"This Linux file system is crazy..."

This FAQ is a quick overview that I have written to summarize how the files are generally organized into their directories. I have often described this, in the past, as looking at a vertical vs. horizontal method of determining where different files should be placed. However, that doesn't really clear things up for 99% of the people I have tried to describe this to, so I will attempt to better explain this here. I will cover the file & directory organization in very broad brush strokes. For more detailed information, there are thousands of in-depth resources to be found with a brief google. Most Linux books cover this as well (but are often skipped as they aren't readily understood by most people who haven't worked with Linux for a while).

Half Full Glasses and Photos
The first thing (or one of the first things, anyhow) that a Windows user who has moved to Linux (or any Unix-like OS) must shed is the belief of How Things Should Be. Just because one OS does things a certain way, doesn't mean that any other way is 'wrong'. Like whether you consider a glass to be half-empty or half-full, it is a matter of perspective.

Consider this to be like your digital family photos on your computer. You may keep your files organized by year (2003, 2004, 2005, etc.) and keep the sets of photos belonging to the separate events (Michael's Birthday or Joshua's Football Game) in their own folders inside each year. It is a convenient way to see all the events of each year, plainly organized in front of you when you open that 2004 folder.

But, that isn't the only way that photos can be organized. If you want to go to all of the "Summer Vacation" photos and reminisce about the all family vacation fun you have had, you will find them scattered into all the various year folders. This brings up another way photos can be organized: by subject. This can be done on your family photos, so that all the Birthdays that Timothy has had through the years can be found in their own specific folder. Modeling agencies would almost have to sort by model name, not by year of photo.

Windows and Linux Directories
What that talk about photos have to do with Linux file organization? Quite a bit, actually.

Windows organizes installed applications by their own folder in "My Programs". Each app creates a folder for everything to go into. This may sound extremely neat, but Windows really isn't as neatly organized as the that made it sound. It scatters .dll files around to different folders and modifies the registry, too. Try copying an app by moving the "Program Files" folder to a new computer, and it won't function.

Linux, on the other hand, puts all "executables" (or "binaries") into a bin/ directory. You won't find the man pages (help files) in there, as those are placed in the man/ directory. Any source code you have on your system will be in a src/ directory. It's a different way of seeing how a file system should be organized, and quite often it seems counter-intuitive to a Windows user.

Ok, Why so Many?
While the concept of bin/ may be easy enough to understand, Windows users will next run into the many different bin/ directories (and sbin/, for system binaries that are usually reserved for users with elevated privileges). Often this will be the next big stumbling block. After all, it just seemed to make sense - then they find that there is no "one place" for these binaries.

This is understood when you really accept that Unix has been a multi-user system for quite a long time (longer than Windows has been around!), and its filesystem reflects this. Different conventions exist, and the following is not a fact for all systems, but is meant to illustrate by example:

• The basic OS installs what it needs in /bin/. Everyone needs to have access to these basic functions.
• The distro will often install and additional apps it uses into /usr/bin/. Again, all users may have access to this.
• Add-ons may be installed into /usr/local/bin/, as these files are local to the machine you are working on, and may not exist on another PC, even if installed with the same version.
• Users without write/execute permissions outside their home directory may use /home/username/bin/, so they can have their own executables that are unique to them (perhaps they are learning C and making test programs).

To see what your particular system is set to look for any binaries you ask it to run, perform anto see what directories it uses.

File Extensions...
This is another piece of baggage from the days of DOS, and even Windows attaches great significance to a file extension. In particular, the whole .exe thing is thrown away in Linux. Not only a compiled binary file, but any file may be made "executable" by setting the proper file attribute (either via chmod at the command line, or by mouse click on Properties using the more advanced file managers available with Gnome or KDE). A text file can be made executable, which is then interpreted like a DOS .bat "batch file".

Hidden Files
Files in Unix don't have a "hidden" attribute. A file (or directory) is considered hidden when you start the filename with a dot. For example, the file .bash_history (which contains the last several hundred commands you have issued at the bash prompt) in your home folder is not shown by a plain ls command. If you wish to see the hidden files, you can use ls -a to display all files.

Almost Everything is a File!
Or, if not a file, then it is represented in the Linux file system. Your devices are all in the /dev/ folder. Various system and kernel information is represented (and can be set!) in the /proc/ directory. An interesting task is to do a cat /proc/cpuinfo and look what Linux tells you about your CPU.

