UNIX File Management
POS 355
Oct-29-2012

UNIX File Management
UNIX file systems consist of six types of files. Normal, directory, special, named pipes, links, and symbolic links. Today’s UNIX Operating Systems support multiple file systems in which the files map into the same underlying system for supporting file systems and allocating disk space. The information for each file resides in an index node (inode). Before a user can access a UNIX file the user must be granted the correct permissions (Stallings, 2012).
UNIX Files
A normal UNIX file is nothing more than data the user or a program has entered into the file. An ordinary UNIX file does not contain any internal structure from the file system; instead it consists of random data in zero or more data blocks. The next UNIX file is called a directory, which are hierarchically organized that contain files as well as other directories. Directories are normal that are setup with special privileges allowing only the file system to write to them. The third UNIX file is called special, these files contain no data, but they provide a method to map physical devices to file names. The next UNIX files are called named pipes. A named pipe is an interprocess communications tool that buffers the data received so a process that reads from the pipe can receive the data on a first-in-first-out basis. The fifth UNIX file is an alternate file name for an existing file. The sixth UNIX file is called a symbolic link. The symbolic link is a data file that contains the name of the file in which it is linked to. Each of these types of files is administered by the OS through inodes (Stallings, 2012).
An inode contains important information on each file, which is needed by the OS. Each file is controlled by one inode, but more than one file name can be associated with a single inode; an active inode is associated only with one file. The information such as attributes, control information and permissions for each file is stored in the inode (Stallings, 2012).
Volume Structure
According to Stallings, a UNIX file system resides on a single logical disk or disk partition that is laid out in four elements; these elements are called boot block, superblock, inode table, and data blocks. The boot block contains the code required to boot the operating system, the superblock contains the attributes and information about the file system, such as inode table size and partition size; the inode table is a collection of inodes for each file, finally the data blocks are storage space available for data files and subdirectories (Stallings, 2012).
UNIX File Access Control
UNIX systems are at least based on the file access control scheme introduced with the early versions of UNIX, where every UNIX user is assigned a unique userid, a member of at least one group. When a file is created the file is designated as owned by the user or program that created the file, the file is also setup with the primary GID of the user or the program or the directory GID. The file is also set with default permissions that allow a user or program either read, read/write, write, or execute permissions. File permissions can be changed only by the owner of the file or a user with root access allowing users or programs permissions at the user level or group level. For example permissions of rw------- would only grant the owner of the file read and write permissions, users in the GID associated with the file and all others on the system would not have any permissions with the file; permissions of rwxrw-rw- would grant the owner of the file read, write, execute permissions, anyone assigned to the GID associated with the file would have read and write permissions; all other users on the system would have read and write permissions.
A system supporting 5,000 users where 4,990 users need to access one file can be managed securely in UNIX. Each user would be assigned a unique userid and the correct GID for the file needing to be accessed. The permissions assigned to the file would need to be set permissions allowing the users to read the file; the permissions would look something like these permissions: rw-rw-r--. The owner of the file could read and write to the file, users who need to have write permissions would be assigned the GID that was given to the file, all others would only have read permissions to the file. Setting up the file permissions in this fashion allows some users the ability to write and prevents all other users from the ability of updating the file.

Reference
Stallings, W. (2012). Operating Systems: Internals and Design Principles (7th ed.). New Jersey, NJ: Prentice Hall.