3% /media/carol/part3
/dev/sdb2      ext4       89M  1.9M   81M   3% /media/carol/part2
/dev/sdc1      fuseblk   299G  223G   76G  75% /media/carol/Samsung Externo

Knowing the type of the filesystem you can filter the output. You can show only filesystems of a given type with -t TYPE, or exclude filesystems of a given type with -x TYPE, like in the examples below.

Excluding tmpfs filesystems:

$ df -hx tmpfs
Filesystem      Size  Used Avail Use% Mounted on
udev            2.9G     0  2.9G   0% /dev
/dev/sda1       106G   25G   76G  25% /
/dev/sdb1        88M  1.6M   79M   2% /media/carol/part1
/dev/sdb3        82M  1.6M   74M   3% /media/carol/part3
/dev/sdb2        89M  1.9M   81M   3% /media/carol/part2
/dev/sdc1       299G  223G   76G  75% /media/carol/Samsung Externo

Showing only ext4 filesystems:

$ df -ht ext4
Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1       106G   25G   76G  25% /
/dev/sdb1        88M  1.6M   79M   2% /media/carol/part1
/dev/sdb3        82M  1.6M   74M   3% /media/carol/part3
/dev/sdb2        89M  1.9M   81M   3% /media/carol/part2

You can also customize the output of df, selecting what should be displayed and in which order, using the --output= parameter followed by a comma separated list of fields you wish to display. Some of the available fields are:

source: The device corresponding to the filesystem.
fstype: The filesystem type.
size: The total size of the filesystem.
used: How much space is being used.
avail: How much space is available.
pcent: The usage percentage.
target: Where the filesystem is mounted (mount point).

If you want an output showing the target, source, type and usage, you can use:

$ df -h --output=target,source,fstype,pcent
Mounted on                    Filesystem     Type     Use%
/dev                          udev           devtmpfs   0%
/run                          tmpfs          tmpfs      1%
/                             /dev/sda1      ext4      25%
/dev/shm                      tmpfs          tmpfs     32%
/run/lock                     tmpfs          tmpfs      0%
/sys/fs/cgroup                tmpfs          tmpfs      0%
/run/user/119                 tmpfs          tmpfs      1%
/run/user/1000                tmpfs          tmpfs      1%
/media/carol/part1            /dev/sdb1      ext4       2%
/media/carol/part3            /dev/sdb3      ext4       3%
/media/carol/part2            /dev/sdb2      ext4       3%
/media/carol/Samsung Externo  /dev/sdc1      fuseblk   75%

df can also be used to check inode information, by passing the following fields to --output=:

itotal: The total number of inodes in the filesystem.
iused: The number of used inodes in the filesystem.
iavail: The number of available inodes in the filesystem.
ipcent: The percentage of used inodes in the filesystem.

For example:

$ df --output=source,fstype,itotal,iused,ipcent
Filesystem     Type      Inodes  IUsed IUse%
udev           devtmpfs  735764    593    1%
tmpfs          tmpfs     744858   1048    1%
/dev/sda1      ext4     7069696 318651    5%
tmpfs          tmpfs     744858    222    1%
tmpfs          tmpfs     744858      3    1%
tmpfs          tmpfs     744858     18    1%
tmpfs          tmpfs     744858     22    1%
tmpfs          tmpfs     744858     40    1%

Maintaining ext2, ext3 and ext4 Filesystems

To check a filesystem for errors (and hopefully fix them), Linux provides the fsck utility (think of “filesystem check” and you will never forget the name). In its most basic form, you can invoke it with fsck followed by the filesystem’s location you want to check:

# fsck /dev/sdb1
fsck from util-linux 2.33.1
e2fsck 1.44.6 (5-Mar-2019)
DT_2GB: clean, 20/121920 files, 369880/487680 blocks

Warning

NEVER run fsck (or related utilities) on a mounted filesystem. If this is done anyway, data may be lost.

fsck itself will not check the filesystem, it will merely call the appropriate utility for the filesystem type to do so. In the example above, since a filesystem type was not specified, fsck assumed an ext2/3/4 filesystem by default, and called e2fsck.

To specify a filesystem, use the -t option, followed by the filesystem name, like in fsck -t vfat /dev/sdc. Alternatively, you may call a filesystem-specific utility directly, like fsck.msdos for FAT filesystems.

Tip	Type `fsck` followed by `Tab` twice to see a list of all related commands on your system.

fsck can take some command-line arguments. These are some of the most common:

-A: This will check all filesystems listed in /etc/fstab.
-C: Displays a progress bar when checking a filesystem. Currently only works on ext2/3/4 filesystems.
-N: This will print what would be done and exit, without actually checking the filesystem.
-R: When used in conjunction with -A, this will skip checking the root filesystem.
-V: Verbose mode, prints more information than usual during operation. This is useful for debugging.

