Chapter 17: Distributed File

Systems

Operating System Concepts – 8th Edition, Silberschatz, Galvin and Gagne ©2009

Chapter
p 17: Distributed File Systems
y
„ Background
„ Naming and Transparency
„ Remote File Access
„ Stateful versus Stateless Service
„ File Replication
„ An Example:
p AFS

Operating System Concepts – 8th Edition 17.2 Silberschatz, Galvin and Gagne ©2009
 Chapter Objectives

„ To explain the naming mechanism that provides location

t
transparency and
d iindependence
d d
„ To describe the various methods for accessing distributed files
„ To contrast stateful and stateless distributed file servers
„ To show how replication of files on different machines in a
distributed file system is a useful redundancy for improving
availability
„ To introduce the Andrew file system (AFS) as an example of a
distributed file system

Operating System Concepts – 8th Edition 17.3 Silberschatz, Galvin and Gagne ©2009

Background

„ Distributed file system (DFS) – a distributed implementation of the

classical
l i l titime-sharing
h i modeld l off a fil
file system,
t where
h multiple
lti l users
share files and storage resources

„ A DFS manages set of dispersed storage devices

„ Overall storage
g space
p managed
g byy a DFS is composed
p of different,,
remotely located, smaller storage spaces

„ There is usually a correspondence between constituent storage

spaces and sets of files

Operating System Concepts – 8th Edition 17.4 Silberschatz, Galvin and Gagne ©2009
 DFS Structure

„ Service – software entity running on one or more machines and

providing
idi a particular
ti l ttype off ffunction
ti tto a priori
i i unknown
k clients
li t

„ Server – service software running on a single machine

„ Client – process that can invoke a service using a set of

p
operations that forms its client interface

„ A client interface for a file service is formed by a set of primitive file

operations (create
(create, delete
delete, read
read, write)

„ Client interface of a DFS should be transparent, i.e., not distinguish

between local and remote files

Operating System Concepts – 8th Edition 17.5 Silberschatz, Galvin and Gagne ©2009

Naming and Transparency

„ Naming – mapping between logical and physical objects

„ Multilevel mapping – abstraction of a file that hides the details of

how and where on the disk the file is actually stored

„ A transparent DFS hides the location where in the network the file
is stored

„ For a file being replicated in several sites, the mapping returns a

set of the locations of this file’s
file s replicas; both the existence of
multiple copies and their location are hidden

Operating System Concepts – 8th Edition 17.6 Silberschatz, Galvin and Gagne ©2009
 Naming Structures
„ Location transparency – file name does not reveal the file’s physical
storage location

„ Location independence
p – file name does not need to be changed
g
when the file’s physical storage location changes

Operating System Concepts – 8th Edition 17.7 Silberschatz, Galvin and Gagne ©2009

Naming Schemes — Three Main Approaches

„ Files named by combination of their host name and local name;

guarantees
t a unique
i systemwide
t id name

„ Attach remote directories to local directories, giving the appearance

of a coherent directory tree; only previously mounted remote
directories can be accessed transparently

„ Total integration of the component file systems

z A single global name structure spans all the files in the system
z If a server is
i unavailable,
il bl some arbitrary
bit sett off directories
di t i on
different machines also becomes unavailable

Operating System Concepts – 8th Edition 17.8 Silberschatz, Galvin and Gagne ©2009
 Remote File Access

„ Remote-service mechanism is one transfer approach

„ Reduce network traffic by retaining recently accessed disk blocks
in a cache, so that repeated accesses to the same information can
be handled locally

z If needed data not already cached, a copy of data is brought

from the server to the user
z A
Accesses are performed
f d on the
th cached
h d copy
z Files identified with one master copy residing at the server
machine, but copies of (parts of) the file are scattered in
different caches
z Cache-consistency problem – keeping the cached copies
consistent with the master file
 Could be called network virtual memory

Operating System Concepts – 8th Edition 17.9 Silberschatz, Galvin and Gagne ©2009

Cache Location – Disk vs. Main Memory

„ Advantages of disk caches

z More reliable
z Cached data kept on disk are still there during recovery and
don’tt need to be fetched again
don

„ Advantages of main-memory caches:

z Permit workstations to be diskless
z Data can be accessed more quickly
z Performance speedup in bigger memories
z Server caches (used to speed up disk I/O) are in main memory
regardless of where user caches are located; using main-
memory caches on the user machine permits a single caching
mechanism for servers and users

Operating System Concepts – 8th Edition 17.10 Silberschatz, Galvin and Gagne ©2009
 Cache Update Policy

