Design Issues of DS

Distributed Systems
Major Design Issues

Introduction
Management of Distributed System mainly

consists of:
Coordination of concurrent distributed processes
Management and networking of distributed resources
Functioning of distributed algorithms
But….
network may be unreliable
Components may be untrusted
These raise the design and implementation
issues.
Introduction
 Following need to be considered for resolving
design and implementation issues:
 How objects in the system are identified
 How to co-ordinate the interaction among objects
 how they communicate with each other
 How can shared/replicated objects be managed in
controlled fashion
 Protection of objects and system security
DESIGN & IMPLEMENTATION ISSUES
Object Models and Naming Schemes

Distributed Coordination
Inter-process Communication
Distributed Resources
Fault Tolerance and Security
[1] OBJECT MODELS AND
NAMING SCHEMES
Objects in a computer system:
 processes, data files, memory, devices, processors, and
networks.
Objects are encapsulated in servers
process servers, file servers, memory servers etc.
A client is a null server that accesses object servers.
[1] Object Models and Naming Schemes
CONT’D
Three possible ways to identify a server
Identification by name (name server)
Identification by either physical or logical address
(network server)
Identification by service that the servers provide
Following all depend on the naming scheme

for system objects:
Structure of the system, management of name space, name
resolution, access methods
[2] DISTRIBUTED
COORDINATION
Processes require coordination to achieve
synchronization
Types of synchronization requirement
Barrier synchronization
Condition coordination
Mutual exclusion
TYPES OF
SYNCHRONIZATION
Barrier synchronization
 Process must reach a common synchronization point before they
can continue
Condition coordination
 A process must wait for a condition that will be set asynchronously
by other interacting processes to maintain some ordering of
execution
Mutual exclusion
 Concurrent processes must have mutual exclusion when accessing a
critical shared resource
[2] Distributed Coordination
Process deadlock problem related to

synchronization
Deadlock detection and recovery tool needed
Four conditions must hold for deadlock to occur
Exclusive use
Hold and wait
No preemption
Cyclical wait
The problem of deadlocks can be handled in

following ways
Prevention
Ensure that deadlock is not possible
 Avoidance
require decisions by the system while it is running in
order to insure that deadlocks will not occur
Detection
When detected, decide which process to rollback or
abnormally terminate
If one of the four conditions is prevented, it will

prevent deadlocks
For example, to impose an order on the resources
and require processes to request resources in
increasing order. This prevents cyclical wait and
thus makes deadlocks impossible
[3] INTER-PROCESS
COMMUNICATION
Inter-process communication can be
accomplished by using simple message passing
primitives
Higher level logical communication methods
provides the transparency
Hide the physical details of message passing
Two important concepts
The client/server model
Remote Procedure Call (RPC)
The client/ server model is a programming
example for structuring processes in distributed
systems
logical communication
request
client server
reply
actual communication
kernel network kernel
Remote Procedure Call (RPC) is a powerful technique
for constructing distributed, client-server based
applications.
It is based on extending the conventional local procedure
calling so that the called procedure need not exist in
the same address space as the calling procedure.
The two processes may be on the same system, or they
may be on different systems with a network connecting
them.
[3] Interprocess Communication
The Remote Procedure Call (RPC) model is

similar to that of the local model
The caller places arguments to a procedure in a specific
location (such as a result register)
The caller temporarily transfers control to the procedure
When the caller gains control again, it obtains the results
of the procedure from the specified location.
The caller then continues program execution.
[4] DISTRIBUTED RESOURCES
Data and Processing Capacity

Load Distribution
 multiprocessor scheduling (Static)
load distribution/sharing (Dynamic)
[4] DISTRIBUTED RESOURCES
Distributed shared memory

Distributed file systems
Issues: Sharing and Replication of data

Requires to maintain: data consistency and
coherency
Difference in implementation: distributed file
systems and distributed shared memory
[5] FAULT TOLERANCE AND SECURITY
Distributed Systems have openness in

operating environment
So vulnerable to failures and security threats
Faults: Failure and Security Violation
[5] FAULT TOLERANCE AND SECURITY
The problem of failures can be alleviated

through: redundancy
Transparent handling of failures (like removal of
machines, network links, and other resources)
without loss of data or functionality
Roll-back recovery for execution states
[5] Fault Tolerance and Security
OS view: trustworthy communication process,

confidentiality and integrity of messages and data
Security
 Authentication: clients and also servers and
messages must be authenticated.
Authorization: access control has to be performed
across a physical network with heterogeneous
components under different administrative units
using different security models.

Design Issues of DS

Uploaded by

Copyright:

Available Formats

Design Issues of DS

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design Issues of DS

Uploaded by

Copyright:

Available Formats

Distributed Systems

Major Design Issues

Management of Distributed System mainly

Object Models and Naming Schemes

Following all depend on the naming scheme

Process deadlock problem related to

The problem of deadlocks can be handled in

If one of the four conditions is prevented, it will

The Remote Procedure Call (RPC) model is

Data and Processing Capacity

Distributed shared memory

Issues: Sharing and Replication of data

Distributed Systems have openness in

The problem of failures can be alleviated

OS view: trustworthy communication process,

You might also like