Shulou(Shulou.com)06/01 Report--

Editor to share with you the example analysis of the file system in Linux, I believe that most people do not know much about it, so share this article for your reference, I hope you can learn a lot after reading this article, let's go to know it!

Generally speaking, the file system on Linux is EXT2 or EXT3, but this article does not intend to talk about them directly, but hopes to learn about Linux's file system step by step by combining the Linux operating system and starting with the hard disk, the foundation of the file system.

1. Physical storage mechanism of mechanical hard disk

Most of the file storage functions of modern computers are provided by devices such as mechanical hard drives. (today's SSD and flash memory are partly inherited from mechanical hard drives in terms of concept and logic, so there is no problem with using mechanical hard drives to understand.)

The mechanical hard disk can realize the function of information storage based on: the magnetic storage medium can be magnetized, and the magnetized state can be retained for a long time after magnetization, and this magnetized state can be read out, and at the same time, this magnetized state can be constantly modified. Magnetization happens to have two directions, so it can represent 0 and 1.

So the hard disk is to make this magnetic storage medium into a disk, each disk is distributed with a large number of magnetic storage units, using a magnetic read-write head to write and read the disk (similar to the playing of vinyl records in principle).

There are hundreds of millions of magnetic storage units in a hard disk (there are about 8 billion magnetic storage units in a 1T hard disk), so a set of rules are needed to plan how information is accessed (for example, a book that stores information is divided into pages, each page is read from top to bottom, from left to right, and there is also a chapter directory).

As a result, there are these physical and logical concepts:

A hard disk has multiple disks stacked, and different disks are numbered. The storage particles on each disk are arranged in a circular circle. Each circle is called a track, and there is a circle of storage particles on each track. Every 512x8 (512byte, 0.5KB) storage particles are used as a sector. The sector is the smallest physical unit stored on the hard disk. N sectors can form clusters. N depends on the file system or the configuration of the file system. The cluster is the smallest storage unit in this file system. The same track on all disks forms a cylinder, called a cylinder, which is the smallest unit of the system partition.

When the magnetic head reads and writes the file, it is first read and written by the partition, and the corresponding track and sector are found by the inode number (introduced after the only number in the area), and then read and write cylinder by cylinder. The read and write control system of a mechanical hard disk is an amazing precision engineering (there are hundreds of millions of storage units on a disk, each track is less than dozens of nanometers wide, and the disk rotates tens of thousands of revolutions per minute). At the same time, there are many details about the logic of reading and writing (for example, the number of sectors is not continuous), which is very interesting, you can search for articles to expand reading.

Having a hard disk does not mean that LInux can store it immediately. It also needs to be integrated into Linux's file system before it can be used by Linux.

2.Linux file system

Linux manages the data and hardware resources in the computer in the form of files, that is, everything is a file, which is reflected in the file types of Linux: ordinary files, directory files (that is, folders), device files, link files, pipe files, socket files (interface for data communication) and so on. These kinds of files are managed by Linux using directory tree. The so-called directory tree is a kind of file structure which is mainly based on the root directory (/) and presents a branch down. Different from the pure file system such as ext2, I call it the file system. The resource management mode of all files and file directory tree constitutes the file system of Linux, which makes it easy for the Linux operating system to use system resources.

Therefore, the content of the file system is much less than that of the file system. The Linux file system mainly focuses on the implementation of the operating system-related things with the carrier of the file: the file system is mounted on the operating system, and the whole operating system is placed in the file system. However, there is not much about the file system in this paper, and the file system can be used instead of the file system in most places.

File types in Linux

Let's talk briefly about the file types in Linux, focusing on normal files, directory files, and symbolic link files.

Ordinary files (-) from the perspective of Linux, file types at the application level such as mp4, pdf and html all belong to ordinary files. Linux users can view, change and delete directory files according to their access rights (ddirectory file) directory files are not easy to understand for users who are used to Windows. A directory is also a directory of files. a file contains the file names under their respective directories and pointers to these files. To open a directory is actually to open a directory file, as long as you have access permissions. You can access the files in these directories at will (the execution permission of ordinary files is the access permission of directory files), but only kernel processes can modify them, although they cannot be modified. But we can look at the contents of directory files through vim. This type of file is similar to the shortcut in Windows. It refers to an indirect pointer to another file, that is, what we often call soft link block device files and character device files. These files are generally hidden in the / dev directory. It will be used when reading and interacting with peripherals, for example, the disk drive is a block device file, and the serial device belongs to the character device file system. All devices in the character device file system are either block device files or character device files. Without exception, FIFO pipe files are mainly used for inter-process communication. For example, using the mkfifo command, you can create a FIFO file, enable a process A to read data from the FIFO file, start process B to write data to FIFO, first-in, first-out, and read as you write. Sockets are used for network communication between processes, and can also be used for non-network communication between local machines. These files are generally hidden in the / var/run directory, proving the existence of related processes.

