Messages in this thread |  | | | Date | Sat, 15 May 1999 20:45:06 +0200 | | From | Manfred Spraul <> | | Subject | ext2 question |
| |
ext2 uses an array (direct, double indirect, tripple indirect) to store the blocks that belong to a file. Every array field stores 1 block, i.e. additional disk accesses are required if the file is > 12 kB even if the file is not fragmented. Most other modern filesystems (Be FS, Hfs, Hfs Extended, NTFS, HPFS) use a btree/list/compressed list of (start+len) to store this information. Are there any plans to support to a (start+len) system?
Perhaps we could replace the 12 direct blocks with 8 direct extents (4 byte start + 2 byte len), while keeping the current system for the contents of the indirect blocks: This would ensure that for 99,99% of all files no indirect blocks are required, but we would avoid a long linear search for highly fragmented/very long files.
-- Manfred
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
|  |