Software Architecture - 1 September 10, 2015September 10, 2015September 10, 2015

Software Architecture Software Architecture - 1- 1

April 19, 2023April 19, 2023

All Architectural PatternsAll Architectural Patterns Data Flow ArchitectureData Flow Architecture

– Batch SequentialBatch Sequential– Pipe and FilterPipe and Filter– Process ControlProcess Control

Data Centered ArchitectureData Centered Architecture– RepositoryRepository– BlackboardBlackboard

Hierarchical ArchitectureHierarchical Architecture– Main-subroutineMain-subroutine– Master-slaveMaster-slave– LayeredLayered– Virtual MachinesVirtual Machines

All Architectural PatternsAll Architectural Patterns Implicit Asynchronous Communication Implicit Asynchronous Communication

ArchitectureArchitecture– Non-buffered event-based Implicit InvocationNon-buffered event-based Implicit Invocation– Buffered messaged-basedBuffered messaged-based

Interaction-Oriented ArchitectureInteraction-Oriented Architecture– Model-View-Controller (MVC)Model-View-Controller (MVC)– Presentation-Abstraction-Control (PAC)Presentation-Abstraction-Control (PAC)

Distributed ArchitectureDistributed Architecture– Client-serverClient-server– Broker Broker – Service-oriented architecture (SOA)Service-oriented architecture (SOA)

Component-based ArchitectureComponent-based Architecture

Architectural Patterns - 1Architectural Patterns - 1 Data Flow ArchitectureData Flow Architecture

– Batch SequentialBatch Sequential– Pipe and FilterPipe and Filter– Process ControlProcess Control

Data Centered ArchitectureData Centered Architecture– RepositoryRepository– BlackboardBlackboard

Hierarchical ArchitectureHierarchical Architecture– Main-subroutineMain-subroutine– Master-slaveMaster-slave– LayeredLayered– Virtual MachinesVirtual Machines

Data-Flow ArchitectureData-Flow Architecture OverviewOverview

– A series of transformation on successive sets of dataA series of transformation on successive sets of data– Data sets and operations are independent of each Data sets and operations are independent of each

other. other. Three subcategoriesThree subcategories

– Batch SequentialBatch Sequential– Pipe and FilterPipe and Filter– Process Control Process Control

Batch SequentialBatch Sequential OverviewOverview

– A transformation subsystem or module cannot start its A transformation subsystem or module cannot start its process until the previous module completes its process until the previous module completes its computationcomputation

– Data flow carries a batch of data as a whole from a Data flow carries a batch of data as a whole from a module to nextmodule to next

– series of transformation on successive sets of dataseries of transformation on successive sets of data– Data sets and operations are independent of each Data sets and operations are independent of each

other. other. ApplicationsApplications

– Business data processing in banking and utility billingBusiness data processing in banking and utility billing

Batch SequentialBatch Sequential

Validate Sort Update Report

data transformation

Data flow


222 U 111 I 333 D -123 U

Validate

222 U 111 I 333 D

sort

111 I 222 U 333 D

Update100 ----111 ----200 ----222 ----444 ----

100 ---200 ---222 ---333 ---444 ---

Generate Report

Sortedtransaction

Validated transaction

Updated Master file

Rejected transaction

-123 U

Master file

Reports

Transaction File


Using Unix Shell ScriptUsing Unix Shell Script(exec) validate trans validTrans invalids(exec) validate trans validTrans invalids

(exec) sort validTrans sortedTrans(exec) sort validTrans sortedTrans

(exec) update master sortedTrans(exec) update master sortedTrans

(exec) generateReports master report1 report2 (exec) generateReports master report1 report2

Batch SequentialBatch Sequential Applicable DomainsApplicable Domains

– Data are batchedData are batched– Intermediate files are sequential filesIntermediate files are sequential files– Each module reads input files and writes output files Each module reads input files and writes output files

BenefitsBenefits– Simple division on subsystemsSimple division on subsystems– Each subsystem can be stand-alone programEach subsystem can be stand-alone program

LimitationsLimitations– Implementation requires external controlImplementation requires external control– No interactive interfaceNo interactive interface– Concurrency not supportedConcurrency not supported