The specific utility for ext2, ext3 and ext4 filesystems is e2fsck, also called fsck.ext2, fsck.ext3 and fsck.ext4 (those three are merely links to e2fsck). By default, it runs in interactive mode: when a filesystem error is found, it stops and asks the user what to do. The user must type y to fix the problem, n to leave it unfixed or a to fix the current problem and all subsequent ones.

Of course sitting in front of a terminal waiting for e2fsck to ask what to do is not a productive use of your time, especially if you are dealing with a big filesystem. So, there are options that cause e2fsck to run in non-interactive mode:

-p: This will attempt to automatically fix any errors found. If an error that requires intervention from the system administrator is found, e2fsck will provide a description of the problem and exit.
-y: This will answer y (yes) to all questions.
-n: The opposite of -y. Besides answering n (no) to all questions, this will cause the filesystem to be mounted read-only, so it cannot be modified.
-f: Forces e2fsck to check a filesystem even if is marked as “clean”, i.e. has been correctly unmounted.

Fine Tuning an ext Filesystem

ext2, ext3 and ext4 filesystems have a number of parameters that can be adjusted, or “tuned”, by the system administrator to better suit the system needs. The utility used to display or modify these parameters is called tune2fs.

To see the current parameters for any given filesystem, use the -l parameter followed by the device representing the partition. The example below shows the output of this command on the first partition of the first disk (/dev/sda1) of a machine:

# tune2fs -l /dev/sda1
tune2fs 1.44.6 (5-Mar-2019)
Filesystem volume name:   
Last mounted on:          /
Filesystem UUID:          6e2c12e3-472d-4bac-a257-c49ac07f3761
Filesystem magic number:  0xEF53
Filesystem revision #:    1 (dynamic)
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype needs_recovery extent 64bit flex_bg sparse_super large_file huge_file dir_nlink extra_isize metadata_csum
Filesystem flags:         signed_directory_hash
Default mount options:    user_xattr acl
Filesystem state:         clean
Errors behavior:          Continue
Filesystem OS type:       Linux
Inode count:              7069696
Block count:              28255605
Reserved block count:     1412780
Free blocks:              23007462
Free inodes:              6801648
First block:              0
Block size:               4096
Fragment size:            4096
Group descriptor size:    64
Reserved GDT blocks:      1024
Blocks per group:         32768
Fragments per group:      32768
Inodes per group:         8192
Inode blocks per group:   512
Flex block group size:    16
Filesystem created:       Mon Jun 17 13:49:59 2019
Last mount time:          Fri Jun 28 21:14:38 2019
Last write time:          Mon Jun 17 13:53:39 2019
Mount count:              8
Maximum mount count:      -1
Last checked:             Mon Jun 17 13:49:59 2019
Check interval:           0 ()
Lifetime writes:          20 GB
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)
First inode:              11
Inode size:	          256
Required extra isize:     32
Desired extra isize:      32
Journal inode:            8
First orphan inode:       5117383
Default directory hash:   half_md4
Directory Hash Seed:      fa95a22a-a119-4667-a73e-78f77af6172f
Journal backup:           inode blocks
Checksum type:            crc32c
Checksum:                 0xe084fe23

ext filesystems have mount counts. The count is increased by 1 each time the filesystem is mounted, and when a threshold value (the maximum mount count) is reached the system will be automatically checked with e2fsck on the next boot.

The maximum mount count can be set with the -c N parameter, where N is the number of times the filesystem can be mounted without being checked. The -C N parameter sets the number of times the system has been mounted to the value of N. Note that command line parameters are case-sensitive, so -c is different from -C.

It is also possible to define a time interval between checks, with the -i parameter, followed by a number and the letters d for days, m for months and y for years. For example, -i 10d would check the filesystem at the next reboot every 10 days. Use zero as the value to disable this feature.

-L can be used to set a label for the filesystem. This label can have up to 16 characters. The -U parameter sets the UUID for the filesystem, which is a 128 bit hexadecimal number. In the example above, the UUID is 6e2c12e3-472d-4bac-a257-c49ac07f3761. Both the label and UUID can be used instead of the device name (like /dev/sda1) to mount the filesystem.

The option -e BEHAVIOUR defines the kernel behaviour when a filesystem error is found. There are three possible behaviours:

continue: Will continue execution normally.
remount-ro: Will remount the filesystem as read-only.
panic: Will cause a kernel panic.