There is no so-called extension for Linux files. Whether a Linux file can be executed or not is related to its executable attributes, as long as you have x in your permissions, such as [- rwx-r-xr-x], it means that the file can be executed, regardless of the file name. Unlike files that can be executed under Windows, the file extension is usually .com .exe. Bat, and so on.

However, being able to be executed is not the same as being successful. For example, the install.log under the root main directory is a text file. Can the file be executed successfully after the permission is changed to-rwxrwxrwx? Of course not, because there is no executable data in its content. So, this x represents the executable ability of the file, but whether it can be executed successfully or not depends on the content of the file.

Even so, we still want to know what the file is from the extension, so we usually use the appropriate extension to indicate what kind of file it is.

So the file name on the Linux system really just lets you know what the file might be used for, and whether it is really executed or not still requires permission specifications. For example, the common / bin/ls instruction that displays file attributes becomes unexecutable if the permission is changed to unexecutable. This problem most often occurs in the process of file transfer. For example, if you download an executable file on the network, but cannot execute it in your Linux system, it may be that the properties of the file have been changed. And if it is transferred from the network to your Linux system, the attribute permissions of the file will indeed be changed.

Linux directory tree

For Linux systems and users, all operable computer resources exist in the logical structure of the directory tree, and the access to computer resources can be regarded as the access to the directory tree. As far as the hard disk is concerned, all access to the hard disk becomes access to a node in the directory tree, that is, a folder, and you do not need to know whether it is a hard disk or a folder in the hard disk.

The logical structure of the directory tree is also very simple, starting from the root directory (/) and constantly expanding down all levels of subdirectories.

3. Hard disk partition

Hard disk partition is the first step in integrating the hard disk into the file system, which essentially transforms the physical concept of "hard disk" into the logical concept of "zone" in preparation for the next step of formatting.

So dividing itself is not necessary, you can use a whole hard disk as an area. However, from the point of view of data security and system performance, partitioning still has many uses, so the hard disk is generally partitioned.

When it comes to partitioning, you have to mention the most important first sector on each hard disk, which contains the hard disk master boot record (Master boot record, MBR) and partition table (partition table), of which MBR occupies 446 bytes and partition table occupies 64 bytes. The hard disk master boot record contains the most basic boot loader, which is the key link of the system boot, which is described in more detail in the appendix. The partition table is related to the partition, which records the information about the partition of the hard disk, but because the partition table only has 64bytes, it can only remember up to four partitions (the partition itself actually sets up the partition table).

It is too little to be divided into four sectors, so there is the concept of extended partition. Since the partition table where the first sector is located can only record four pieces of data, can I use additional sectors to record more partition information.

The ordinary accessible partition is called the primary partition, the extended partition is different from the primary partition, it has no content, it provides space for further logical partition. After a partition is designated as an extended partition, the extended partition can be further divided into multiple logical partitions. The operating system stipulates that:

Each of the four partitions can be a primary partition or an extended partition can only have at most one (and not necessarily multiple) extended partitions that can be further divided into multiple logical partitions, just a logical concept and cannot be accessed itself. that is, a partition that cannot be formatted as data access, only the number of primary and logical partitions that can be accessed as data varies according to the operating system. In Linux systems, IDE hard drives have up to 59 logical partitions (numbers 5 to 63) and SATA drives have 11 logical partitions (numbers 5 to 15)

Generally, when partitioning a hard disk, it is best to have one primary partition and one extended partition, and then divide the extended partition into N logical partitions.

Is it possible not to have the primary partition? Do not know, but do not seem to care, when you create a partition will automatically configure you with a special type, you'd better separate a swap area (memory replacement space), it is a unique category, the function is: when there is data stored in physical memory, but the data is not often used by CPU, then these infrequently used programs will be thrown into the swap replacement space of the hard disk On the other hand, the faster physical memory space is released for the programs that really need it.

4. Formatting

We know that the Linux operating system supports many different file systems, such as ext2, ext3, XFS, FAT, and so on, while Linux gives access to different file systems to VFS (virtual file system), and VFS can access and manage different file systems. So once you have the zone, you need to format it into a specific file system for VFS access.