Pipe and FilterPipe and Filter OverviewOverview

– A system consists of data source, filters, pipes, and A system consists of data source, filters, pipes, and data sinksdata sinks

– Connections between components are data streamsConnections between components are data streams– Each data stream is a first-in-first-out bufferEach data stream is a first-in-first-out buffer– Each filter reads data from its input stream, processes Each filter reads data from its input stream, processes

it, and writes it over a pipe for next filter to process it, and writes it over a pipe for next filter to process – A filter processes data once received, does not wait A filter processes data once received, does not wait

until the whole data is receiveduntil the whole data is received– A filter only knows its connected pipes, does not know A filter only knows its connected pipes, does not know

what are at the other end of the pipewhat are at the other end of the pipe– Filters are independent of each otherFilters are independent of each other– Pipes and filters may run concurrently Pipes and filters may run concurrently

Pipe and FilterPipe and Filter Active filterActive filter

– Pulls in (read) data from upstream and pushes out Pulls in (read) data from upstream and pushes out (write) data to downstream(write) data to downstream

Passive filterPassive filter– It lets connected upstream pipes push (write) data to it It lets connected upstream pipes push (write) data to it

and lets downstream pipes to pull (read) data from itand lets downstream pipes to pull (read) data from it

Pipe and FilterPipe and Filter

data source:Data Source

filter1:Filter

pipe:pipe

filter2:Filter

data sink:Data Sink

R

write

W

read

Filter1

Data Source Filter1 Filter2 Data Sink

pipe

Filter2


H 1L 2

HL LO WUpper caseConversionFilter

orldMatch (H,L)Filter

Hello World

Sort & Count Filter


Pipe and FilterPipe and Filter Applicable DomainsApplicable Domains

– System can be broken into a series of steps over data System can be broken into a series of steps over data streamsstreams

– Data format in the streams is simple, stable and Data format in the streams is simple, stable and adaptableadaptable

– Significant amount can be pipelined for performanceSignificant amount can be pipelined for performance BenefitsBenefits

– Concurrency: all filters may operate at the same timeConcurrency: all filters may operate at the same time– Reusability: filters are easy to plug and playReusability: filters are easy to plug and play

LimitationsLimitations– No user interactionNo user interaction– Difficult to configure Difficult to configure – Concurrency not supportedConcurrency not supported

Data-Centered Data-Centered Architecture Architecture OverviewOverview

– A centralized data store is shared by all A centralized data store is shared by all surrounding software componentssurrounding software components

– The surrounding components are normally The surrounding components are normally independent each otherindependent each other

– Significant amount can be pipelined for Significant amount can be pipelined for performanceperformance

Two categories:Two categories:– RepositoryRepository– BlackboardBlackboard

RepositoryRepository OverviewOverview

– Data store is passiveData store is passive– Clients of the data store are activeClients of the data store are active– Supports interactive data processing Supports interactive data processing – Clients control the computation and flow of Clients control the computation and flow of

logiclogic

RepositoryRepository



RepositoryRepository ApplicationsApplications

– Suitable for large complex information systems where Suitable for large complex information systems where many software component clients need to access it in many software component clients need to access it in different waysdifferent ways..

– Data transactions to drive the control flow of computation.Data transactions to drive the control flow of computation. Benefits:Benefits:

– Easy to backup and restoreEasy to backup and restore..– Easy to add new software componentsEasy to add new software components– Reduce the overhead of transient dataReduce the overhead of transient data

Limitations:Limitations:– Centralized repository is vulnerable to failure compared to Centralized repository is vulnerable to failure compared to

distributed repository with data replication.distributed repository with data replication.– High dependency between the structure of the data store High dependency between the structure of the data store

and its agents. and its agents.

