Linux Hard and logical file system connection?
 

This is a large topic I'm sure but a question that encapsulates the idea is;

How can a directory mount the hard drive? How can a list of files sitting on a hard drive that is not in RAM run your computer and then mount the physical device that stores the code that it is running? Obviously the CPU can access the hard drive before this so how does this massive abstraction make physical sense?I'm thinking that if this is abstracted to this extent so is the "/.../.../" path nomenclature as well. Is this just to change directories and or an environmental variable? Is the file really farther the more slashes that are in it?

Any search terms or explanation that can connect how this is associated with the physical hard drive would be greatly appreciated!

Hi,

Quote:

Originally Posted by theKbStockpiler (Post 4148921) This is a large topic I'm sure but a question that encapsulates the idea is;

Especially when you look at things so broadly!


It doesn't. You need to look at the 'kernel' and how things work.

You should learn the basics of GNU/Linux or Linux and filesystems.


Rute Tutorial & Exposition


Just a few links to aid you to gaining some understanding;


1 Linux Documentation Project

2 Rute Tutorial & Exposition

3 Linux Command Guide 

4 Bash Beginners Guide

5 Bash Reference Manual

6 Advanced Bash-Scripting Guide

7 Linux Newbie Admin Guide

8 LinuxSelfHelp

9 Getting Started with Linux

The above links and others can be found at 'Slackware-Links'. More than just Slackware® links!

I'm trying to point out how abstract it is. I can read the same guide over again to the same effect
 

Those are some great links and Slackware has some of the best documentation that there is but I don't see the connection between what a programmer thinks it is and what is actually happening bit for bit. I am not a piece of data so understanding it from a data only plane does not allow an understanding of how it is stored. I could go into more detail but I think its i just a concept that has been compounded by abstraction and not understood in the physical presence.

Any direct concepts to search for would be appreciated!

Hi,

Your queries are to broad nor provide detail. If you want to understand how things are handled by a GNU/Linux then you should see how the Linux kernel works. Your queries are not specific, but spinning!

Linux Kernel contains several links that should help you to understand some basics.

Linux File System should provide links for further investigation.

From a programmers level of understanding then one should understand the terminology and specifics in handling issues. Just apply some thought into the generation of keywords that should be applied to a search engine that will hopefully provide the needed information.
:hattip:

Thanks for your expertise!
 

Just so you know, my questioning is broad because of the lack of knowledge. I am working from the outside in. If I knew what I was talking about it would be the other way around and this thread would not be here.
I will read your recommendations and I thank you for them!


May the Linux Deities be with you!Serendipity to All! :hattip::hattip::hattip:

The kernel, which is code in memory, mounts (or instructs the CPU to mount, if you prefer thinking of it that way) the file system which basically involves reading a description of the file system from disk.

No, farther from what?

Logical file systems are not logical
 

Linux Logical file systems are over emphasised. Most people make a big deal out of it and it is not important.It's the educational standard to make a big ;over emphasized, deal out of something that does not matter so it appears instrumental when it is not. Why for instance would anyone assume that the windows method confuses the Linux method? They are both equally dysfunctional. I need a software explanation that crosses over to hardware or vise versa. The tree and the hierarchy-directory thing only exist on a piece of paper and is synonymous to nothing. I'm just pointing out how ridiculous it is.If you take the logical file system for how it is presented.


We have a big nothing that is Root.
How can root which is nothing hold encapsulate anything?

Root is a directory.
A directory is a list of files

All of these files share the same hierarchy.
This basically means that they all have the same number of slashes.

What does a slash represent?
Distance from the directory? Probably not but it is probably a measure of how much parsing takes place before the O.S or whatever can access it.


The mounting thing.
I don't understand why everything is not always mounted but it is apparent that an application has to keep it mounted in RAM so if the RAM is cleared the directory has to be remounted. Mounting is not like an electrical circuit that always has continuity. This looks more like VFS topic and nothing else.

My main focal point is the expression of Mountpoint/device/hard-drive.
The mount point is ON the Hard Drive which makes for a schizophrenic logical File System notation.
How can a hard drive need a file to access its self?