The standard Linux file system Ext2 uses "inode-based file systems"

We know that in addition to the actual content of the file, the file data of the general operating system also has many attributes, such as file permissions (rwx) and file attributes (owner, group, time parameters, etc.) of the Linux operating system. The file system usually stores the attributes and the actual content in different chunks. In the inode-based file system, permissions and attributes are placed in inode. The actual data is placed in the data block block, and both inode and data block are numbered

Ext2 file system on this basis

There is a boot sector (boot sector) at the front of the file system, which allows us to install boot management programs. This design allows us to install different boot loaders to the front end of the individual file system without having to overwrite the unique MBR of the entire hard disk. Only in this way can we achieve the function of multi-boot and further divide each area into multiple blocks (block group). Each block group has an independent inode/block system if the file system up to hundreds of GB, put all the inode and block together because the number of inode and block is too large, not easy to manage this is actually very easy to understand, because the partition is the user's partition, the actual computer management there is a most suitable size, so the computer will further partition in the partition (but this may not lead to large files can not put the problem? Is there any mechanism to clean up the aftermath? Each block group is actually divided into six parts, with four subsidiary modules in addition to inode table and data block, which can optimize and improve the performance of the system.

So the whole partition will probably be divided like this:

Inode table mainly records the attributes of the file and which block the actual data of the file is placed in. It records at least this information: size, real content block number (one or more) access mode (read/write/excute) owner and group (owner/group) various times: the time of establishment or state change, the time of the last read, the time of the last modification, no file name! The file name is in the block of the directory! A file occupies an inode, each inode has a number Linux system, there is a situation where the inode number is used up, but the disk space is used up. Note, the file here is not just an ordinary file, the directory file, that is, a folder is actually a file, and other things, that is, the number and size of inode are fixed when formatted. Each inode size is fixed to 128bytes (the new ext4 and xfs can be set to 256bytes) the number of files that can be created by the file system is related to the number of inode. If there is enough space but not enough inode, the system needs to find the inode first when reading the files, and analyze whether the permissions recorded by the inode are consistent with the user. If it meets the requirements, it can actually start reading the contents of the block. Inode has a lot of data to record, but only 128bytes. While it takes 4byte for inode to record a block number, suppose I have a file with 400MB and each block is 4K, then I need at least 100, 000 block number records! How can inode have so much space to store? For this reason, our system cleverly defines the areas where inode records block numbers as 12 direct, one indirect, one double indirect and one three indirect recording areas (see appendix for details). When formatted, the size of the block is fixed, and each block is numbered to facilitate the recording of the inode. In principle, the size of the block is fixed. The size and number of block cannot be changed after formatting (unless reformatted) the block sizes supported in Ext2 file systems are 1K, 2K and 4K, due to the difference in block size. It will cause the maximum disk capacity that the file system can support is different from the maximum single file capacity: Block size 1KB 2KB 4KB maximum single file limit 16GB 256GB 2TB maximum file system total capacity 2TB 8TB 16TB each block can only hold the data of one file, but a file can be placed in multiple block (large) if the file is less than block Then the remaining capacity of the block can no longer be used (disk space will be wasted), so if your files are very small, but your block uses the largest 4K when formatting, it may result in a waste of capacity. Since a large block may cause a serious waste of disk capacity, shall we set the block size at 1K? This is also inappropriate, because if the block is smaller, then large files will occupy more block, and inode will also have to record more block numbers, which may lead to poor read and write performance of the file system. In fact, the disk capacity is too large, so it is common to choose a 4K block size superblock to record information about the entire file system. The size is generally 1024bytes. The main information recorded is: the total amount of block and inode is unused and the number of inode / block used is a valid bit value. If the file system has been mounted, the valid bit is 0. If it is not mounted, the valid bit is the size of 1 block and inode (block is 1,2,4K 128bytes or 256bytes). Other information about the file system: the mount time of the filesystem, the time of the last write, the time of the last fsck verification Superblock is very important, without Superblock, there would be no file system, so if superblock dies, your file system may take a lot of time to save each block may contain superblock, but we also say that a file system should have only one superblock What was that all about? In fact, except for the superblock in the first block, the subsequent block does not necessarily contain superblock, and if it contains superblock, the superblock is mainly used as a backup of the first superblock in the block, so that you can rescue superblock. Filesystem Description file system description this section can describe the block number of the beginning and end of each block group. And explain which block number each section (superblock, bitmap, inodemap, data block) is between which block bitmap block comparison table if you want to add a file, which block should be used to record it? Of course, choose "empty block" to record. Then how do you know which block is empty? This has to be done through block bitmap, which records which block is empty, so our system can quickly find available space to record. Also, when you delete certain files, the block number originally occupied by those files has to be released, and the flag bit corresponding to the block number in block bitmap has to be changed to "unused." inode bitmap and block bitmap are similar functions. It's just that block bitmap records used and unused block numbers, while inode bitmap records used and unused inode numbers.

