Enumerating metadata in file system directories
View Patent ↗One embodiment is a method that stores metadata about files separately from directories for the files in storage devices and represents the directories as a tree structure in which entries refer to separate storage blocks that include the metadata with each of the separate storage blocks storing metadata for only a single file.
1. A method, comprising:
storing metadata about files separately from directories for the files in storage devices in a cluster file system;
representing the directories as a tree structure in which entries refer to separate storage blocks that include the metadata with each of the separate storage blocks storing metadata for only a single file;
storing the metadata in inodes with each of the separate storage blocks including only one corresponding inode; and
enumerating the files in the directories while extracting some of the metadata about the files.
2. The method of claim 1 further comprising, traversing the tree structure to extract a time when at least one of the files was last modified.
3. The method of claim 1 further comprising, storing the metadata in the inodes with consecutive inode numbers being in contiguous storage blocks.
4. The method of claim 1 further comprising, storing the metadata in consecutive storage blocks to increase efficiency of traversing the directories to enumerate the files.
5. A non-transitory computer readable storage medium having instructions for causing a computer to execute a method, comprising:
storing metadata about files as inodes in storage blocks in a cluster file system with the inodes being arranged by increasing inode number in consecutive storage blocks, wherein each of the storage blocks includes only a single inode; and
enumerating the files in a tree structure and extracting some of the metadata about the files.
6. The non-transitory computer readable storage medium of claim 5 , wherein the instructions are to further cause the computer to arrange the files in the tree structure with the files containing reference numbers that refer to the storage blocks.
7. The non-transitory computer readable storage medium of claim 5 , wherein the instructions are to further cause the computer to traverse the tree structure to extract information from the metadata to determine a last time at least one of the files was changed.
8. The non-transitory computer readable storage medium of claim 5 , wherein the instructions are to further cause the computer to store new inodes in consecutive storage blocks to increase efficiency in traversing the tree structure and gathering metadata about the files.
9. A server, comprising:
a memory;
a processor coupled to the memory; and
a file system manager coupled to the processor, the file system manager to store metadata about files separately from directories for the files in storage devices and to represent the directories as a tree structure in which entries refer to separate storage blocks that include the metadata with each of the separate storage blocks storing metadata for only a single file, wherein the metadata is stored in inodes with each of the separate storage blocks including only one corresponding inode.
10. The server of claim 9 , wherein the storage devices are distributed in a cluster file system.
11. The server of claim 9 , wherein the metadata is stored in the inodes with consecutive inode numbers being in consecutive storage blocks.
12. The server of claim 9 , wherein the metadata is stored in consecutive storage blocks to increase performance of searching the directories to enumerate the files.