I assume that this is meant to imply that something has hierarchy (control) over the hard drive and it is its' master. The data can't exists without the drive but the mount command has control over the drive. (This is looking like the answer to my question which I would not have arrived to with out responding to your reply.)


The whole path thing is also demented. let's say that we have /home/whoever/desktop/filename. A path implies travel which implies that "filename" has to travel through desktop to get to whoever.

Thanks for the reply!

To GURU & theKbStockpiler,
I must admit that the "path" the discussion has chosen to take is fascinating! There is a school of thought which beleives in the theory of"ORDER IN CHAOS"! I am one of them. Out of confusion comes clarity and i empathise with the @OP for complaining that none of you are giving "DIRECT ANSWERS" culled from your own understanding of the way files,et get mounted on a hard disk and the way the Linux file system was organized and why?

"GURU's"link was top-notch, but for a semi-novice like me who wants to find out where all the pre-installed software in ubuntu is located, would appreciate the actual path to "type" to get there, and see the software! Yes, /usr is what was mentioned in the link,but that is not the whole path! I accept the concept of"do it yourself" philosophy in Linux, but i am afraid that will not suit all! It implies that you not only have a basic knowledge of computers and worse, be a"bit intelligent"! It's so abstract and must have turned away many an enthusiastic newbie from exploring, understanding and finally mastering "at least" one flavor of Linux! I am like the @OP, who is not satisfied until i go to the root of it all and i suffered for that "perfectionist streak" throughout my medical studies!

Thx for reading this rather long post, but i hope the "pro's" and the "Geeks" understood what i meant!

warm regards,
vyver.
PS:Bridge the divide!

If you want to go to the root, you have to be a "bit intelligent", and you have to have (at least) the basic knowledge. You also have to know the basics to go to the root in your medical studies. There is no way to explain complex topics, like filesystems, without implying you to know at least the basics.

I fail to understand why you are flaming me and this is the second time! The first was reg. Saurabhgarg's post on "installing Windows over a Linux installation" today! This post is very personal, as far as my medical studies are concerned and i consider that an insult and casting a "slur" on my intelligence! I AM A LINUX NEWBIE and i will humbly accept that i have a mountain to climb! I suggest we shake hands on this in a spirit of friendship and i fully accept that you( among the many seniors here) surely know more about Linux than me! I am here hone my skills in Linux.period!
regards,
vyver.:study:

To TobiSGD,
Off topic, but i felt the need to clear a few things! I have finished my MD in Internal Medicine as a topper in 1991and completed a two year diploma course in Diabetology(cum Laude) in Australia. Presently i am a consultant physician with 19 years experience behind me and yet, i bet ,no Doc. worth his salt will vouch that he is a complete master in Medicine and i guess the same goes for Linux too! Otherwise why the "bug reporting protocol"? Why the "raw hide versions in Fedora"? If you peruse my first post in this thread i have ack., GURU's links as top-notch and were a great help to a linux Newbie like me, to unravel the Linux file systems and the posts were intellectually stimulating and at the same time very useful!
regards,
vyver.
vyver.

Hi

I haven't been involved in software development for a very long time so am probably totally wrong or out of date, but...

The first program to run on a PC is not an OS (Windows, Linux, Unix...) but a bootstrap or bios program. This will live in a non-volatile (doesn't forget when powered down) memory of one form or another (I can just remember editing programs on paper tape with sticky-tape and a punch). The bios may be in memory the processor can immediately access or it may be loaded automatically by hardware (serial link) before the processor is allowed to run.
The bios knows just enough to load the next program into ram and instruct the processor to jump into it. In my laptops case the program loaded is GRUB, depending on which option I select this loads an image of the kernel (including modules) into ram from the disk (hard drive, CD, DVD, USB stick) and then jumps into it.

Martin

I apologize if you feel flamed by me, that was not my intention. I quoted you and gave it as an example that you actually have to be a "bit intelligent" (said by you) to get to the root in both fields (Linux and your medical field), and you need to know the basics. So if any one gives you an answer that you are not able to understand because not knowing the basics, you actually have to learn the basics. No offense meant here.