5. Mounting

After an area is formatted as a file system, it can be used by the Linux operating system, but at this time the Linux operating system can not find it, so we also need to "register" the file system into the file system of the Linux operating system, this operation is called "mount".

Mounting uses a directory as an entry point (similar to choosing a ready-made directory as a proxy) and places the file system under that directory, that is, entering the directory can read the contents of the file system. Similar to the entire file system is just a folder (directory) of the directory tree.

The directory of this entry point is called the mount point.

Since the most important thing for the entire Linux system is the root directory, the root directory must be mounted to a partition. And other directories can be mounted to different points according to the needs of users.

At this point, the construction process of the Linux file system is basically over. To sum up, the hard disk has been partitioned and formatted, and each area has become a file system. After mounting this file system, you can let the Linux operating system access the hard disk through VFS as if it were a normal folder. Here is a practical example of reading files in a directory tree to detail directory files and ordinary files.

6. The process of reading a directory tree

First, we need to know

Each file (whether a general file or a directory file) occupies an inode and allocates one or more block to the file according to the size of the file content. After creating a file, the complete information of the file is distributed in three places and two new files are generated: the file name is recorded in the block of the directory file where the file is located. There is no new file generation file attribute, permission information, block number to record the specific content is recorded in inode, inode is a newly generated file file specific memory record in block, block is a newly generated file because the record of the file name is in the block of the directory, "add / delete / rename file name" is related to the w permission of the directory

So in Linux/Unix, the file name is just an attribute of the file, alias or nickname, just to facilitate user memory and use, but the system does not need to use the file name to determine the location of the file, so the most intuitive advantage is that you can be used to rename the file, change the directory, or even put into the wastebasket, will not affect the use of the current file, which is unimaginable in Windows. For example, if you open a Word file and rename it, Windows will tell you that there is no way, close the file first! But there is no pressure in Mac, because Mac's operating system also uses the design of inode.

File creation proc

When creating a general file under ext2, ext2 assigns an inode and the number of block relative to the size of the file to the file

For example, suppose one of my block is 4 Kbytes, and I want to create a 100 KBytes file, then linux will allocate an inode and 25 block to store the file, but at the same time, note that since inode has only 12 direct points, there is an additional block to create the directory as the record of the block number

When a directory is created in the ext2 file system (that is, a new directory file is created), the file system assigns an inode and at least one block to the directory

Inode records the relevant permissions and attributes of the directory, and records the assigned block number, while block records two records automatically in the inode number block corresponding to the file name in this directory. Folder record, inode points to itself, and the other is.. Folder record, inode points to the parent folder the process of reading a file from the directory tree because the file name is recorded in the block of the directory, so when we want to read a file, we must go through the inode and block of the directory, and then we will be able to find the inode number of the file to be read, and finally read the information in the block of the correct file. Because the directory tree starts from the root directory, the operating system first finds the inode number of the mount point through the mount information, from which the inode content of the root directory is obtained, and the block information of the root directory is read according to the inode, and then the correct file is read down layer by layer.

For example, how does the system read the / etc/passwd file if I want to read it?

First take a look at this file and the information about the path folder:

$ll-di / / etc/ etc/passwd128 dr-xr-x rmurx. 17 root root 4096 May 4 17:56 / 33595521 drwxr-x rmurx. 131 root root 8192 Jun 17 00:20 / etc36628004-rw-r-- rmuri -. 1 root root 2092 Jun 17 00:20 / etc/passwd

So the reading process of the file is as follows:

/ inode: find the root directory inode with inode number 128through the information of the mount point, and the permission specified by inode allows us to read the block of the content of the block (with r and x) /: get the number of the block through the previous step and find the content with the inode number of the etc/ directory (33595521) etc/ 's inode: read inode 33595521 and know that it has the permission of r and x Therefore, you can read the block content of etc/, the block of etc/: get the block number after the previous step, and find that the content has the inode number of the passwd file (36628004) inode of passwd: read inode 36628004 and know that you have the permission of r, so you can read the block content of passwd, the block of passwd: finally, read out the data of the block content.

