FileSystem Interface

1
File-System Interface

• Chapter 11

2
Chapter 11 File-System Interface

• 11.1 File Concept
• 11.2 Access Methods
• 11.3 Directory Structure
• 11.4 File System Mounting
• 11.5 File Sharing
• 11.6 Protection

3
Chapter 11 Summary

• File Concept File definition, structure,
  attributes
• Access Methods direct, sequential access
• Directory Structure levels, tree, acyclic
• File System Mounting multi-partition directory
• File Sharing permissions, client-server, DNS
• Protection access levels, access control

4
File Concept

• Contiguous logical address space
• File Types
• Data
• numeric
• character
• binary
• Program
• A file is a sequence of bit, bytes, lines or
  records, the meaning of which is defined by the
  files creator and user.

11.1
5
File Structure

• None - sequence of words, bytes
• Simple record structure
• Lines
• Fixed length
• Variable length
• Complex Structures
• Formatted document
• Relocatable load file
• Can simulate structures by inserting appropriate
  control characters.

11.1
6
File Structure

• Magic number

Text size
Data size
Module name
bss size
Date
Header
Symbol table size
Owner
Entry point
Protection
. . .
Size
Flags
Text
Archive
Data
Relocation bits
Symbol table
Executable
7
File Attributes

• Name only information kept in human-readable
  form.
• Type needed for systems that support different
  types.
• Location pointer to file location on device.
• Size current file size.
• Protection controls who can do reading,
  writing, executing.
• Time, date, and user identification data for
  protection, security, and usage monitoring.
• Information about files is kept in the directory
  structure, which is maintained on the disk.
• File name, unique file identifier

11.1.1
8
File Operations

• Create space in file system, entry in
  directory
• Write search directory, write, update write
  pointer
• Read search directory, read, update read
  pointer
• Seek reposition pointer within a random access
  file
• Delete release file space, erase directory
  entry
• Truncate erase contents, keep attributes
• Open place entry in open file table, with file
  attributes
• Close erase entry from open file table
• Append write data to the end of a file
• Rename change name of an existing file
• Copy copy file to another directory or I/O
  device

11.1.2
9
Open-File Tables

• Per process table for each process
• File pointer position within file
• Access rights
• System-wide table entry made for any file open
• Location of file on disk
• Access date
• File size
• Open count number of processes accessing file
• File open count re-use open file entry
• (Tracks opens and closes, reaches zero on last
  close)

11.1.2
10
File Types Name, Extension
Win32 uses extensions UNIX extensions for apps
only
11.1.3, Figure 11.1
11
Access Methods

• Direct Access
• read n
• write n
• position to n
• read next
• write next
• rewrite n
• where n relative block number (an index
  relative to the beginning of the file).

Used for databases
11.2.2
12
Sequential-access File

• Sequential Access
• read next
• write next
• reset
• skip n

Used for editors and compilers
11.2.1, Figure 11.2
13
Simulation of Sequential Access on a
Direct-access File
where cp current position
11.2.2, Figure 11.3
14
Index and Relative Files
Index contains pointers to various blocks to
find a record, search the index, use pointer to
access file
11.2.3, Figure 11.4
15
Directory Structure

• A collection of nodes containing information
  about all files.

Directory
Files
F 1
F 2
F 3
F 4
F n
Both the directory structure and the files reside
on disk. Backups of these two structures are kept
on tapes.
16
Typical File-system Organization
Partitions are like virtual disks.
11.3, Figure 11.5
17
Information in a Device Directory

• Name
• Type
• Address
• Current length
• Maximum length
• Date last accessed (for archival info)
• Date last updated (for dump information)
• Owner ID (who owns it)
• Protection information (who accesses)

18
Operations Performed on Directory

• Search for a file name pattern searches
• Create a file add to the directory
• Delete a file remove from the directory
• List a directory file names and attributes
• Rename a file
• Traverse the file system access every directory

11.3
19
Benefits of Organized Directory

• Efficiency locating a file quickly.
• Naming convenient to users.
• Two users can have same name for different files.
• The same file can have several different names.
• Grouping
• Logical grouping of files by properties,
• (e.g. all Java programs, all games, )

11.3.1
20
Single-Level Directory

• A single directory for all users.

Naming problem Grouping problem Useful for
small embedded systems
11.3.1, Figure 11.6
21
Two-Level Directory

• Separate directory for each user.

• Path name
• Can have the same file name for different user
• Efficient searching
• No grouping capability

Windows \usr\ast\mailbox UNIX /usr/ast/mailbox MU
LTICS gtusrgtastgtmailbox
11.3.2, Figure 11.7
22
Tree-Structured Directories
11.3.3, Figure 11.8
23
Tree-Structured Directories

• Efficient searching
• Grouping Capability
• Current directory (working directory)
• cd /spell/mail/prog
• type list

11.3.3
24
Tree-Structured Directories

• Absolute or relative path name
• Creating a new file is done in current directory.
• Delete a file
• rm ltfile-namegt
• Creating a new subdirectory is done in current
  directory.
• mkdir ltdir-namegt
• Example if in current directory /mail
• mkdir count

mail
prog
copy
prt
exp
count
Deleting mail ? deleting the entire subtree
rooted by mail.
11.3.3
25
Acyclic-Graph Directories

• Have shared subdirectories and files.