„ Write-through – write data through to disk as soon as they are placed on

an cache
any
z Reliable, but poor performance

„ Delayed-write – modifications written to the cache and then written

through to the server later
z Write accesses complete quickly; some data may be overwritten
before they are written back, and so need never be written at all
z Poor reliability; unwritten data will be lost whenever a user machine
crashes
z Variation – scan cache at regular intervals and flush blocks that have
been modified since the last scan
z Variation – write-on-close, writes data back to the server when the file
is closed
 Best for files that are open for long periods and frequently modified

Operating System Concepts – 8th Edition 17.11 Silberschatz, Galvin and Gagne ©2009

Cachefs and its Use of Caching

Operating System Concepts – 8th Edition 17.12 Silberschatz, Galvin and Gagne ©2009
 Consistency

„ Is locally cached copy of the data consistent with the master copy?

„ Client-initiated approach
z Client initiates a validity check
z Server checks whether the local data are consistent with the
master copy

„ Server-initiated approach
z Server records, for each client, the (p
(parts of)) files it caches
z When server detects a potential inconsistency, it must react

Operating System Concepts – 8th Edition 17.13 Silberschatz, Galvin and Gagne ©2009

Comparing Caching and Remote Service

„ In caching, many remote accesses handled efficiently by the local

cache;
h mostt remote
t accesses will
ill b
be served
d as ffastt as llocall ones
„ Servers are contracted only occasionally in caching (rather than for
each access)
z Reduces server load and network traffic
z Enhances potential for scalability
„ Remote server method handles every remote access across the
network; penalty in network traffic, server load, and performance
„ Total network overhead in transmitting
g big
g chunks of data ((caching)
g)
is lower than a series of responses to specific requests (remote-
service)

Operating System Concepts – 8th Edition 17.14 Silberschatz, Galvin and Gagne ©2009
 Caching and Remote Service (Cont.)

„ Caching is superior in access patterns with infrequent writes

z With frequent writes, substantial overhead incurred to
overcome cache-consistency problem
„ Benefit from caching when execution carried out on machines with
either local disks or large main memories
„ Remote access on diskless, small-memory-capacity machines
should be done through remote
remote-service
service method
„ In caching, the lower intermachine interface is different form the
upper user interface
„ In remote-service, the intermachine interface mirrors the local user-
file-system interface

Operating System Concepts – 8th Edition 17.15 Silberschatz, Galvin and Gagne ©2009

Stateful File Service

„ Mechanism
z Client opens a file
z Server fetches information about the file from its disk, stores it
in its memory, and gives the client a connection identifier
unique
i tto th
the client
li t and
d th
the open fil
file
z Identifier is used for subsequent accesses until the session
ends
z Server must reclaim the main-memory space used by clients
who are no longer active

„ Increased performance
z Fewer disk accesses
z Stateful server knows if a file was opened for sequential access
and can thus read ahead the next blocks

Operating System Concepts – 8th Edition 17.16 Silberschatz, Galvin and Gagne ©2009
 Stateless File Server

„ Avoids state information by making each request self-contained

„ Each request identifies the file and position in the file

„ No need to establish and terminate a connection by open and close

operations

Operating System Concepts – 8th Edition 17.17 Silberschatz, Galvin and Gagne ©2009

Distinctions Between Stateful & Stateless Service

„ Failure Recovery
z A stateful server loses all its volatile state in a crash
 Restore state by recovery protocol based on a dialog with
clients, or abort operations that were underway when the
crash occurred
 Server needs to be aware of client failures in order to
reclaim space allocated to record the state of crashed client
processes (orphan detection and elimination)
z With stateless server, the effects of server failure sand
recovery are almost
l t unnoticeable
ti bl
 A newly reincarnated server can respond to a self-
contained request without any difficulty

Operating System Concepts – 8th Edition 17.18 Silberschatz, Galvin and Gagne ©2009
 Distinctions (Cont.)