BlackboardBlackboard OverviewOverview

– Data store is activeData store is active– Clients are passiveClients are passive– Clients are also called knowledge sources, Clients are also called knowledge sources,

listeners, and subscriberslisteners, and subscribers Two partitionsTwo partitions

– Blackboard: store data (hypotheses and facts)Blackboard: store data (hypotheses and facts)– Knowledge sources: domain-specific Knowledge sources: domain-specific

knowledge is storedknowledge is stored– Controller: initiate the blackboard and Controller: initiate the blackboard and

knowledge sources and take overall knowledge sources and take overall supervisionsupervision

BlackboardBlackboard CollaborationCollaboration

– Knowledge sources register with the Knowledge sources register with the blackboard in advance in a blackboard in advance in a publish/subscribe fashionpublish/subscribe fashion

– Data changes in the blackboard trigger Data changes in the blackboard trigger one or more matched knowledge sources one or more matched knowledge sources to continue processingto continue processing

– Data changes may be caused by new Data changes may be caused by new deduced information or hypothesis results deduced information or hypothesis results by knowledge sourcesby knowledge sources

– Knowledge sources normally do not Knowledge sources normally do not interact with each otherinteract with each other

BlackboardBlackboard

BlackboardBlackboard

Blackboard – example 1Blackboard – example 1

Animal identification System(KBS). Animal identification System(KBS). – The knowledge is represented as production The knowledge is represented as production

rules. rules. R1: IF animal gives milk THEN animal is mammalR1: IF animal gives milk THEN animal is mammalR2: IF animal eats meat THEN animal is carnivoreR2: IF animal eats meat THEN animal is carnivoreR3: R3: IF animal is mammal AND animal is carnivore AND IF animal is mammal AND animal is carnivore AND

animal has tawny color AND animal has black stripes animal has tawny color AND animal has black stripes THEN animal is tigerTHEN animal is tiger

– A set of facts A set of facts F1: animal eats meatF1: animal eats meatF2: animal gives milkF2: animal gives milkF3: animal has black stripesF3: animal has black stripesF4: animal has tawny colorF4: animal has tawny color


Forward reasoning Forward reasoning – New knowledge is added to blackboard New knowledge is added to blackboard

N1: animal is carnivore (from R2 + F1)N1: animal is carnivore (from R2 + F1)N2: animal is mammal (from R1 + F2) IF animal gives milk N2: animal is mammal (from R1 + F2) IF animal gives milk

THEN animal is mammalTHEN animal is mammalN3: animal is tiger (R3 + N3 + N2 + F3 + F4)N3: animal is tiger (R3 + N3 + N2 + F3 + F4)


Travel Consulting System Travel Consulting System – Participating agentsParticipating agents

Airline, hotel reservation, auto rental, and attraction Airline, hotel reservation, auto rental, and attraction agents agents

– BlackboardBlackboard Budget, available time, locations, etcBudget, available time, locations, etc

– Clients of the systemClients of the system Clients fill out a initial travel form Clients fill out a initial travel form The system will respond with many optional plans The system will respond with many optional plans

for client to choosefor client to choose


Travel Consulting System Travel Consulting System – ProcessProcess

A client submits a requestA client submits a request The system stores all the data in the blackboardThe system stores all the data in the blackboard The blackboard makes a request to the air agentThe blackboard makes a request to the air agent Once air reservation data is returned and stored in Once air reservation data is returned and stored in

blackboard, the change triggers hotel, auto rental, blackboard, the change triggers hotel, auto rental, attraction agents for a travel plans under budge and attraction agents for a travel plans under budge and timetime

Client chooses one of the planClient chooses one of the plan The system triggers the billing processThe system triggers the billing process

BlackboardBlackboard ApplicationsApplications

– Suitable for open-ended and complex problems (AI)Suitable for open-ended and complex problems (AI)– The problem spans multiple disciplines, each of them The problem spans multiple disciplines, each of them

has complete different knowledge expertisehas complete different knowledge expertise– Partial, or approximate solution is acceptable to the Partial, or approximate solution is acceptable to the

problems.problems. Benefits:Benefits:

– Easy to add new or update existing knowledge source.Easy to add new or update existing knowledge source.– Concurrency: all knowledge sources can work in Concurrency: all knowledge sources can work in

parallelparallel– Reusability of knowledge source agents.Reusability of knowledge source agents.

BlackboardBlackboard Limitations:Limitations:

– Tight dependency between the blackboard and Tight dependency between the blackboard and knowledge source,knowledge source,

– Difficult to make a decision when to terminate Difficult to make a decision when to terminate reasoning, since only partial or approximated solutions reasoning, since only partial or approximated solutions are expectedare expected

– Synchronization of multiple agents is an issue. Synchronization of multiple agents is an issue. – Debugging and testing of the system is a challenge. Debugging and testing of the system is a challenge.

