### Overview of HPFS
The HPFS file system was first introduced with OS/2 1.2, aiming to enhance the ability to access larger hard disks emerging in the market at that time. Additionally, a new file system was needed then that could expand the naming system, organization, and security to meet the growing demands of the network server market. HPFS retained the directory organization of FAT while adding automatic directory sorting based on file names. File names were extended to up to 254 double-byte characters. HPFS also allowed files to be composed of "data" and special attributes, thereby increasing flexibility in supporting other naming rules and security. Moreover, the allocation unit was changed from clusters to physical sectors (512 bytes), which reduced the waste of disk space.
Under HPFS, directory entries contain more information than under FAT. As for attribute files, they also include information about modification, creation, access dates, and times. Under HPFS, directory entries do not point to the first cluster of a file but to an FNODE. An FNODE can contain the file's data, pointers to the file's data, or other structures that ultimately point to the file's data.
HPFS attempts to allocate a file in as many consecutive sectors as possible. This is done to improve the speed of sequential file processing.
HPFS organizes a drive into a series of 8 MB bands, and files are included in one of these bands whenever possible. Between bands are 2K allocation bitmaps to track which sectors within a band are allocated and which are not. Banding can improve performance because the drive head does not have to return to the logical top of the disk (usually cylinder 0), but only needs to return to the nearest band allocation bitmap to determine where the file is stored.
In addition, HPFS also includes two unique special data objects