11.3.4, Figure 11.9
26
Acyclic-Graph Directories

• Implementing shared files and directories
• Use symbolic links, pointers to other directory
  entries
• (Links are indirect pointers)
• Or use duplicate entries (but consistency a
  problem)
• If a file is deleted, a link can be left dangling
• Search for links and de-allocate, but takes time
• Leave the link until an attempt to use it, then
  delete
• (but be careful, it could point to a new file)
  (Win32)
• Keep a record of all links and directory entries,
  delete file only when the file-reference list is
  empty
• (UNIX does this, but just keeps a count of links)

11.3.4
27
General Graph Directory
11.3.5, Figure 11.10
28
General Graph Directory

• How do we guarantee no cycles?
• Allow only links to files, not subdirectories.
• Garbage collection to find cycles, and
• delete file if the only reference is to itself.
• Every time a new link is added,
• use a cycle detection algorithm to
• determine whether it is cyclic.
• Or use algorithm to bypass cycle in directory
  traversals.

11.3.5
29
File System Mounting

• A file system must be mounted before it can be
  accessed.
• The directory structure can be built out of
  multiple partitions.
• An unmounted file system is attached to a file
  system (mounted) at a mount point.
• Typically, a mount point is an empty directory
• A system may disallow mounting over non-empy
  directories
• A system may allow the same file system to be
  mounted repeatedly, at different points
• Win32 discovers and mounts devices at boot time
• UNIX has both automatic and manual execution of
  mounting

11.4
30
Mounting Point Example
Unmounted Partition
Existing File System
11.5, Figure 11.11
31
Mounting Point Example
File System mounted over /users
11.5, Figure 11.12
32
Remote File Systems

• Evolution of Network file sharing
• File transfer protocol (ftp)
• Anonymous access
• Authenticated access
• Distributed file system (DFS)
• Remote directories visible
• World Wide Web
• Browser gains access to the remote files
• ftp with a wrapper accesses the files

11.5.2
33
Remote File Systems

• Client-server systems
• Clients are identified by network name or IP
  address
• Encryption algorithms counter spoofing
• Require a secure exchange of keys
• UNIX Network File System (NFS) is
• a common distributed file-sharing method.
• Allows many clients to connect to many servers
• Domain Name System (DNS) for Internet
• A type of distributed naming service
• Host name to network address translations

11.5.2
34
Remote File Systems

• Network Information Service (NIS) from Sun
• Centralized storage of user names, host names
• Insecure, clear text system
• Common Internet File System (CIFS)
• Network information plus user authentication
• Domains distributed naming structure (older)
• Active directory (Windows 2000 and later)
• Lightweight Directory Access Protocol (LDAP)
• Efforts to create industry agreement on this
  standard
• Secure single log-in across multiple computer
  types

11.5.2
35
Domain Name System

• DNS is a mechanism to resolve the host name into
  a host-id that describes the destination system
  to the networking hardware
• DNS specifies the naming structure of the hosts
• The system resolves addresses by examining the
  host name components in reverse order
• Each component has a name server that accepts a
  name and returns the address of the name server
  responsible for that name
• bob.cs.brown.edu turns into 128.148.31.100, after
  edu name server has returned server address for
  brown.edu, etc.

11.5.2.2
36
Failure Modes

• Local file systems
• Failure of the disk or network hardware
• Corruption of the directory structure
• Corruption of metadata (disk management info)
• (RAID systems have redundancy for secure data)
• (Redundant Arrays of Inexpensive Disks)
• Remote file systems
• Network interruptions
• Server crash
• Scheduled server downtimes

11.5.2.3
37
Failure Recovery

• DFS protocols allow for delay of operations
• So remote server can come back into operation
• State information preserved at client and server
• Export status of file
• Mounting status of file systems
• NFS implements a stateless DFS
• Assumes legitimacy of interrupted client
• Assumes remote mounting of the file system
• Locates appropriate file, completes operation

11.5.2.3
38
File Consistency

• Consistency Semantics
• determine when modifications of data by one user
  are observable by another
• Defined over a file session, a series of accesses
  between open and close operations
• UNIX
• A file has a single, shareable image
• Writes to an open file are immediately visible to
  other users of the file
• Movement of pointer to current location within a
  file is shared among all users of the file

11.5.3
39
File Consistency

• Andrew File System (AFS)
• Can span over 5000 workstations
• A file may have many different images at the same
  time
• Atomic read and write operations ensure
  consistency between local and remote data
• Writes are not immediately visible to other users
  
• File changes are seen by other users only after
  they close and reopen the file
• Results in rapid, concurrent access

11.5.3
40
Protection

• File owner/creator should be able to control
• what can be done
• by whom
• Types of controlled access
• Read
• Write
• Execute
• Append
• Delete
• List

11.6.1
41
Access Lists and Groups

• Associate with each file and directory an access
  list user name and access level
• Enables complex access methodology
• But access lists get long
• Condense the access list to three classifications
  of user
• Owner the one who created the file
• Group a set of users needing similar access
• Universe all other users
• May also have specific users in an access list

11.6.2
42
Access Lists and Groups

• Mode of access read, write, execute
• Three classes of users
• RWX
• a) owner access 7 ? 1 1 1 RWX
• b) group access 6 ? 1 1 0
• RWX
• c) public access 1 ? 0 0 1
• Manager creates group G comprised of some users
• For a particular file or subdirectory, define an
  appropriate access.

11.6.4
43
Access Lists and Groups
Access by Owner, Group, Public
Directory?
Owner Size Date File Name

• -rw-rw-r-- 1 staff 31200 Sep 3 intro.ps
• drwx------ 5 staff 512 Jul 8 private/
• drwxrwxr-x 2 staff 512 Jul 8 doc/
• drwxrwx--- 2 student 512 Aug 3 student-proj/
• -rw-r--r-- 1 staff 9423 Feb 24 program.c
• -rwxr-xr-x 1 staff 20471 Feb 24 program

Number of links
Attach a group to a file chgrp G game
11.6.4, Figure 11.13