Hierarchical ArchitectureHierarchical Architecture

OverviewOverview– The software system is decomposed into logical modules The software system is decomposed into logical modules

(sub-systems) at different levels in the hierarchy. (sub-systems) at different levels in the hierarchy. – Modules at different levels are connected by explicit or Modules at different levels are connected by explicit or

implicit method invocations. implicit method invocations. – A lower level module provides services to its adjacent A lower level module provides services to its adjacent

upper level modulesupper level modules– Upper level modules invoke the methods or procedures Upper level modules invoke the methods or procedures

in lower level. in lower level. Four categoriesFour categories

– Main-subroutineMain-subroutine– Master-slaveMaster-slave– LayeredLayered– Virtual MachinesVirtual Machines

Main-subroutineMain-subroutine OverviewOverview

– The main-subroutine architecture has dominated The main-subroutine architecture has dominated the software design methodologies for a very long the software design methodologies for a very long time. time.

– Reuse the subroutines and have individual Reuse the subroutines and have individual subroutines developed independently. subroutines developed independently.

– Using this style, a software system is decomposed Using this style, a software system is decomposed into subroutines hierarchically refined according to into subroutines hierarchically refined according to the desired functionality of the system. the desired functionality of the system.

– Refinements are conducted vertically until the Refinements are conducted vertically until the decomposed subroutine is simple enough to have decomposed subroutine is simple enough to have its sole independent responsibility, and whose its sole independent responsibility, and whose functionality may be reused and shared by functionality may be reused and shared by multiple callers above.multiple callers above.

Main-subroutineMain-subroutine


Data Flow Diagram to Main-subroutineData Flow Diagram to Main-subroutine– Transform flow: incoming flow feeds data in Transform flow: incoming flow feeds data in

an external format, and the data is then an external format, and the data is then transformed to another format, and then the transformed to another format, and then the outgoing flow carries the data outoutgoing flow carries the data out

– Transaction flow: evaluates its incoming data, Transaction flow: evaluates its incoming data, and decided which path to follow among and decided which path to follow among many action paths.many action paths.

– A transform flow is mapped by a controlling A transform flow is mapped by a controlling module for incoming, transform and outgoing module for incoming, transform and outgoing information processing.information processing.

– The transaction node becomes a dispatcher The transaction node becomes a dispatcher control module that controls all subordinate control module that controls all subordinate action modulesaction modules



Applicable DomainsApplicable Domains– Data are batchedData are batched– Intermediate files are sequential filesIntermediate files are sequential files– Each module reads input files and writes output files Each module reads input files and writes output files

BenefitsBenefits– Easy to decompose the system based on the definition Easy to decompose the system based on the definition

of the tasks in a top-down refinement manner of the tasks in a top-down refinement manner LimitationsLimitations

– Globally shared data in classical main-subroutines Globally shared data in classical main-subroutines introduces vulnerabilities.introduces vulnerabilities.

– Tight coupling may cause more ripple effects of Tight coupling may cause more ripple effects of changes as compared to OO Design.changes as compared to OO Design.

Master-slaveMaster-slave

OverviewOverview– A variant of the main-subroutine architecture style that A variant of the main-subroutine architecture style that

supports fault tolerance and system reliability. supports fault tolerance and system reliability. – Slaves provide replicated services to the master, and Slaves provide replicated services to the master, and

the master selects a particular result among slaves by the master selects a particular result among slaves by certain selection strategy. certain selection strategy.

– The slaves may perform the same functional task by The slaves may perform the same functional task by different algorithms and methods or totally different different algorithms and methods or totally different functionality.functionality.



Applicable DomainsApplicable Domains– Software systems where reliability is critical.Software systems where reliability is critical.

Layered ArchitectureLayered Architecture OverviewOverview

– The system is decomposed into a number of higher The system is decomposed into a number of higher and lower layers in a hierarchyand lower layers in a hierarchy

– Each layer consists of a group of related classes Each layer consists of a group of related classes that are encapsulated in package, in a deployed that are encapsulated in package, in a deployed component, or as a group of subroutines in the component, or as a group of subroutines in the format of method library or header file. format of method library or header file.

