Method and apparatus for distributed processing of file
View Patent ↗A method and apparatus for the distributed processing of a file are disclosed. The apparatus includes a shared data block selection unit, a modified data block selection unit, a first file distributed-processing unit, and a second file distributed-processing unit. The shared data block selection unit selects at least one first data block to remain without change after distributed processing of a file. The modified data block selection unit selects at least one second data block to be modified after the distributed processing of a file. The first file distributed-processing unit allows an inode after the distributed processing of a file to point to the first data blocks so that the first data block is shared before and after the distributed processing of a file. The second file distributed-processing unit allows the inode after the distributed processing of a file to point to at least one third data block.
1. A computer-implemented method for distributed processing of a file stored in a memory space, comprising:
selecting at least one first data block stored in the memory space to remain without change after distributed processing of a file;
selecting at least one second data block stored in the memory space to be modified after the distributed processing of a file;
pointing, by an inode after the distributed processing of a file, to the first data block so that the first data block is shared before and after the distributed processing of a file; and
pointing, by the inode after the distributed processing of a file, to at least one third data block stored in the memory space and modified from the second data block,
wherein the distributed processing of a file corresponds to any one of file segmentation, file merging, and addition into a front of a file, and
wherein if the distributed processing of a file corresponds to the addition into the front of a file, the second data block is a first data block of a file before the distributed processing of a file and there is an empty space corresponding to a size of new data to be added to the second data block, the third data block corresponds to a data block that is obtained by moving existing data of the second data block backward and copying the new data in front of the existing data.
2. The method of claim 1 , wherein if the distributed processing of a file corresponds to the file segmentation, the third data block corresponds to two data blocks copied from two block fragments of the second data block that has been internally fragmented after the file segmentation.
3. The method of claim 1 , wherein if the distributed processing of a file corresponds to the file merging and a sum of sizes of data fragments of the internally fragmented second data block is equal to or smaller than a size of one first data block, the at least third data block is one in number.
4. The method of claim 1 , wherein if the distributed processing of a file corresponds to the addition into the front of a file, the second data block is a first data block of a file before the distributed processing of a file and there is no empty space corresponding to a size of new data to be added to the second data block, the third data block becomes the index block, and an inode of the index block sequentially points to a data block including the new data and the first data block.
5. A computer system for distributed processing of a file stored in a memory space, having a processor, comprising:
a shared data block selection unit configured to select at least one first data block stored in the memory space to remain without change after distributed processing of a file;
a modified data block selection unit configured to select at least one second data block stored in the memory space to be modified after the distributed processing of a file;
a first file distributed-processing unit configured to allow an inode after the distributed processing of a file to point to the first data blocks so that the first data block is shared before and after the distributed processing of a file; and
a second file distributed-processing unit configured to allow the inode after the distributed processing of a file to point to at least one third data block stored in the memory space and modified from the second data block,
wherein the distributed processing of a file corresponds to any one of file segmentation, file merging, and addition into a front of a file, and
wherein if the distributed processing of a file corresponds to the addition into the front of a file, the second data block is a first data block of a file before the distributed processing of a file and there is an empty space corresponding to a size of new data to be added to the second data block, the third data block corresponds to a data block that is obtained by moving existing data of the second data block backward and copying the new data in front of the existing data.
6. The apparatus of claim 5 , wherein if the distributed processing of a file corresponds to the file segmentation, the third data block corresponds to two data blocks copied from two block fragments of the second data block that has been internally fragmented after the file segmentation.
7. The apparatus of 5 , wherein if the distributed processing of a file is the file merging and a sum of sizes of data fragments of the internally fragmented second data block is equal to or smaller than a size of one first data block, the at least third data block is one in number.
8. The apparatus of claim 5 , wherein if the distributed processing of a file corresponds to the addition into the front of a file, the second data block is a first data block of a file before the distributed processing of a file and there is no empty space corresponding to a size of new data to be added to the second data block, the third data block becomes the index block, and an mode of the index block sequentially points to a data block including the new data and the first data block.