1.9 Types of Distributed System: Grid, Cluster, Cloud

TYPES OF DISTRIBUTED SYSTEM

Distributed systems can be classified into different types based on their design, architecture, and purpose.

1. Distributed Computing System

A distributed computing system is a network of multiple computers or nodes working together to achieve a common goal. 

This distributed system is used in performance computation which requires high computing.

a) Cluster Computing

A collection of connected computers that work together as a unit to perform operations together, functioning in a single system.

Clusters are generally connected quickly via local area networks & each node is running the same operating system.

Features :

Collection of similar workstations/PCs, closely connected by means of a high-speed LAN:

• Each node runs the same OS. 
• Homogeneous environment(computers using similar configurations and protocols 
• Can serve as a supercomputer
• Excellent for parallel programming

Examples: Linux-based Beowulf clusters, MOSIX (from Hebrew University).

Advantages 

High Performance

Easy to manage

Scalable

Expandability

    Availability

Flexibility

Cost effectiveness 

DISADVANTAGES

High cost

The problem in finding the fault

More space is needed

Increased infrastructure needed

APPLICATION

• In many web applications functionalities such as Security, Search Engines, Database servers, web servers, proxy, and email.
• Assist and help to solve complex computational problems
• Cluster computing can be used in weather modeling
• Earthquake, Nuclear, Simulation, and tornado forecast

2. Grid Computing:

 

Grid Computing is a computing infrastructure that combines computer resources spread over different geographical locations to achieve a common goal. It uses widely distributed computer resources to reach a common goal.

In Grid Computing ,each system can belong to a different administrative domain and can differ greatly in terms of hardware, software, and implementation network technology.

Features

1. Resource Sharing –
   Multiple heterogeneous resources (computers, storage, applications, data) are shared across different organizations and locations.
    
2. Heterogeneity –
   Grid computing integrates different types of resources (PCs, clusters, supercomputers, storage devices) running on different operating systems and architectures.
    
3. Geographical Distribution –
   Resources are located in different geographical locations but are connected through a network (usually the Internet).
    
4. High Performance & Parallel Processing –
   Tasks are divided into sub-tasks and executed simultaneously across multiple systems, improving speed and efficiency.
5. Collaboration –Promotes collaboration among organizations, research centers, and individuals by pooling resources.

Advantages 

• Can solve bigger and more complex problems in a shorter time frame. 
• Easier collaboration with other organizations and better use of existing equipment
• Existing hardware is used to the fullest.
• Collaboration with organizations made easier

• Examples: Drug discovery simulations, protein folding analysis, large-scale climate modeling.

3. Cloud Computing:

 

Cloud computing is the on-demand delivery of computing resources (like servers, storage, databases, networking, software, and AI) over the internet (“the cloud”) with pay-as-you-go pricing.

Instead of owning and maintaining physical data centers or servers, users can rent computing power and services from cloud providers like Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP).

Features

1. On-demand self-service – Users can get computing resources instantly.
    
2. Broad network access – Services are available over the internet from anywhere.
    
3. Resource pooling – Multiple users share resources (multi-tenancy).
    
4. Rapid elasticity – Resources can scale up or down automatically.
    
5. Measured service – Pay only for what you use.

Types of Cloud Service Models

1. IaaS (Infrastructure as a Service):
    
   • Virtual machines, storage, networks.
      
   • Example: AWS EC2, Google Compute Engine.
      
2. PaaS (Platform as a Service):
    
   • Developers build apps without managing servers.
      
   • Example: Google App Engine, Heroku.
      
3. SaaS (Software as a Service):
    
   • End users use software over the internet.
      
   • Example: Gmail, Microsoft 365, Dropbox.
      

Types of Cloud Deployment Models

1. Public Cloud – Services offered over the public internet (AWS, Azure).
    
2. Private Cloud – Exclusive to a single organization (banks, govt.).
    
3. Hybrid Cloud – Mix of public + private.
    
4. Community Cloud – Shared by multiple organizations with a common purpose.

• Examples: Web hosting, online storage, software as a service (SaaS), platform as a service (PaaS), infrastructure as a service (IaaS).

