fadvise

Preload a file into cache

$ fadvise [[-a|--advice]] willneed [path/to/file]

$ fadvise [path/to/file]

$ fadvise [[-h|--help]]

fadvise file offset length advice

file
    The path to the file to which the advice applies.

offset
    The starting offset, in bytes, of the region to which the advice applies.

length
    The length, in bytes, of the region to which the advice applies. A length of 0 means from offset to the end of file.

advice
    The advice to provide to the kernel. Possible values include:

• normal: No special advice. This is the default behavior.
• sequential: Expect sequential access.
• random: Expect random access.
• willneed: Expect access in the near future.
• dontneed: Don't expect access in the near future.
• noreuse: Expect access only once.

The fadvise command is a Linux system call wrapper that allows a program to advise the kernel about the expected usage pattern of a file or file region. This advice can help the kernel optimize disk caching, read-ahead, and write-behind operations to improve I/O performance.
The advisory nature of fadvise means that the kernel is free to ignore the advice provided. However, when the advice is heeded, it can lead to significant performance gains, particularly for applications that perform large sequential reads or writes. The goal is to reduce latency and improve throughput by providing the kernel with information about how the file will be accessed.

The fadvise command will return an error if the specified file does not exist, if the user does not have permission to access the file, or if an invalid advice value is provided. The system call may also return an error if it encounters a problem during its execution.

Example: `fadvise large_file.dat 0 104857600 sequential` advises the kernel that the first 100MB of `large_file.dat` will be accessed sequentially. `fadvise data.txt 0 0 random` advises the kernel that file `data.txt` will be accessed randomly.

The fadvise system call and command-line utility were introduced to provide a mechanism for applications to communicate their expected file access patterns to the kernel. This became increasingly important as storage devices evolved and the kernel needed more information to optimize I/O operations. The command has been part of the Linux kernel for many years and continues to be used to improve performance in a variety of applications.

open(2), read(2), write(2), mmap(2)