Dirty n00b Questions :)

[Wallybanger]

Hey guys,

Now that I have my box up and running I have some silly n00b questions to ask. Hopefully ya won't mind answering them. :)

Well, I just discovered the SySRq key while trying to take a snapshot... no real questions there yet but it was interesting. Apparently I need to be more careful with the keyboard!
So from the photo below:

1. Any idea why it says "AMD64" in the first line? I've got an intel CPU.
2. It looks like local is my domain. Does it have to be local or can it be anything?
2. After running ls why do the results begin with ./ ../? What do those signify?
3. I'm pretty sure the entries beginning with a dot are hidden files, right? Why don't they have extensions? What is COPYRIGHT? and what are the entries followed by @?
4. I noticed that when changing directories I could change into most of them with cd /dir, cd /dir/ or cd dir but with my DirtyData zpool (?) I could only use cd DirtyData. Any reason why? Are the forward slashes more significant that I may realize?
5. How important is it to have groups or to be a member of a group?
6. I created a user and it set the home directory as /nonexistant. I was doing some reading and it seems pretty common to have a folder called /home that has the users and user files in it. This wasn't automatically created. Is it common to create/use a /home directory in FreeNAS?

It's kinda confusing trying to figure this stuff out for the first time.

Now onto FreeNAS stuff:

7. When I initiated the install of freenas it asked me to create a zpool out of my disks, which I did as a RAIDz2 pool called DirtyData. It then asked me to create a share which I did called movies. It then gave me the structure you see above. Are Jails and Movies VDevs or is there only one VDev called DirtyData? Should I create a zvol (I'm assuming that's a VDev) called movies and then move that share (?) for movies into the movies VDev?

8. I created a RaidZ2 pool out of 8x 4TB disks. Why does the first entry in that list say I have 29tb available and the second entry say I only have 20tb available. Pretty sure there is supposed to be about 21-22tb available.

Thanks Guys! :)

 

[Mirfster]

Any idea why it says "AMD64" in the first line? I've got an intel CPU.

Common identifier for an x64 CPU. Nothing to worry about.

It looks like local is my domain. Does it have to be local or can it be anything?

Can change it as desired, it is easier to use the Web GUI to do so

I'm pretty sure the entries beginning with a dot are hidden files, right? Why don't they have extensions? What is COPYRIGHT? and what are the entries followed by @?

Best to just leave those alone... ;)

I noticed that when changing directories I could change into most of them with cd /dir, cd /dir/ or cd dir but with my DirtyData zpool (?) I could only use cd DirtyData. Any reason why? Are the forward slashes more significant that I may realize?

The "/" is not needed when you are already in the parent folder.

How important is it to have groups or to be a member of a group?

Important if you want to adequately control permissions. Of course you don't have to use groups, but then that makes administration much more of a PITA if you have a lot of users. Say you have a share and you grant a group "Users" permissions to Read/Execute... So it would be easy just to add new users to that group instead of adding them individually to the Share and assigning them Read/Execute permissions.

I created a user and it set the home directory as /nonexistant. I was doing some reading and it seems pretty common to have a folder called /home that has the users and user files in it. This wasn't automatically created. Is it common to create/use a /home directory in FreeNAS?

For some it is, but for proper "Home" Directories you would want to create a separate share and specify "Use as home share". For reference, please see "Table 10.4a: Options for a CIFS Share" in the Manual.

It's kinda confusing trying to figure this stuff out for the first time.

The Manual is a great start and should pretty much answer most (if not all) of your questions... ;)

When I initiated the install of freenas it asked me to create a zpool out of my disks, which I did as a RAIDz2 pool called DirtyData. It then asked me to create a share which I did called movies. It then gave me the structure you see above. Are Jails and Movies VDevs or is there only one VDev called DirtyData?

Short answer: Neither, "DirtyData", "Jails" or "Movies" are vDevs. Now instead of answering further, I will leave you to figure it out with the link I provided to the Manual. :)

Should I create a zvol (I'm assuming that's a VDev) called movies and then move that share (?) for movies into the movies VDev?

Short answer: No a ZVol is not a vDev and unless you are planning on using iSCSI you don't need it.

I created a RaidZ2 pool out of 8x 4TB disks. Why does the first entry in that list say I have 29tb available and the second entry say I only have 20tb available. Pretty sure there is supposed to be about 21-22tb available.

This is explained a lot in the forums, so I foresee a forum search in your future

*** Hint: If you check out the links listed under "Recommended Reading" in my sig you will find great threads that will also go a long way in helping you.

 

[Wallybanger]

Thanks for the reply, Mirfster. I feel silly having asked about the vdevs. Totally forgot that the vdev is my drive array :P

 

[Mirfster]

Thanks for the reply, Mirfster. I feel silly having asked about the vdevs. Totally forgot that the vdev is my drive array :p

No worries, it took a while for things to sink into my thick skull and I still find myself learning.