The Forward Slash: The 'root' of all Goodness
Unlike what a DOS/Windows user is used to, drives (or, more accurately, partitions are not set up each as separate "tops" of the file tree. There is no C:\filename.ext with no file system above. All partitions that are mounted are assigned to a folder (typically in the /mnt/ directory). Unix and all variants use the forward slash '/' as the starting point for all file references, regardless of which drive, network attachment, or device you are referring to. There is no need to care whether /home/, /boot/, /mnt/cdrom or /mnt/windows_c are on your "Linux drive", or if they are elsewhere. Even devices such as scanners are found in the file tree. It can be a strange concept when you first migrate to Linux, but it makes sense when you become used to it.

More awaits you...
Alas, I will end this here. This little FAQ just touched on the filesystem that Linux (and other Unix variants) use to keep things organized. It may seem totally foreign at first, but once you get used to thinking the *nix way, things do get easier.

In closing, I would like to invite anyone else that wants to add to this with some additional detail or information that I did not cover, please post this information as an "addendum" to this first post. Or, if you feel particularly energetic, write your own short HOWTO or FAQ for the Linux section.



As always, those of you who are reading this and notice an error in what I have said, please let me know so that it can be corrected.

EDIT: I forgot to include a couple of links to some more detailed references.
http://www.comptechdoc.org/os/linux/comman...x_crfilest.html
and
http://web.archive.org/web/20050310142547/...p?articles_id=6
may be of interest...
Also: Filesystem Hierarchy Standard:
http://www.pathname.com/fhs/

Awesome... thanks for the info Not a big Linux person, yet, so this really helps!

Wow, thanks for that info

/bin : system-wide binaries essential to the running of the system
/boot : often separated from the root filesystem (ie on another partition) and contains the actual kernel as well as bootloader config files
/dev : a directory full of devices, populated either by the "devfs" filesystem, or more recently "udev". Really old systems had to have this directory filled with device nodes manually.
/etc : contains configuration files for the system, as well as stuff which doesn't fit into any other directory
/home : contains users' home directories
/lib : system-wide libraries essential to the running of the system
/lost + found : the directory to find things after your drive has been fscked (fsck = file system check)
/media : removable drives (such as CDs, DVDs, floppies) should be mounted to a sub-directory of this dir
/mnt : static filesystems (such as windows drives, etc) should be mounted to a sub-directory of this dir
/opt : used for add-on apps and some other software packages
/proc : a virtual filesystem which contains information about your system
/root : the root user's home directory
/sbin : system binaries, where things that are essential to system maintenance but not to normal users go
/sys : another virtual filesystem, which gives information on the running kernel, and can be used to pass options to the kernel during operation
/tmp : system-wide temporary files
/usr : this directory normally contains a whole new filesystem heirarchy, with its own bin, include, lib, and sbin subdirectories. This is where most installed software should put its binaries and libraries, as well as default option files (normally in /usr/share/<appname>). The /usr directory helps to keep files essential to the running of the system/OS separate from user-installed applications.
/var : contains variable data files such as mailboxes, news directories, spooling files, and logs.

Thank you very much!

The primary reason file extensions are considered a bad idea are because they contain important meta-data about a file (what it is and to a lesser extent what to do with it) in a place that's easy to edit/change or corrupt. Why should it matter what your files are named when you have a billions of CPU cycles that could be used to figure out what they are and what to do with them? UNIX-like operating systems follow this mantra as do other operating systems like BeOS and Mac OS.

Discovering Specific File-type Information Without Extensions Using a CLI
The file[1] is used for finding content-type information about a file. The command file somefile.xyz will look at somefile.xyz and determine what kind of information is stored inside by performing a series of tests. The exact nature of the tests and the file-type information is controlled by the magic[2] file is typically found at /usr/share/file/magic. It is possible to add tests and associated file types to the magic file but there is little need because most distributions ship with a fairly comprehensive magic file.

An example of the file command's out put is:


Generic File Information With the ls Command
If you want more generic information about a file like whether it is a directory, block device, ascii file, link, socket, etc. then you can use the ls[3] command along with either the -G or --color arguments (variable depending on your OS).BSD tends to use -G to colors output, Linux tends to respond to --color or -G, and some *NIX operating systems won't color output at all so you will need to install GNU/ls from the GNU File Utilities[4] to get color output. ls -G or ls --color will show directories in a dark blue, named pipes in yellow, symbolic links in purple, executable files in green, etc.

