RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

DwarFS (file system)

Product
Date of the premiere of the system: November, 2020
Branches: Information technologies

2020: The announcement of DwarFS - the file system with a record compression ratio of data

In November, 2020 on the GitHub portal there were source codes of the DwarFS project within which the developer Marcus Holland-Moritz develops the file system of the same name with "very much a high compression ratio of data".

The developer gave several examples of effective data compression. So, when archiving an image as a part of which there were nearly 1140 installations of the Perl language (more than 1.9 million files and 330 thousand directories entered there), the DwarFS system could squeeze it from initial 47.65 Gbytes to 555.7 MB, having spent for it 15 minutes. For comparison, creation of an image in the SquashFS file system demanded 69 minutes, and at the exit the image contracted up to 4.7 GB.

DwarFS can be used, for example, for creation of Live-images and reduction of the size of archives with a large number of duplicates and the repeating data (for example, storages of images of virtual machines or a collection of different versions of programs).

On GitHub the source code DwarFS - the file system with a record compression ratio of data appeared

Besides, the main part of the operating system of similar devices which is not subject to change and at the expense of algorithms of DwarFS occupying a minimum of disk space can be stored in such file system.

DwarFS uses the FUSE mechanism and works in a user space. The code is written on With ++ and is distributed under license GPLv3.

The DwarFS file system can work in the mode "only for reading" (Read Only) that does not allow to change available or to add new files to the images created with its help.

To distinctive features of the product Markus Holland-Moritz carried the following:

  • Clustering of files on similarity to use a hash function;
  • The analysis of segmentation on blocks of the file system for the purpose of reduction of the size of the uncompressed file system;
  • High multithreaded implementation;
  • Optional experimental support of Lua for providing the configured functions of filtering and ordering.[1]

Notes