Appendix: boot process and hard disk master boot record

You can talk a little bit about the boot process and the hard disk master boot record (MBR, or master boot partition).

A normal operating computer will set up a boot hard disk on the BIOS, in fact, each hard disk can be used as a boot disk, the design of the hard disk itself provides this possibility, which starts from the first sector on the hard disk, this sector has the hard disk master boot record (Master boot record, MBR) and partition table (partition table), in which MBR occupies 446 bytes, while the partition table occupies 64 bytes.

There is a program BIOS,BIOS written to the motherboard on the computer motherboard, which is the first program that the computer system will execute actively after it is turned on. BIOS will analyze what storage devices are in the computer. Let's take the hard disk as an example. BIOS will get the hard disk that can be booted according to the user's settings, and go to the hard disk to read the MBR location of the first sector.

MBR this only 446 bytes hard disk capacity will put the most basic boot loader (Boot loader), its purpose is to load the operating system kernel files, because the boot loader is provided by the operating system during installation, so it will understand the file system format in the hard disk, so it will be able to read operating system kernel files. Then there is the work of the kernel file, which is known as the task of the operating system.

So to put it simply, the boot process is:

BIOS: boot active program, will identify the first bootable device MBR- boot loader: the first bootable device in the first sector of the main boot partition boot loader, can read operating system kernel files operating system kernel files: different operating systems about opening their own programs

From the above instructions, we will know that BIOS and MBR are functions supported by the hardware itself, and Boot loader in MBR is a program written by the operating system on MBR. Because MBR has only 446 bytes, this boot loader is very small and beautiful, and its main tasks are:

Provide menu: users can choose different boot items, this is also an important function of multi-boot to load the operating system kernel: directly point to the bootable program section to start the operating system to transfer the other loader: it is interesting to transfer the boot loading function to other loader, it is possible to have more than two boot loaders in your computer system. Isn't there only one MBR on our hard drive? This is true, but the boot loader can not only be installed in the MBR, but also can be installed in the boot sector of each partition (boot sector). This feature creates a "multi-boot" function (see Chapter 3, Section IV of Bird's book). Hard disk drive giant magnetoresistance (GMR) head: from micron to nanometer [J]. Physics, 2004, 33 (07): 0-0. In recent years, the most critical factor for the rapid growth of computer hard disk storage density is the spin valve nano-multilayer structure, that is, the application of giant magnetoresistance (GMR) reading sensor magnetic head. The giant magnetoresistance magnetic head reading sensor has been transformed from microelectronic devices to nanoelectronic devices. this process includes spintronics, material science, microelectronic engineering, chemistry, micromechanics and engineering and other disciplines and related micromachining technologies to challenge the limits. How the disk works reveals that most permanent or semi-permanent computer data is achieved by magnetizing a small piece of metal material on the disk. These magnetic maps can then be converted into raw data. The internal hardware structure and working principle of the mechanical hard disk are explained in detail. The easiest way to give sector numbering is the sequential numbering of lmagin2, 3, 4, 4, 5, 6, etc. If sectors are numbered sequentially around tracks, the controller rotates the disk too far beyond the interval between sectors (which is very small) during the processing of data from one sector. the next sector to be read or written by the controller has passed through the head, perhaps a considerable distance. In this case, the disk controller can only wait for the disk to rotate again for almost one week before the desired sector can reach under the head. It's a waste of time. Many years ago, an outstanding engineer at IBM came up with a brilliant idea: instead of numbering sectors sequentially, they used an interleave to number sectors. Other details of formatting vary from operating system to operating system. For example, the pre-windows 98 micro-soft operating system mainly uses the file system FAT (or FAT16), the later version of windows 2000 has the so-called NTFS file system, and the orthodox file system of Linux is Ext2 (Linux second extended file system, ext2fs). And by default, the windows operating system does not recognize Linux's Ext2. In traditional disk and file system applications, a partition can only be formatted into a file system, so we can say that a file system is a partition. However, due to the use of new technologies, such as LVM and Software disk Array (software raid), these technologies can format a partition into multiple file systems or compose multiple partitions into one file system, so at present, we are no longer talking about formatting for partitions. Usually we can call a mountable data a file system rather than a partition. Relationship between inode/block and file size (interesting)

Let's briefly analyze the relationship between EXT2's inode / block and file size. Inode has a lot of information to record, but there is only 128bytes, and it takes 4byte for inode to record a block number. Suppose I have a file with 400MB and each block is 4K, then I need at least 100, 000 block number records! How can inode have so much space to store? For this reason, our system cleverly defines the areas where inode records block numbers as 12 direct, one indirect, one double indirect and one three indirect recording areas. What is this? Let's draw the structure of inode.