If you have trouble seeing color or you are using a terminal device that does not support ANSI colors then you can use the -F argument with ls to display special characters after the file names instead of or in addition to color information. symbolic links have an "@" character after their name, executables are marked with a "*", and directories with a "/", etc. More information can be found in the man page for ls[3].

The colors used for differentiating files are controlled by the LS_COLORS environmental variable which can be configured using the dircolors[5] command. The specifics of colorizing ls output are worth of their own FAQ.

It is also perfectly acceptable to use file extensions if you want to. You are free to continue naming your JPEG images with .jpg extensions, doing so may save you a lot of hassle if you ever need to access those files on Windows (as an example). *NIX doesn't force you to use file extensions a certain way, but it doesn't penalize you either.

--Edit 1--
[1] Die.net: file(1): Determine File Type Information - Linux Man Page
[2] Die.net: magic(5): File Commands Numbers File - Linux Man Page
[3] Huntsville Macintosh User Group: Mac OS X / Darwin Man Pages
[4] GNU Operating System - Free Software Foundation: Fileutils
[5] Rt.com: Unix Man Pages: dircolors(1)

Nice overview, but I thought this topic would be about file systems like ext2, ext3, XFS, ReiserFS, etc...

And there are no "folders" in a file system, only directories.

That's interesting, what's the difference between a Folder and a Directory?

Arent they the same thing

Normally, in the *nix world they are referred to properly as 'directories'. A 'folder' is a GUI representation that sort of symbolizes the concept visually.

Very useful guide, thanks

Another good one Mark
Perhaps adding the differences between FS types might be beneficial? I see a lot of users pick ext3 by default since they are unaware of the differences. But I guess I'm just a die-hard gentoo / reiserfs fan

Thanks! But I don't know much about the different filesystem types, except what I have read. Maybe you could write something up, and explain the differences (and advantages/disadvantages)?

Thanks for a very good guide!

Whoa! Awesome!! Double thanks for the info. Tracking this topic for future reference while trying to figure out this OS.

Nice one, helped me understand alot more about how things are organised

Thank you, I've been weaning myself from m$ to Linux for a wee while now and this is probably one of the most helpful threads. Informative and very pertinent. Thank you.

Haha, I'll try, I know you can fill in the rest

I'll make this as simplistic as possible, and I know there are many more, but these would be the most common choices.

Ext2: Commonly used for /boot to keep everything nice and simple, but is by no means the way to go for your main filesystem. This FS does NOT have journaling, meaning if you crash, you can very easily lose data.

Pros:
Clean, safe to use for /boot
Well rounded when it comes to speed
Wide range of support, easily readable and writable from within windows

Cons:
Non-Journaling -- Data Loss is probable in a crash

Ext3: The typical choice most users make, however the ReiserFS has surpassed this FS. Ext3 is well rounded for speed and reliability.

Pros:
Very well rounded FS
Wide range of support, easily readable from within windows

Cons:
Potential problems when handling large numbers of files
Inferior to ReiserFS

Reiser3: Most likely the best choice for users, this FS is extremely fast, very reliable, and can handle the extremes in a filesystem very well due to the BTree algorithm it uses. From massive 100+ gig files to 1+ million files in a directory, reiser can take it, and handle it very efficiently.

Pros:
Extremely fast and reliable in almost any situation
Stable / Bug Free, code is no longer altered unless mandatory
More space efficient than Ext3
Does not fragment over time

Cons:
Limited readability from within windows

Reiser4: The new and improved Reiser3, however support is not readily available in most distros, and not yet widely supported for 3rd party programs. You will NOT be able to read this FS in windows.

Pros:
Far exceeds Reiser3 in nearly every aspect

Cons:
Lack of support - it's relatively new, you typically won't find distros / kernels supporting it without patches
Actively developed, bugs may show up
NO readability from within windows

XFS: This FS is very efficient when it comes to handling very large files, however it lacks greatly in other areas. I'd recommend staying away from this FS unless you have a special use with large files.

Pros:
Efficient handling of large files

Cons:
Normally a "specialized" application filesystem
Performance drops like a brick when it comes to handling multiple small files
Readability in windows (?)

JFS: I'm not really going to go into this one, I'd recommend staying away from it altogether. It's similar to Ext3, but lacks when it comes to performance.



There's a quick overview, I **KNOW** things may be missing and may also be inaccurate, please correct me on anything you find. I'm unable to bring up the FS lists at the moment, so this is entirely off the top of my head. I'll try and fix / add whatever I can tomorrow, when I can bring up my spec sheet for the filesystems.