The default behaviour is to continue. remount-ro might be useful in data-sensitive applications, as it will immediately stop writes to the disk, avoiding more potential errors.

ext3 filesystems are basically ext2 filesystems with a journal. Using tune2fs you can add a journal to an ext2 filesystem, thus converting it to ext3. The procedure is simple, just pass the -j parameter to tune2fs, followed by the device containing the filesystem:

# tune2fs -j /dev/sda1

Afterwards, when mounting the converted filesystem, do not forget to set the type to ext3 so the journal can be used.

When dealing with journaled filesystems, the -J parameter allows you to use extra parameters to set some journal options, like -J size= to set the journal size (in megabytes), -J location= to specify where the journal should be stored (either a specific block, or a specific position on the disk with suffixes like M or G) and even put the journal on an external device with -J device=.

You can specify multiple parameters at once by separating them with a comma. For example: -J size=10,location=100M,device=/dev/sdb1 will create a 10 MB Journal at the 100 MB position on the device /dev/sdb1.

Warning

tune2fs has a “brute force” option, -f, which will force it to complete an operation even if errors are found. Needless to say, this should be only used with extreme caution.

Maintaining XFS Filesystems

For XFS filesystems, the equivalent of fsck is xfs_repair. If you suspect that something is wrong with the filesystem, the first thing to do is to scan it for damage.

This can be done by passing the -n parameter to xfs_repair, followed by the device containing the filesystem. The -n parameter means “no modify”: the filesystem will be checked, errors will be reported but no repairs will be made:

# xfs_repair -n /dev/sdb1
Phase 1 - find and verify superblock...
Phase 2 - using internal log
        - zero log...
        - scan filesystem freespace and inode maps...
        - found root inode chunk
Phase 3 - for each AG...
        - scan (but do not clear) agi unlinked lists...
        - process known inodes and perform inode discovery...
        - agno = 0
        - agno = 1
        - agno = 2
        - agno = 3
        - process newly discovered inodes...
Phase 4 - check for duplicate blocks...
        - setting up duplicate extent list...
        - check for inodes claiming duplicate blocks...
        - agno = 1
        - agno = 3
        - agno = 0
        - agno = 2
No modify flag set, skipping phase 5
Phase 6 - check inode connectivity...
        - traversing filesystem ...
        - traversal finished ...
        - moving disconnected inodes to lost+found ...
Phase 7 - verify link counts...
No modify flag set, skipping filesystem flush and exiting.

If errors are found, you can proceed to do the repairs without the -n parameter, like so: xfs_repair /dev/sdb1.

xfs_repair accepts a number of command line options. Among them:

-l LOGDEV and -r RTDEV: These are needed if the filesystem has external log and realtime sections. In this case, replace LOGDEV and RTDEV with the corresponding devices.
-m N: Is used to limit the memory usage of xfs_repair to N megabytes, something which can be useful on server settings. According to the man page, by default xfs_repair will scale its memory usage as needed, up to 75% of the system’s physical RAM.
-d: The “dangerous” mode will enable the repair of filesystems that are mounted read-only.
-v: You may have guessed it: verbose mode. Each time this parameter is used, the “verbosity” is increased (for example, -v -v will print more information than just -v).

Note that xfs_repair is unable to repair filesystems with a “dirty” log. You can “zero out” a corrupt log with the -L parameter, but keep in mind that this is a last resort as it may result in filesystem corruption and data loss.

To debug an XFS filesystem, the utility xfs_db can be used, like in xfs_db /dev/sdb1. This is mostly used to inspect various elements and parameters of the filesystem.

This utility has an interactive prompt, like parted, with many internal commands. A help system is also available: type help to see a list of all commands, and help followed by the command name to see more information about the command.

Another useful utility is xfs_fsr, which can be used to reorganize (“defragment”) an XFS filesystem. When executed without any extra arguments it will run for two hours and try to defragment all mounted, writable XFS filesystems listed on the /etc/mtab/ file. You may need to install this utility using the package manager for your Linux distribution, as it may not be part of a default install. For more information consult the corresponding man page.

Guided Exercises

Using du, how can we check how much space is being used by just the files on the current directory?
Using df, list information for every ext4 filesystem, with the outputs including the following fields, in order: device, mount point, total number of inodes, number of available inodes, percentage of free space.
What is the command to run e2fsck on /dev/sdc1 in non-interactive mode, while trying to automatically fix most errors?
Suppose /dev/sdb1 is an ext2 filesystem. How can you convert it to ext3, and at the same time reset its mount count and change its label to UserData?
How can you check for errors on an XFS filesystem, without repairing any damage found?