Here's a table summarizing the key differences:

Feature	Cluster Computing	Grid Computing	Cloud Computing
Architecture	Tightly coupled	Loosely coupled	On-demand via internet
Purpose	High-performance computing	Large-scale problem solving	Flexible, on-demand resources
Resource Sharing	Controlled, within cluster	Voluntary, geographically dispersed	Pooled, dynamic allocation
Centralization	High	Lower	Highly centralized
Examples	Scientific computing, rendering	Drug discovery simulations, climate modeling	Web hosting, data analytics, SaaS pen_spark

 

2. Distributed Information System 

A Distributed Information System is a system in which data, applications, and services are distributed across multiple computers connected via a network, but appear to the user as a single integrated system.

Features:

• Resource sharing (data, files, apps, computing power)

• Scalability (easy to add more nodes)

• Fault tolerance (system continues to work even if some nodes fail)

• Transparency (users don’t feel the system is distributed)

a) Distributed Transaction Processing Systems (DTPS)

A distributed transaction processing system manages transactions that span across multiple, distributed databases or applications. Ensures ACID properties (Atomicity, Consistency, Isolation, Durability) of transactions across different systems. Often uses a Transaction Manager and techniques like Two-Phase Commit (2PC) to coordinate between nodes.

Examples:

• Online banking system (fund transfer between accounts in different banks/databases).

• Airline reservation systems.

• E-commerce checkout (payment gateway + inventory + shipping system).

Advantages:

• Reliable multi-database operations.

• Data consistency across sites.

• Supports real-time applications.

Challenges:

• High communication overhead.

• Complexity in synchronization.

• Performance bottlenecks if not designed well.

 

b) Enterprise Application Integration (EAI)

EAI refers to the process of integrating different enterprise applications (ERP, CRM, SCM, HR, etc.) into a single unified system so that data and processes can flow seamlessly. Provides a way for disparate applications to communicate and share data without replacing them.

Approaches:

1. Point-to-Point Integration (direct connections between apps).

2. Hub-and-Spoke (a central hub controls communication).

3. Message Bus / Middleware (e.g., ESB – Enterprise Service Bus).

4. Service-Oriented Architecture (SOA) / APIs (modern approach).

Examples:

• Integrating SAP ERP with Salesforce CRM.

• Linking e-procurement systems with financial systems in universities.

• Linking hospital management system with lab and pharmacy systems.

Advantages:

• Eliminates data silos.

• Increases efficiency and automation.

• Improves decision-making through integrated data view.

Challenges:

• Complex to implement.

• Expensive for large organizations.

• Security and data governance issues.

3. Distributed Pervasive System

A Distributed Pervasive System is a type of distributed system where computing is embedded into the environment and becomes part of everyday life. The goal is to provide anytime, anywhere computing by integrating different devices, sensors, and applications in a seamless manner.

It is based on the concept of Pervasive Computing (or Ubiquitous Computing) — proposed by Mark Weiser — where computing devices “disappear” into the background and users interact naturally without realizing they are using a computer.

Key Features

1. Ubiquity – Devices and services are available everywhere (smartphones, sensors, IoT devices).

2. Context-awareness – The system adapts to user needs and environment (location, time, activity).

3. Autonomy – Devices can operate and make decisions with minimal human intervention.

4. Interoperability – Multiple heterogeneous devices communicate and cooperate.

5. Scalability – Supports a large number of devices and users.

Examples

• Smart Home Systems: Lights, thermostats, and appliances controlled via IoT.

• Smart Health Monitoring: Wearable devices transmitting health data to doctors.

• Smart Cities: Traffic control, pollution monitoring, and surveillance integrated in real-time.

• Ubiquitous Learning (u-learning): Learning resources available on-demand anytime, anywhere.

• Smart Transportation: Self-driving cars and vehicle-to-vehicle communication.