OK, I agree with most of your points here, but I'd like to jump in with another viewpoint:

ReiserFS - Yes, its good. Yes, its better than ext3. Yes, its possible to read from within windows. Yes, most people should probably be using it. But no, most people won't notice a performance difference between Ext3 and ReiserFS. Ext3 is still journalled, so crashes and hard reboots etc won't screw up your filesystem, and Ext3 is an absolutely rock solid filesystem, and has been for a long time - longer than ReiserFS.

Reiser4 - Reiser4 is the newest version of the Reiser filesystem, and aims to improve on ReiserFS (which is what all previous versions of the Reiser filesystem were known as) in every area. A lot of money is being spent on developing Reiser4 by some large corporations, and Hans Reiser seems to have excelled himself. This filesystem seems to be ahead of every other filesystem available today, for any platform whatsoever. Unfortunately, this includes crashes and instability - this filesystem is so new that it hasn't been accepted as stable enough to be included in the kernel by Linus and Andrew yet. Having said that, I know many people who happily use it, and despite some teething issues, it is useable on a day-to-day basis, as long as you're not averse to having to search around for some solutions to small issues.

awesome thread thanks

I can't wait until Reiser4 is stable enough to become mainstream...

How come Red Hat (Fedora) doesn't use ReiserFS?

I was googling to see if I could find out why Fedora didn't include ReiserFS, when I found this:
http://www.fedorafaq.org/#reiserjfs

Apparently, even though Fedora doesn't officially support Reiser, they include it, and all you have to do is type linux reiserfs at the initial install prompt.

Well, damn, I just upgraded to FC3... well whenever FC4 comes out, I'll have to do this. Thanks for the info

Two questions from a Linux noob.

Installing programs: Where should I install my programs? I installed Firefox in my home directory because that was the default place in the installer. Should I have pointed it to /usr directory? Where exactly? Also, when I install tar.gz files, like for XMMS, where should I put that directory? I don't really want it in my home directory. Also, how do I know where RPM files are installed?

File System: I'm not sure what file system I installed for my Linux distribution (Mandrake 10.1). Is there a way to check? Also, what file systems in Linux can be seen from Windows, because that will be a major plus. On a related note, is there a way that Linux can view my Windows partitions. It'd be cool if I could listen to my music in Linux. They're currently in NTFS.

Thanks for any help. Man, I have so many questions... but I'm learning

Depends mostly on which users will need access to the program, and also on how organized you want your directories.

If you are the only user that will be using that program, and you know you aren't going to have other users using it, then your home is ok. A better place for a system-wide binary would probably be /usr/local/bin. Ultimately, it doesn't really matter too much for a home user (at least I have seen no reason it would). If you have an automatic backup script, you will want all your manually added apps to be in a controlled number of locations, to make backup easier.

The different parts go to their correct parts (i.e. man pages to the man/ directory and executables to the bin/ directory). If you want to find out where your executable went (continuing from the first part I quoted), you can do a which firefox to find out. In my case, the result of that yields /usr/bin/firefox, so I know the executable is in /usr/bin/.

Your answers here are:

• Use the mount command to find out what is currrently mounted (accessable).
  
• I believe that ext2/3 can be read from Windows with a third-party app for Windows...
  
• Yes. I think Mandrake includes this ability by default. DiskDrake or something like that may help. (it may also tell you what filesystem format your drives use, like the mount command does)
  

LOL That was a bit more than the "two questions" you initially stated you had. But we are pretty friendly folk here, and are happy to help.

nice thread. mark and contributors!!!!!!

Very comprehensive FAQ. Thanks again mark

Again, thanks for all the input and compliments.

I also want to note that I added a paragrpah about the forward slash and the file tree.

Very interesting read, I've read the whole of Page 1 (slowly getting through Page 2). But I've been looking for this sort of guide to help me learn more about the folder...I mean, Directory Structure. I seriously did not know that a 'Folder' and a 'Directory' are two different things, but have the same result.

Directory = Physical (Think Command Line)
Folder = Graphical Representation of a Directory

Ext3 = /boot
Reiser3/FS = /root

What about a SWAP partition? Since we're on file structures, I'm confused about the following:

Why does Linux need 3 partitions to install Fedora Core or Gentoo when Windows only needs 1 partition?
/boot (+32M)
/swap (+512M)
/root (remaining space)

My incorrect theory (so please correct me):

- The kernel is stored in /boot as a compressed 2-4MB file.
- The reason it's stored here is because when you turn your system on, the MBR detects GRUB and loads up the kernel here as opposed to /root for file protection???
- /boot decompresses into /swap for your session (for some reason).
- /root all your files, system files...everything (except for the kernel).

