File System Basics

File Concept

  • File System: A collection of files used to store data.
  • Directory Structure: Organizes information about files.
  • Files: A sequence of bits, bytes, lines, or records defined by its creator and user.
  • Data cannot be stored on secondary storage unless within a file.
  • Types of information: source programs, executables, text, images, sound recordings, etc.

File Attributes

  • File Attributes:

    • Name: Human-readable information.
    • Identifier: Unique tag that identifies a file.
    • Type: File classification; tells the system what to do with the file (e.g., JPG).
    • Location: Pointer to the file's location in the file system.
    • Size: Current file size (e.g., 826 KB).
    • Protection: Access control over file permissions (read, write, execute).
    • Time, date, and user info: Usage monitoring (e.g., last modified time).
      • Created: Tuesday, July 14, 2009, 06:32:38
      • Modified: Tuesday, July 14, 2009, 06:32:31
      • Accessed: Tuesday, July 14, 2009, 06:32:31

  • EXIF Meta-Data (Exchangeable image file format):

    • Contains date, time, physical/logical location, software version, owner, etc.
    • Examples:
      • Catching criminals using image meta-data.
      • Catching hackers using image meta-data.
      • Criminals exploiting vulnerabilities based on software versions.
      • Phishing attacks based on owner information.

File Operations

  • Basic Operations:
    • Create a file: Allocate space on disk and link to directory.
    • Deleting a file: Deallocate space and unlink from directory (in Windows, space may not be immediately deallocated).
    • Write and Read a file: Use a pointer to the current file position.
    • Reposition within a file: Change the current file position pointer.
    • Truncating a file: Erase file content but keep attributes.
  • Other Operations:
    • Rename file, append info, copy file, get/set attributes.

Opening Files

  • Open-file table: Tracks all opened files.
  • File pointer: Unique per process, points to last read/write location.
  • File-open count: Tracks number of open and close operations per file; when the counter is zero, the file entry is removed from the open-file table.
  • Disk location of file: Pointer to the disk location.
  • Access rights: Access mode per process.

File Locking

  • File lock: Processes can block access to a file or section of a file, useful when multiple processes access the same file.
  • Read/Write locks:
    • Shared lock: Multiple processes can acquire concurrently (reader lock).
    • Exclusive lock: Only one process can acquire (writer lock).
  • Other lock types:
    • Mandatory: Prevents access until the exclusive lock is released.
    • Advisory: Returns the lock status of a file, letting software developers decide.

Access Methods

  • Sequential Access:
    • Process information in order.
    • Operations: read next, write next, reset to file beginning.
    • Current pointer.
  • Direct Access:
    • File is split into blocks.
    • Operations: read i, write i.
    • Can link to sequential access: position to i, read next, write next.
  • Other Methods:
    • Extend direct access with an index to blocks.
    • Order blocks based on some method and use an index (e.g., books ordered by author's last name).
    • Similarity to non-contiguous memory allocation.

Directory and Disk Structure

  • Disks are split into partitions, and each partition has a directory.
  • Directory: Listing of files and their attributes.
  • Metadata
  • Index to File 1
  • Data for File 1

Directory Overview

  • Directory Structure: Collection of nodes containing information about all files.
  • Located on disk.
  • Operations: Create, delete, search, rename, list files.
  • Organized logically for efficiency, naming, grouping.

Directory Structure Types

  • Single-Leveled Directory:
    • All files in the same directory.
    • Advantage: simple and efficient.
    • Disadvantage: requires unique file names, can become very large.
  • Two-level Directory:
    • Separate directory per user.
    • Users cannot access other user spaces.
    • No grouping allowed.
  • Tree-Structured Directory:
    • Generalization of two-level, allowing grouping.
    • Root directory with groupings.
    • Permissions for each file and directory: read (r), write (w), execute (x).
    • Permissions per user: owner, group, others.
    • Caution: No Sharing
  • Acyclic Graph Directories:
    • Tree-structured with sharing allowed.
    • Multiple links to files: symbolic (logical) and hard links (physical).
  • General Graph Directory:
    • Tree-structured with sharing and cycles.
    • Cycles are a problem; cycle detection is needed on each new link.
    • Garbage Collection: Cycles lead to self-referencing

Key Concepts

  • The current directory may be included in the search path.
  • A relative path name defines a path from the current directory.
  • If the current directory is /home/user/jane, then the correct relative path to the file /home/user/mike/prog.c is ../mike/prog.c.