The leftmost picture above is inode itself (128bytes). There are 12 comparisons that point directly to the block number. These 12 records can directly get the block number! As for the so-called indirect is to take another block as the recording area of the block number, if the file is too large, the indirect block will be used to record the number. As shown in the figure above, it is only indirectly using a block to record the extra number. By the same token, if the file continues to grow, it will take advantage of the so-called double indirect, the first block only points out where the block of the next record number is, and the actual record is in the second block. And so on, three indirectly is to use the third layer of block to record the number!

How many block can inode specify in this way? Let's illustrate it with a smaller 1K block, and you can specify the following:

12 direct points: because 12*1K=12K is a direct point, a total of 12 records can be recorded, so the total size is as shown above indirect: 256*1K=256K each record of block number takes 4bytes, so the size of 1K can record 256 records, so a file size that can be recorded indirectly is the same as above Double indirect: 2562561K=256 2K layer 1 block specifies 256layer 2, each layer 2 can specify 256numbers, so the total size is as large as above; third indirect: 256256256*1K=256 3K layer 1 block specifies 256layer 2, each layer 2 can specify layer 3, and each layer 3 can specify 256numbers, so the total size is the same as above Total: direct, indirect, double indirect, three indirect sum, get 12 + 256256 + 256256x256 (K) = 16GB at this time we know that when the file system formats block into 1K size, the maximum file that can hold is 16GB, compare the results of the file system restriction table can be found to be consistent! However, this method can not be used in the calculation of 2K and 4K block size, because the block larger than 2K will be limited by the Ext2 file system itself, so the calculated results will not quite match the file system size and disk read performance.

With regard to the efficiency of the use of the file system, when your file system is planned to be very large, such as 100GB, because the data on the disk always comes and goes, the files on the whole file system usually cannot be written together continuously (block numbers are not contiguous), but fill in the data into the unused block. If the block written in the file is really scattered, there will be the so-called problem of discretization of file data.

As mentioned earlier, although our ext2 has recorded all the block numbers recorded in the file at inode, so the data can be read at once, if the file is really too discrete, there will still be a problem of low reading efficiency. Because disk read heads still have to come and go frequently throughout the file system! If so, the problem can be solved by copying all the data in the entire file system, reformatting the file system, and then copying the data back to him.

In addition, if the file system is really too large, then when a file is recorded in the first and last block number of the file system, it will cause the mechanical arm of the disk to move too much. It will also cause a decline in the performance of data reading. And when the read head searches the entire file system, it will spend more time searching. Therefore, the planning of the partition is not the bigger the better, but really needs to be planned for your host use.

Everything in Linux is a file.

Various things in Linux, such as documents, directories (called folders in Mac OS X and Windows systems), keyboards, monitors, hard drives, removable media devices, printers, modems, virtual terminals, and input / output resources such as interprocess communication (IPC) and network communications are all byte streams defined in the file system space.

Everything can be thought of as a file, and the most obvious benefit is that you only need the same set of Linux tools, utilities, and API for the input / output resources listed above. You can use the same set of api (read, write) and tools (cat, redirection, plumbing) to handle most of the resources in unix.

The ultimate goal of designing a system is often to find the atomic operation, once the atomic operation is locked, the design work will become simple and orderly. "File" as an abstract concept, its atomic operation is very simple, only reading and writing, this is undoubtedly a very good model. Through this model, the design of API can be simplified, users can access any resources in a general way, and have their own corresponding middleware to adapt to the underlying layer.

Modern operating systems introduce files to solve the problem that information can be stored independently of processes for a long time. As logical units of process creation information, files can be used concurrently by multiple processes. In the UNIX system, the operating system designs a set of general API for text and image on disk, input devices such as mouse and keyboard, and network interaction, so that they can all use byte stream mode. In other words, everything in the UNIX system except the process is a file, and Linux maintains this feature. To facilitate file management, Linux also introduced the concept of directories (sometimes referred to as folders). Directories enable files to be classified and managed, and the introduction of directories makes Linux's file system form a hierarchical directory tree.

The above is all the contents of the article "sample Analysis of File Systems in Linux". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.