^^ You can see why I'm confused.

I think, and please correct me if I'm wrong, that the swap partition is used as virtual memory... like what Windows does with the paging file.

Linux can install just fine on a single partition, if you'd prefer it that way. Here's a few reasons why most distributions have traditionally used 3 or sometimes even more partitions:

- because it facilitates easier backups and reinstalls of the OS

- with a seperate /boot partition, your kernel can be placed in there and then the partition not mounted automatically at boot so that your vital boot files aren't available to be messed around with unless you specifically mount them

- using a swap partition instead of a swap file (linux can happily handle either) again prevents silly errors like people deleting the file by mistake and so not getting any swap space available at next boot

- using multiple partitions can just make managing your system easier - for example many servers use a seperate partition for /var which is traditionally where the http documentroot is on an apache webserver and also where the mail spooler normally resides on mailservers

- many people keep their /home mounted on a seperate partition which they mount from multiple operating systems - back when I used redhat 9 alongside gentoo, I had a seperate /home dir which got mounted from both distros, so I kept the same home folder all the time

Hope that gives you some idea of why multiple partitions are normally used.

Thanks rezza... very interesting

Thanks rezza, I think I understand it all, and according to your notes hopefully this would be a possible partition table (3 HDDs):

hda1 /boot (kernel storage)
hda2 /swap (still confused as to what gets stored here)
hda3 /root (operating system)
hdb1 /home (personal files)
hdb2 /var/httpd/website_pages (client's web sites)
hdc1 /var/mysql_db (mysql database server)

For a mid-size business it would be better to have this (3 PC's):

PC1 - hda1 /boot (kernel storage)
PC1 - hda2 /swap (still confused as to what gets stored here)
PC1 - hda3 /root (operating system)
PC1 - hdb1 /home (personal files)
PC2 - hda1 /var/httpd/website_pages (client's web sites)
PC3 - hda1 /var/mysql_db (mysql database server)

^ I decided to make more then three partitions and reinstall Fedora Core 3 to see what the differences were.

I have two 20GB hard drives and they are partitioned as followed:

hda1 /dev/hda1 /boot reiserfs 102MB
hda1 /dev/mapper/VolGroup00-LogVol01 / reiserfs 5.2GB
hda1 /dev/mapper/VolGroup00-LogVol03 /tmp reiserfs 5.2GB
hda1 /dev/mapper/VolGroup00-LogVol07 /opt reiserfs 5.0GB
hda1 /swap swap 1GB
hda1 none /dev/shm tmpfs 251MB

hda2 /dev/mapper/VolGroup01-LogVol04 /usr reiserfs 5.0GB
hda2 /dev/mapper/VolGroup01-LogVol06 /usr/local reiserfs 5.0GB
hda2 /dev/mapper/VolGroup01-LogVol02 /home reiserfs 5.0GB
hda2 /dev/mapper/VolGroup01-LogVol05 /var reiserfs 5.0GB

So far I haven't seen any benefit in splitting it up like this, but I don't plan on changing it anytime soon though.

Swap= virtual memroy= page file in windows.








Why all reiserfs? Ext3 seems to have better compatibility on Ubuntu.

I was just playing around... I might switch them later on to see if there are any differences, but I'm satisfied with reiserfs right now.

I have a doubt!

In windows there is a folder named "Program Files" In which by defauly all the files go into..



In linux for example, if I get a .bzip or .tar file I usually extract it to /opt folder and do a ./configure, make, make install. Actually In which folder in Linux is it **RECOMMENDED** to extract and Install?


---------------------------------------------------------------
Naveen Chandran
http:\\naveenchandran.blogspot.com

^^^ Only some of the stuff goes into "Program Files". There are .dlls and registry entries and so forth.

But *nix is set up for multi-users, so it is set up a bit differently. There is /sbin/ for sytem binaries (executables), /bin/ for all-user binaries that are part of the installation. Individual users can put their personal executables in the bin/ directory in their home, and "extra" apps that you (as sys admin) want all users to have access to are usually thrown in /usr/bin/ or /usr/local/bin/

I guess for a single-user machine, "Program Files" is fine. For a system that was designed for multi-user, though, it doesn't cut it.