To the thread from gargsaurabh31: Instead of giving him an appropriate answer, (you remember, he wanted to remove Linux and install Windows) you gave him the answer that he cannot do this and gave him links how to install Windows in Virtualbox. That is not what he intended to do, and taking him from the zero reply list with an answer that doesn't fit his problem is in my opinion bad manner. Pointing out that your answer is not helpful is in my opinion not flaming.

If you feel insulted by my answers feel free to report me, or just rate my answer negative.

Hi,


Terminology can confuse a newbie therefore to communicate without talking down to someone, I prefer to provide links or information for that same person to get some basic understanding. After which we should be able to communicate about a problem. I intentionally don't use tech speak when communicating with someone outside of the field. Look at it in this way, as a MD you need to convey or converse with patients. These same patients don't always understand medical terminology, so you need to convey the diagnosis in the manner they will understand.

As to direct answers for the OP, a lot of the query was not specific and was bloviate to the extent it actually spun around the topics with pompous intent.

If questions were direct with specifics then of course a direct answer would have been presented. I believe the links that were presented should provide the answers with some effort on the readers part. If some confusion from that then present back queries that are clear & specific.

I like to recommend newbies to look at 'How to Ask Questions the Smart Way' so in the future their queries provide information that will aid us in diagnosis of the problem or query. Another good read is 

 How to Answer a Linux Question , which should be read by everyone.

That is when you should return with specific queries so we can provide the means or solution that will address your needs. Sometimes things can be lost in translation when communicating textually.

Quote:

Originally Posted by vyver (Post 4150915) I accept the concept of"do it yourself" philosophy in Linux, but i am afraid that will not suit all! It implies that you not only have a basic knowledge of computers and worse, be a"bit intelligent"! It's so abstract and must have turned away many an enthusiastic newbie from exploring, understanding and finally mastering "at least" one flavor of Linux! I am like the @OP, who is not satisfied until i go to the root of it all and i suffered for that "perfectionist streak" throughout my medical studies!

That's part of the problem for most things that people are not fully aware of how things work. One of the reasons for a hold-your-hand distribution, attempting to automate or allow simple user interfaces for general operations thus shielding the user from learning the intrinsic parts of the system.

If indeed you are the type that is curious, driven and investigative with good diagnostic skills then you will master GNU/Linux. The GNU/Linux you choose will dictate the level of expertize that you will achieve. If you select a GUI driven GNU/Linux then you'll learn the GUI. If you want to learn system level then selection of a GNU/Linux like Slackware or even a BSD will achieve understanding of the intrinsic aspects. Sometimes the "perfectionist streak" as you speak can be a deterrent or cause you to miss or even overlook the simplest terms.

"Do what you love. Know your own bones; gnaw at it, bury it, unearth it, and gnaw it still." - Thoreau

I really don't see a divide but a slight difference in how things have been conveyed. Sometimes my tone can come across as hard or stern with how I present things. My true intent is to help whenever possible but sometimes my impatience can come across as terse.

“Light is the task when many share the toil.”- Homer
:hattip:

What do the commas represent in this address? 123 Fake St, Big City, Nice Country. Not how far Fake St is from Big City. The parsing time of the path is trivial (nanoseconds) next to reading a single page from disk (milliseconds).

It's not about access, it's about figuring about where the data is. To get the data inside the file you have know where on disk its data is stored. To get the location of the file's data you have to read the first page of the file. To find the first page of the file you have to read the file's entry in it's parent directory (etc up to root). To find the first page of the root directory you have to read the description of the file system (this is called mounting). To get the description of the file system you have to know where it starts. To know where the file system starts you have to read the partition table. The partition table is always located at the beginning of the disk.

There are other ways to organize things. For example, colorforth, by Chuck Moore has no file system. The data is stored in 256 word (forth words, not machine words) blocks. There is no need for a file system because Chuck knows where everything is. It's much simpler but not as flexible.