 

[depasseg]

./ is a representation for the folder that you are currently in and ../ is the representation for the folder one level up. so if you are in /mnt/pool/mystuff and type cd .. you will go to the /mnt/pool directory.

files that begin with a . are usually configuration files and/or usually hidden. but that's different than the ./ and ../

 

[JDCynical]

Why don't they have extensions? What is COPYRIGHT?

The COPYRIGHT file is text of the FreeBSD copyright. It's safe to read it, try using the less command. :)

As for why they don't have extensions, forget extensions when dealing with most UNIX and UNIX-workalike systems. Filename extensions are a microsoft-ism (or at the very least, they are mostly to blame for it still being a thing for this long), *NIX systems have a thing called magic and the file command (seriously, google something like 'unix magic file'). IIRC, MacOS 7-9 environments (not sure about earlier versions) had something called a resource fork which could be used for things like this.

Sadly, it seems that a lot of systems and environments use it minimally, if at all. In my mind, it's debatable if it's due to an idea that it's an 'archaic system' that's outlived it's usefulness, laziness in not implementing calls to the functions, or not wanting to deal with updating the magic file itself (never mind that most systems have a way of adding to the magic 'database').

 

[Wallybanger]

That's exactly the kind of info I was looking for. Thanks guys. I find it a little strange that you don't think extensions are a good idea. It seems to me like a really quick way to identify files by type but it does seem as though the norm with *nix systems is to put all files of x.type in their own folder, is that right?

I read some stuff about magic but it was pretty complex. The only thing I really caught was that it identifies files by code/bits/info at the beginning of the file but it looks like the "file" command does that also. I wish I had started learning *nix when I got my first computer. Shoulda had dual boot! What a fool!

 

[JDCynical]

I find it a little strange that you don't think extensions are a good idea.

Well, I didn't say that they are not a good idea exactly, but there are, IMO, much better ways to deal with 'what program can read this file properly' than file extensions.

For evidence, I point to the number of visual basic trojans out there that take advantage of the stupid default microsoft put into windows, hiding file name extensions for known extensions, by naming the file image.jpg.vbs, tricking a user who doesn't know any better into running a script and getting their machine compromised. Don't get me wrong, magic file checks wouldn't have stopped all of these things, but it wouldn't have been so bleeding easy IMO.

File name extensions are good for quickly identifying a file by name alone, but I don't think that it should be the only way for the environment to determine what program is used to read/run a given file.

It seems to me like a really quick way to identify files by type but it does seem as though the norm with *nix systems is to put all files of x.type in their own folder, is that right?

Not really.

Lemme see if I can put this into a nutshell, somewhat...

First off, *NIX systems don't use the 'drive letter' stuff, as I'm sure you have noticed. Every directory can be considered a filesystem. Most live on the same drive, but they do not have to. A directory could be on a machine on the opposite side of the world accessed via NFS. *NIX systems don't give a flying fsck where the data is, just if it's accessible where the system is configured to look for it. Years ago when I worked on the helpdesk for Mentor Graphics, several of our devs had their *NIX workstations configured like this, with home directories located on NAS boxes that were local (on the came campus), and sometimes project files stored in another state accessed across high speed WAN links.

The directory structure used has a hierarchy as well. Machine specific configs traditionally were located in /etc, binaries lived in /bin, /sbin, /usr/bin or /usr/sbin, depending on the type and use. Binaries and configs that were machine specific were in /usr/local on *BSD style hierarchies, sysV systems tended to use /opt for storage of said stuff. Of course, no one was forced to use these, so depending on the system, you could find a mix, or something even stranger (*cough HP-UX cough*).

I read some stuff about magic but it was pretty complex. The only thing I really caught was that it identifies files by code/bits/info at the beginning of the file but it looks like the "file" command does that also. I wish I had started learning *nix when I got my first computer. Shoulda had dual boot! What a fool!

:) I cut my teeth on *NIX systems on Slackware back when Walnut Creek was still a thing and selling CD copies of their archives (pretty sure I still have the disk set as well somewhere) and free nets were just starting out (my ex hated the long distance bill I would run up).

I still remember a due-hard NT fan I used to work with, poo-pooing my interest in UNIX systems and how UNIX was a 'dead OS'. Funny that, I guess that necromancer is going a really damned good job then, huh? ;)

Google is your friend, there is a metric ton of info out there on the history and why things are the way they are.

There is a bit of a learning curve with *NIX systems, but in my experience, once you reach 'critical mass', it all just clicks into place and suddenly everything makes sense.

Remember, UNIX is user friendly, it's just really picky about who it makes friends with ;)

 

[Mirfster]

Remember, UNIX is user friendly, it's just really picky about who it makes friends with ;)

That is a nice phrase... Going to eventually plagiarize it one day... :)

 

[JDCynical]

That is a nice phrase... Going to eventually plagiarize it one day... :)

Be my guest. IIRC, I first saw it back in my days of reading and posting in the Scary Devil Monastery ;)

(darn, now I want to set up a leaf node again...)