Linux file system issues, Linux file system

After booting up, find the bootloader according to the BIOS settings, start the bootloader, load the kernel image, and after the load is successful, start to mount the root file system, you can set the mount root file system. When you say boot, do you mean after the kernel is loaded successfully?

What is the file you want to mount Is it an image file Single?

You need to find the corresponding configuration file backwards.

Linux file types are divided into common files, D directory files, B block device files, C character device files (serial port devices), S socket files (special files Scoket), and L link files.

For example, in the following figure, d starts with a directory file, - starts with a normal file, and l starts with a linked file.

Files created with tool commands such as touch, cp, tar, echo, cat, etc., are ordinary files, and ordinary files are divided into the following three types:

2) Binary: executable file (command) in Linux

3) Data format file (a special kind of file data).

ls -ld can display all directory files.

It refers to the link through the inode, and multiple files in the system each point to the same inode, which is normally allowed, and the file in this case is called a hard link. In the actual production environment, users can use hard links to prevent accidental deletion of important data.

ln source filename link filename Create a hard-linked file.

ln -s source filename link filename to create a soft-linked file.

1) When deleting the source file, the hard-linked file is used normally, and the soft-linked file becomes invalid.

2) When a hard link is deleted, it has no effect on the source file and the soft link.

3) When deleting a soft link, it has no effect on the source file and the hard link.

4) Once you delete the source file and hard-linked file, the entire file will be deleted.

1) The directory is not a hard link file, but a soft link file can be created.

2) Soft links to directories are very useful in a production environment.

3) All directories have a hard link"."Directory hard links do not cross file systems.

Linux file extensions are mainly used to make it easier to distinguish between different files, and are not the same as Windows strongly typed extensions.

1、.tar . tgz .A zip type of file means a compressed file.

2、.sh indicates a shell script file.

3、.html .php .JSP web page file.

4、.conf configuration file for system services.

5、.rpm stands for rpm installation package file.

Each file has its own attributes, including the number of nodes, file types, permissions, owners, groups, last modified, size, file or directory names, and the number of hard links.

For example, d represents a directory file, rwxr-xr-x represents permissions, the first rwx represents the permissions of the file owner user, the second r-x represents the permissions of the group to which it belongs, and the third r-x represents the permissions of other users. 2 indicates the number of hard links, the first root indicates that the file owner is root, the second root indicates the group it belongs to, 42 indicates the file size, and may 27 2021 indicates the last modification time, which is the file name.

The number of the inodes in it needs to be seen by ls -i, and the string of numbers displayed on the far left is shown in the following figure

Neither of the first two seems to be true.

Let's talk about the relationship between mount points and devices (storage space).

A mount point is just a directory, or rather"Location"(equivalent to "place names", like Jiangsu, Hebei), just for the convenience of accessing the device;

Data (here only files) are stored on devices (here only block devices, such as hard disks and USB flash drives), not on mount points, and the device is the real storage space.

1.All devices are mounted in the root directory or its subdirectories, so the root directory is not equivalent to the C system disk, but to the entire system.

2 and 3Files are stored on the device (not the mount point), so if you mount multiple devices (usually disk partitions) separately to directories such as usr home var lib, then the device corresponding to those directories is "getting bigger", not the root directory is getting bigger.

4.Regarding Linux partitioning, there are many recommended solutions on the Internet.

Desktop system partitions are mainly considered for data security, and some partitions are usually formatted when reinstalling the system, and if you put data files in a separate partition, you don't need to delete these files when you reinstall the system.

btw: In order to be more rigorous, I have to be a little more verbose...

Linux, UNIX, AIX directory is just a file system node, the whole file system is like a tree, the directory is like a branch, these branches can be grafted with different tree species, that is, the root, boot, swap, home, user can be hung in different hard disk partitions, can also be placed on the same partition. See how you want to plan, generally open other partitions in the user directory one level below home.

Forget about Windows partitions, the Linux Unix file system is a tree structure that is the starting point for all.

Generally, the partition planned by Redhat by default is sufficient. There is LVM, which is easy to expand, and you basically don't have to worry about partitioning

This statement is wrong that Linux systems do not support FAT32 file systems

Whether you can read the file system fs depends on whether the kernel module is enabled and whether there is a corresponding module in the kernel module.

If the module is compiled directly into the kernel, it will definitely be able to read or read at a level that depends on how open the fat spec is.

This is not that mathematics does not care too much about the accurate interpretation of words, and practice is the only criterion for testing truth.

It says that it does not support Linux, but it does not affect the reading of data.

Be careful when installing OSD on Linux, as errors can damage the file system. Not only does the command execution fail, but the device also doesn't work properly.

The command execution failed and the log is as follows:

Practical impact. My osdnode1 device doesn't work, embodied in:

After the relevant commands are listed, the result will not be returned, and the IO error will always be reported if the command fails to execute, such as create

Solution. After the command is repaired by the following method: yes next all the way during the command execution is OKFSCK -t ext4 dev sda3 is successfully repaired.

The above is all the content of how to fix Linux file system failure, although file system damage and failure are more likely to occur, but the repair method is relatively simple.