Explorational Exercises

Consider you have an ext4 filesystem on /dev/sda1 with the following parameters, obtained with tune2fs:
```
Mount count:              8
Maximum mount count:      -1
```
What will happen at the next boot if the command tune2fs -c 9 /dev/sda1 is issued?
Consider the following output of du -h:
```
$ du -h
216K	./somedir/anotherdir
224K	./somedir
232K	.
```
How much space is occupied by just the files on the current directory? How could we rewrite the command to show this information more clearly?
What would happen to the ext2 filesystem /dev/sdb1 if the command below is issued?
```
# tune2fs -j /dev/sdb1 -J device=/dev/sdc1 -i 30d
```
How can we check for errors on a XFS filesystem on /dev/sda1 that has a log section on /dev/sdc1, without actually making any repairs?
What is the difference between the -T and -t parameters for df?

Summary

In this lesson, you learned:

How to check for used and free space on a filesystem.
How to tailor the output of df to suit your needs.
How to check the integrity of and repair a filesystem with fsck and e2fsck.
How to fine tune an ext filesystem with tune2fs.
How to check and repair XFS filesystems with xfs_repair.

The following commands were discussed in this lesson:

du: View the amount of disk space in use on a filesystem.
df: View the amount of disk space that is available (free) on a filesystem.
fsck: The filesystem check repair utility.
e2fsck: The filesystem check repair utility specific to extended (ext2/3/4) filesystems.
tune2fs: Modifies filesystem parameters on an extended (ext2/3/4) filesystem.
xfs_repair: The equivalent of fsck for XFS filesystems.
xfs_db: This utility is used to view various parameters of an XFS filesystem.

Answers to Guided Exercises

Using du, how can we check how much space is being used by just the files on the current directory?

First, use the -S parameter to separate the output of the current directory from its subdirectories. Then, use -d 0 to limit the output depth to zero, meaning “no subdirectories”. Do not forget -h to get an output in an “human-readable” format:
```
$ du -S -h -d 0
```
or
```
$ du -Shd 0
```
Using df, list information for every ext4 filesystem, with the outputs including the following fields, in order: device, mount point, total number of inodes, number of available inodes, percentage of free space.

You can filter filesystems with the -t option followed by the filesystem name. To get the output needed, use --output=source,target,itotal,iavail,pcent. So, the answer is:
```
$ df -t ext4 --output=source,target,itotal,iavail,pcent
```
What is the command to run e2fsck on /dev/sdc1 in non-interactive mode, while trying to automatically fix most errors?

The parameter to automatically try to fix most errors is -p. So the answer is:
```
# e2fsck -p /dev/sdc1
```
Suppose /dev/sdb1 is an ext2 filesystem. How can you convert it to ext3 and at the same time reset its mount count and change its label to UserData?

Remember that converting an ext2 filesystem to ext3 is just a matter of adding a journal, which can be done with the -j parameter. To reset the mount count, use -c 0. To change the label use -L UserData. The correct answer is:
```
# tune2fs -j -c 0 -L UserData /dev/sdb1
```
How can you check for errors on an XFS filesystem, without repairing any damage found?

Use the -n parameter, like in xfs -n, followed by the corresponding device.

Answers to Explorational Exercises

Consider you have an ext4 filesystem on /dev/sda1 with the following parameters, obtained with tune2fs:
```
Mount count:              8
Maximum mount count:      -1
```
What will happen at the next boot if the command tune2fs -c 9 /dev/sda1 is issued?

The command will set the maximum mount count for the filesystem to 9. Since it is currently 8, the next system boot will cause a filesystem check.
Consider the following output of du -h:
```
$ du -h
216K	./somedir/anotherdir
224K	./somedir
232K	.
```
How much space is occupied by just the files on the current directory? How could we rewrite the command to show this information more clearly?

Of the total of 232 K used, 224 K are used by the subdirectory somedir and its subdirectories. So, excluding those, we have 8K being occupied by the files on the current directory. This information can be shown more clearly by using the -S parameter, which will separate the directories in the count.
What would happen to the ext2 filesystem /dev/sdb1 if the command below is issued?
```
# tune2fs -j /dev/sdb1 -J device=/dev/sdc1 -i 30d
```
A journal will be added to /dev/sdb1, converting it to ext3. The journal will be stored on the device /dev/sdc1 and the filesystem will be checked every 30 days.
How can we check for errors on a XFS filesystem on /dev/sda1 that has a log section on /dev/sdc1, without actually making any repairs?

Use xfs_repair, followed by -l /dev/sdc1 to indicate the device containing the log section, and -n to avoid making any changes.
```
# xfs_repair -l /dev/sdc1 -n
```
What is the difference between the -T and -t parameters for df?

The parameter -T will include the type of each filesystem in the output of df. -t is a filter, and will show only filesystems of the given type on the output, excluding all others.