DISTRIBUTED SYSTEMS

Lecture four

middle ware
 Middle ware

• Refers to a layer of services or softwares added

between those of a network OS and the actual
applications.
• Also known as software in the middle.
• It is a Set of tools that provide a uniform means
and style of access to system resources across
all platforms
 Middleware
 Middleware runs between client /server processes.

• Enable programmers to build applications that look and feel

the same
• Enable programmers to use the same method to access data
 It Shields the client and application programmers
from the complexity of the communication protocol,
NOS functions and hardware configurations.
Software and software Layers in Distributed
Systems

Applications, services

Middleware

Operating system
Platform

Computer and network hardware

4
 Middle ware functions

1.Facilitate the implementation of distributed

applications
2. Attempt to hide the heterogeneity of the
underlying system architectures (both hardware
and software).
 Middle ware services
There are a number of services common to
many middle ware systems. These include:

• communication services
• Naming services
• Distributed transaction services
• Security services
 communication services

• All middle ware attempts to implement access

transparency by offering high level
communication facilities that hide the low- level
message passing through computer networks.

• E.g http protocol allow transparent fetching of

documents in the web
 Naming services

• Naming services allow entities to be shared and

looked up.

• This implemented by providing directory services

which are comparable to telephone books and
yellow pages

• The location of the entity is assumed to be fixed.

 Naming services

• E.g a url contains the name of the server where the

document to which the url refers is stored.
• if the document is moved to another server, its url
ceases to work
 Distributed transaction services
• An important property of a transaction is that it allows
multiple read and write operations to occur atomically.

• Atomically means that the transaction either

succeeds so that all its write operations are actually
performed or it fails leaving all referenced data
unaffected
 Distributed transaction services

• Distributed transaction services operate on data that

are spread across multiple machines.

• These services are offered in order to hide failures in

the system
 Security services
• Virtually all middleware systems provide Facilities
for security.

• For example facilities to provide identification and

authentication before a shared resource is
accessed
 Middle ware models

• most middleware is based on some model, or

paradigm,.
• This Models are used for two main reasons:
• 1.To describe distribution and communication.
• 2. To make development and integration of
distributed applications as simple as possible,

13
 Middle ware models

Middle ware Models include the following:

1.Remote Procedure Calls (RPCs)

2.Distributed objects.

3.Distributed documents

4.Distributed file systems

14
 1.Remote Procedure Calls (RPCs
• The idea behind RPC is to call code remotely as if
we were just calling a procedure

• involves hiding network work communication by

allowing a process to call a procedure whose
implementation is on a remote machine.

15
 1.Remote Procedure Calls (RPCs
• From the client's point of view, calling a service on
the server should look just like calling a local
function, and the RPC framework is the responsible
for serializing parameters to the service, making sure
the correct service is run, and fetching the result
back to the client.

16
 1.Remote Procedure Calls (RPCs)

• Parameters are transparently shipped to the remote

host where the procedure is subsequently executed,
after which the results are sent back to the caller.

• It appears as if the procedure call was executed

locally:

• The calling process remains unaware of the fact that

n/w communication took place, except perhaps for
some loss of performance.

17
1.Steps of a Remote Procedure Call (RPCs)

18
1.Steps of a Remote Procedure Call (RPCs)

1. Client procedure calls client stub in normal way

2. Client stub builds message, calls local OS

3. Client's OS sends message to remote OS

4. Remote OS gives message to server stub

5. Server stub unpacks parameters, calls server

19
 1.Steps of a Remote Procedure Call (RPCs)

6. Server does work, returns result to the stub

7. Server stub packs it in message, calls local OS

8. Server's OS sends message to client's OS

9. Client's OS gives message to client stub

10. Stub unpacks result, returns to client

20
 Remote procedure calls(RPC)

• Parameters are transparently shipped to the remote

host where the procedure is subsequently executed,
after which the results are sent back to the caller.

21
 Remote procedure calls(RPC)
Rpc provides the following services
1. access transparency-a call to remote procedure may
be similar to a local procedure

2.location transparency – The developer can refer to

the procedure by name, unaware of
Where exactly the remote procedure is located.

3. synchronization – since the process invoking the

RPC remains suspended (blocked) until the
Remote procedure is completed,just as a call to
Local procedure.
22
 Remote method invocation
• The only real difference between RPC and RMI is
that there is objects involved in RMI instead of
invoking functions through a proxy function and
we invoke methods through a proxy .

23
 Remote method invocation

– In this approach, local object

invokes(calls) methods of an object
residing on a remote computer

- invocation appears as if it was a local

method call

24
Remote method invocation

25
 Remote method invocation
• The objective of RMI is that each object implements
an interface that hides all the internal details of the
object from its users.