I have tryed linux, liked it, and headed back to windows. Why ? 'cause i got scared of the thought of learning a hole new file system! after this, i really have my head set on linux, wich pops 2 questions on my mind (and yes its only 2 questions )

Wich distro is more suitable to handle media, games, and all around fun? (Yes my computer is my favorite toy !)
How can i have complex games (like counter strike) on my linex (probably using wine, if you have a guide to this prog, please share )

Thanks for the attention, and thank you markjensen and everione else, for showing that linux is not a 2 headed monster !

Which distro? All are about the same, and anything lacking due to potential patent issues can easily be added with a command or two. Same for games (though, it may be easier to keep Windows around for your easy gaming needs).

For those people that install from source, if you do "make install" as root it will actually put the application in the appropriate directories. You don't have to worry about where to put the files.

Oh, well i'll just draw straws then
I'm inclined to fedora, since i see it as the most used, but 'im going to read a litle more about it, and about suse too. thanks, and i'll comeback for those "comand or two"

Recommended Link: (suggested to be added to first post)

Filesystem Hierarchy Standard:
http://www.pathname.com/fhs/

This is the basis of the layout of the *nix file system.

Good reference from my quick overview.

Added to first post. Thanks!

cool thanks ive been meaning to read up on this

A graphical example layout would probably be helpful, I'll make one when I understand it myself (which I more or less do, except for the no top-level hard drive thing. What is the utmost highest level?)

and, ****, I'd hoped I'd be able to get rid of the annoying 'all users' 'username' program folders nonsense. What if you're the only one using your computer, all the time, and are kind of a neatfreak to boot? Whatevs

Thanks for info.

Graphics would be nice!

As far as top level, that is the root, or "/". You can attach (mount) hard drive, networked drives and anything else to locations in the regular filesystem.

And, *nix OSes have been multi-user for a very long time, so the foundation to separate users is a fundamental part of the OS design. You can, however, run as root if you like, and symlink your bin directories to one common spot, if you like. However, if you ever do want to set up (for testing, or what-not) an additional user account, you kinda just screwed yourself and made things a bit more complex by doing by it that way.


EDIT: A few filesystem graphics for you to get ideas from, for starters can be found using Google Image.

Well, I don't know how things are the Linux way, but if you're intending to use multiple disks, consider using a volume manager. LVM or what it is in Linux. It treats all added disks as one large logical volume. I think that's preferable over mounting different drives in a specific locations and play the remaining disk space game for each of those mounts. Instead, set up your system as you want, and when you run out of space, put disk into the computer, add it to the LVM and grow the filesystem running out of space.

Myself I've been spoiled by ZFS and don't want to miss it (Solaris user). I guess under Linux, putting at least your home folders on ZFS/FUSE may be an option if it becomes stable in future, since those will be the ones growing out of proportions while the rest staying more or less stable.

Here, it currently looks like this:

CODE

root@bigmclargehuge:~ > zpool status
  pool: tank
state: ONLINE
scrub: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        tank        ONLINE       0     0     0
          c3d0s7    ONLINE       0     0     0
          c0d0s1    ONLINE       0     0     0

errors: No known data errors
root@bigmclargehuge:~ > zpool list
NAME                    SIZE    USED   AVAIL    CAP  HEALTH     ALTROOT
tank                    293G   68.9G    224G    23%  ONLINE     -
root@bigmclargehuge:~ > zfs list  
NAME                         USED  AVAIL  REFER  MOUNTPOINT
tank                        68.9G   219G  8.03M  /tank
tank/home                   65.5G   219G    21K  /export/home
tank/home/root               336K   219G   336K  /export/home/root
tank/home/servo             65.5G   219G  1003M  /export/home/servo
tank/home/servo/Documents   1.12G   219G  1.12G  /export/home/servo/Documents
tank/home/servo/LargeFiles  11.3G   219G  11.3G  /export/home/servo/LargeFiles
tank/home/servo/Music       2.74G   219G  2.74G  /export/home/servo/Music
tank/home/servo/Pictures    1.10G   219G  1.10G  /export/home/servo/Pictures
tank/home/servo/Video       48.3G   219G  48.3G  /export/home/servo/Video
tank/opt                     448M   219G   448M  legacy
tank/usr                    2.87G   219G  2.87G  legacy
tank/var                    82.0M   219G  82.0M  legacy
root@bigmclargehuge:~ >

"http://www.micronux.com/catalog/article_info.php?articles_id=6" is a dead link, you should get rid of it