– Also, each layer has its own sole responsibility in Also, each layer has its own sole responsibility in the system.the system.

– A request to layer i +1 invokes the services A request to layer i +1 invokes the services provided by the layer i via the interface of layer i. provided by the layer i via the interface of layer i.

– The response may go back to the layer i +1 if the The response may go back to the layer i +1 if the task is completed; otherwise layer i continually task is completed; otherwise layer i continually invokes services from the layer i -1 below. invokes services from the layer i -1 below.

Layered ArchitectureLayered Architecture



FTP

TCP

IP

Ethernet

FTP

TCP

IP

Ethernet

TCP protocol

IP protocol

Physical connection

Ethernet protocol

FTP protocol


Layered ArchitectureLayered Architecture Applicable DomainsApplicable Domains

– Any system that can be divided between the application specific Any system that can be divided between the application specific portions and platform specific portions .portions and platform specific portions .

– Applications that have clean divisions between core services, Applications that have clean divisions between core services, critical services, user interface services, etc.critical services, user interface services, etc.

BenefitsBenefits– Incremental development based on levels of abstraction.Incremental development based on levels of abstraction.– Enhanced independence of upper layer to lower layer as long as Enhanced independence of upper layer to lower layer as long as

their interfaces remain unchanged. their interfaces remain unchanged. – separation of the standard interface and its implementation.separation of the standard interface and its implementation.– Component-based technology may be used to implement the Component-based technology may be used to implement the

layered architecture; this makes the system much easier to allow layered architecture; this makes the system much easier to allow for plug-and-play of new components.for plug-and-play of new components.

– Promotion of portability: each layer can be an abstract machine Promotion of portability: each layer can be an abstract machine deployed independently. deployed independently.

– Easy to decompose the system based on the definition of the tasks Easy to decompose the system based on the definition of the tasks in a top-down refinement manner in a top-down refinement manner


LimitationsLimitations– Lower runtime performance since a client’s request or Lower runtime performance since a client’s request or

a response to client must go through potentially a response to client must go through potentially several layers. There are also performance concerns several layers. There are also performance concerns on overhead on the data marshaling and buffering by on overhead on the data marshaling and buffering by each layer.each layer.

– Many applications cannot fit this architectural design.Many applications cannot fit this architectural design.– Breach of interlayer communication may cause Breach of interlayer communication may cause

deadlocks and “bridging” may cause tight coupling.deadlocks and “bridging” may cause tight coupling.– Exceptions and error handling is an issue in the layered Exceptions and error handling is an issue in the layered

architecture, since faults in one layer must propagate architecture, since faults in one layer must propagate upwards to all calling layers. upwards to all calling layers.

Virtual MachineVirtual Machine

OverviewOverview– A virtual machine is built up on an existing A virtual machine is built up on an existing

system and provides a virtual abstraction, a system and provides a virtual abstraction, a set of attributes, and operations. set of attributes, and operations.

– In most cases, we find that a virtual machine In most cases, we find that a virtual machine separates a programming language or separates a programming language or application environment from a execution application environment from a execution platform. platform.

– Some people say that a virtual machine looks Some people say that a virtual machine looks like emulation software. like emulation software.



Virtual MachineVirtual Machine Applicable DomainsApplicable Domains

– Suitable for solving a problem by simulation or translation Suitable for solving a problem by simulation or translation if there is no direct solution.if there is no direct solution.

– Sample applications include interpreters of Sample applications include interpreters of microprogramming, XML processing, script command microprogramming, XML processing, script command language execution, rule-based system execution, language execution, rule-based system execution, Smalltalk and Java interpreter typed programming Smalltalk and Java interpreter typed programming languagelanguage

BenefitsBenefits– Portability and machine platform independency.Portability and machine platform independency.– Simplicity of software development.Simplicity of software development.– Simulation for disaster working model.Simulation for disaster working model.

LimitationsLimitations– Slow execution of the interpreter due to the interpreter Slow execution of the interpreter due to the interpreter

nature.nature.– Additional overhead due to the new layer.Additional overhead due to the new layer.

Documents

Software Architecture - 1 September 10, 2015September 10, 2015September 10, 2015