• An interface consists methods that the objects

Implements.

• The only thing that a process sees of an object

is its interface.

26
 Remote method invocation
• Each object is located on a single machine and it’s
interface available on other machines.

• A method invocation is simply transformed into a

message that is sent to the object.

• The object executes the requested method and

sends back the result (reply message) to the
invoking process.

27
 A Remote Object

• A remote object will advertise and implement a remote interface

• Remote invocation can only invoke methods in the remote interface

Example of an interface
A Java Server example
RMI System
 RMI System

 Stub and Skeleton Layer: intercepts calls

from client and redirects to Remote
Reference Layer
 They unmarshals the request and invokes the
corresponding method in the remote object.
these RMI components are generated automatically by an
interface compiler.
 Remote Reference Layer interprets
references to remote objects, knows what to
do with them. Passes messages to the
 Transport Layer: sends messages using
request-reply protocol
 RMI System

• Proxy: the local representative of the

remote object.

• Dispatcher: relays a request to the

appropriate skeleton method.
 Remote method invocation

• The client hold references to remote objects that

it can invoke methods on.

• These references behave just like local objects,

but when invoked dispatch the method
invocation to the remote object.

34
 3. Distributed documents
• information is organized into documents, with
each document residing at a machine
transparently located somewhere in the world.
• Documents contain links that refer to other
documents.
• By following a link, the document to which that
link refers is fetched from its location and
displayed on the user’s screen.
• e.g world wide web

35
 4. distributed file systems
• A distributed file system is a subsystem of
distributed system that performs file management
activities where multiple users share files and
storage resources that physically dispersed.

• It is designed to provide abstraction of a storage

device.i.e convenient mechanism for storing and
retrieving information.

36
 DFS components

• Service – software entity running on one or more

machines and providing a particular type of
function to a priori unknown clients

• Server – service software running on a single

machine

• Client – process that can invoke a service using a

set of operations that forms its client interface
 DFS structure

DFS structure : There are two models

 4. Distributed file systems
• In the DFS model communication between
processes is done using shared files.

• Major difference other paradigms and the DFS

paradigm is that the resources (files) in DFS are
much longer lived.

• Clients access files and directories that are provided

by one or more file servers

39
 Distributed file systems

• A file server provides a client with a file service

interface and a view of the file system.

• Note that the view provided to different clients by the

same server may be different, for example, if clients
only see files that they are authorized to access.

40
 Distributed file systems

• Access to files is achieved by clients

performing operations (such as create,
delete, read, write, etc.) from the file service
interface on a file server.

• Depending on the implementation the

operations may be executed by the servers
on the actual files, or by the client on local
copies Of the file.

41
 Features of a good DFS
1.Flexible:
• It must be possible to add or replace file servers.

• A DFS should also support multiple underlying file

system types (e.g., various Unix file systems, various
Windows file systems, etc.)

42
 Features of a good DFS

1.provides transparency – e.g structure transparency,

access, naming, replication transparency

2. user mobility – user should have the flexibility to work

on different times

3. Scalability -a dfs should be designed to easily cope

with growth of nodes and users in the system
 Features of a good DFS

4. Simplicity and ease of use

The semantics of a distributed file System
should be easy to understand and the user interface
to the file system should be simple.

5. High performance:
Time of accessing secondary devices and cpu
processing time should be minimal.
Time needed to satisfy clients requests in dfs should
be comparable to that of a centralized file system.
 DFS services

Distributed file system typically provides

three types of services.
1.Storage service- allocating and managing
of space on a secondary storage device
that is used for storage of file system
2. True file service- concerned with the
operations on individual files such as
operations for accessing and modifying
data in files and creating and deleting files
 Distributed file systems services

Name service
 - provides a mapping between text names for file
and references to files that is file Ids

 - Most file systems use directories to perform

this mapping.
 File models

Different file systems use different conceptual

Models of a file.
There are two types of file models.

1. Unstructured files
2. structured files.
3. mutable files
4. immutable files
 1. Unstructured files
• In this model, the content of each file of the file
System appears to the file server as a uninterrupted
sequence of bytes

• The operating system is not interested in the

information stored in the files.

• It is only the application programs that interrupt

• The data stored in the files.

 structured files
• In this model, a file appears to the file server as an
ordered sequence of records.

• records of different files of the same file system can

be of different size.

• In this model, a record is the smallest unit of file data

can be accessed.
 Immutable files

• A file cannot be modified once it has been created exce

to be deleted.

• They are easy to support consistent sharing.

 Mutable files
• an update performed on a file overwrites its old contents
to produce the new contents
 Take away
• Discuss the following topics max (7pgs):

• Distributed Computing Environment (DCE)

• Distributed objects

• CORBA

• Process Migration

• Kerberos