„ Penalties for using the robust stateless service:

z longer request messages
z slower request processing
z additional constraints imposed on DFS design

„ Some environments require stateful service

z A server employing server-initiated cache validation cannot
provide stateless service, since it maintains a record of which
y which clients
files are cached by
z UNIX use of file descriptors and implicit offsets is inherently
stateful; servers must maintain tables to map the file
descriptors to inodes, and store the current offset within a file

Operating System Concepts – 8th Edition 17.19 Silberschatz, Galvin and Gagne ©2009

File Replication

„ Replicas of the same file reside on failure-independent machines

„ Improves availability and can shorten service time
„ Naming scheme maps a replicated file name to a particular replica
z Existence of replicas should be invisible to higher levels
z Replicas must be distinguished from one another by different
lower-level names
„ Updates – replicas of a file denote the same logical entity, and thus
an update to any replica must be reflected on all other replicas
„ Demand replication – reading a nonlocal replica causes it to be
cached locally, thereby generating a new nonprimary replica

Operating System Concepts – 8th Edition 17.20 Silberschatz, Galvin and Gagne ©2009
 An Example: AFS

„ A distributed computing environment (Andrew) under development

since
i 1983 att C
Carnegie-Mellon
i M ll U University,
i it purchased
h dbby IBM and
d
released as Transarc DFS, now open sourced as OpenAFS
„ AFS tries to solve complex issues such as uniform name space,
location-independent file sharing, client-side caching (with cache
consistency), secure authentication (via Kerberos)
z Also includes server-side caching
g ((via replicas),
p ), high
g
availability
z Can span 5,000 workstations

Operating System Concepts – 8th Edition 17.21 Silberschatz, Galvin and Gagne ©2009

ANDREW (Cont.)

„ Clients are presented with a partitioned space of file names: a

l
local
l name space and
d a shared
h d name space
„ Dedicated servers, called Vice, present the shared name space to
the clients as an homogeneous, identical, and location transparent
file hierarchy
„ The local name space is the root file system of a workstation, from
p
which the shared name space descends
„ Workstations run the Virtue protocol to communicate with Vice, and
are required to have local disks where they store their local name
space
„ Servers collectively are responsible for the storage and
management of the shared name space

Operating System Concepts – 8th Edition 17.22 Silberschatz, Galvin and Gagne ©2009
 ANDREW (Cont.)

„ Clients and servers are structured in clusters interconnected by a

b kb
backbone LAN

„ A cluster consists of a collection of workstations and a cluster

server and is connected to the backbone by a router

„ A key
y mechanism selected for remote file operations
p is whole file
caching
z Opening a file causes it to be cached, in its entirety, on the
local disk

Operating System Concepts – 8th Edition 17.23 Silberschatz, Galvin and Gagne ©2009

ANDREW Shared Name Space

„ Andrew’s volumes are small component units associated with the

fil off a single
files i l client
li t
„ A fid identifies a Vice file or directory - A fid is 96 bits long and has
three equal-length components:
z volume number
z vnode number – index into an array containing the inodes of
files in a single volume
z uniquifier – allows reuse of vnode numbers, thereby keeping
certain data structures, compact
„ Fids are location transparent; therefore, file movements from server
to server do not invalidate cached directory contents
„ Location information is kept on a volume basis, and the information
is replicated on each server

Operating System Concepts – 8th Edition 17.24 Silberschatz, Galvin and Gagne ©2009
 ANDREW File Operations

„ Andrew caches entire files form servers

z A client workstation interacts with Vice servers only during
opening and closing of files
„ Venus – caches files from Vice when they are opened, and stores
modified copies of files back when they are closed
„ Reading and writing bytes of a file are done by the kernel without
Venus intervention on the cached copy
„ Venus caches contents of directories and symbolic links, for path-
name translation
„ Exceptions to the caching policy are modifications to directories
that are made directly on the server responsibility for that directory

Operating System Concepts – 8th Edition 17.25 Silberschatz, Galvin and Gagne ©2009

ANDREW Implementation

„ Client processes are interfaced to a UNIX kernel with the usual set
off system
t calls
ll
„ Venus carries out path-name translation component by component
„ The UNIX file system is used as a low
low-level
level storage system for
both servers and clients
z The client cache is a local directory on the workstation’s disk
„ Both Venus and server processes access UNIX files directly by
their inodes to avoid the expensive path name-to-inode translation
routine

Operating System Concepts – 8th Edition 17.26 Silberschatz, Galvin and Gagne ©2009
 ANDREW Implementation (Cont.)

„ Venus manages two separate caches:

z one for status
z one for data
„ LRU algorithm used to keep each of them bounded in size
„ The status cache is kept in virtual memory to allow rapid servicing
of stat() (file status returning) system calls
„ The data cache is resident on the local disk, but the UNIX I/O
buffering mechanism does some caching of the disk blocks in
memory y that are transparent
p to Venus

Operating System Concepts – 8th Edition 17.27 Silberschatz, Galvin and Gagne ©2009

End of Chapter 17

Operating System Concepts – 8th Edition, Silberschatz, Galvin and Gagne ©2009
 Fig. 17.01

Operating System Concepts – 8th Edition 17.29 Silberschatz, Galvin and Gagne ©2009