Distributed Systems: Difference between revisions

Revision as of 07:49, 6 July 2025

Distributed Systems

A distributed system is a model in which components located on networked computers communicate and coordinate their actions by passing messages. The components interact with one another in order to achieve a common goal. Distributed systems can be categorized based on their architecture, networking topology, and consistency models, amongst other factors. They are increasingly important in computing, as they facilitate the development of applications that are more scalable, resilient, and accessible.

Introduction

Distributed systems are prevalent in modern computing and form the backbone of many major applications and services. They provide key advantages such as resource sharing, fault tolerance, scalability, and improved performance. In a distributed system, components located on multiple networked computers work together to perform tasks, effectively giving the appearance of a single coherent system to the user. The emergence of cloud computing, web services, and peer-to-peer systems has further propelled the relevance and use of distributed systems.

While distributed systems may seem similar to cluster computing or grid computing, they present unique challenges in terms of coordination, data consistency, and security. As technology advances and the demand for effective data management increases, distributed systems will continue to evolve and adapt.

History

The concept of distributed systems has its roots in the early days of computing when multiple computers were connected via networks to share resources. The genesis of distributed systems can be traced back to the following milestones:

During the 1970s, early efforts such as the ARPANET showcased the potential of connecting computers remotely, facilitating communication and collaboration among researchers.
By the 1980s, the introduction of distributed file systems and early database management systems allowed organizations to manage data across multiple nodes, albeit with significant limitations in performance and scalability.
The 1990s saw the emergence of more sophisticated mechanisms such as remote procedure calls (RPC) and various protocols for inter-process communication, which laid the groundwork for modern distributed systems.
The late 1990s and early 2000s witnessed the rise of web-based applications and the shift towards service-oriented architectures enabling distributed computing on a global scale.
Recent developments in cloud computing and microservices have further transformed the landscape of distributed systems, allowing for highly scalable and fault-tolerant applications.

Design and Architecture

The design of a distributed system can vary greatly based on the intended use case, architecture, and protocols employed. It typically involves several key components and design patterns:

1. Components

Distributed systems consist of multiple autonomous components that work collaboratively. The primary component types include:

Clients: Users or systems that request services from servers.
Servers: Components that provide services to clients, typically by processing requests and returning results.
Middleware: Software that lies between client applications and server resources, aiding communication and data management.

2. Architectural Models

Several architectural models guide the design of distributed systems:

Client-Server Architecture: In this model, clients request resources or services from centralized servers which provide responses. This is the most common distributed system architecture.
Peer-to-Peer (P2P) Architecture: All nodes have equal responsibilities and can act as both client and server. This model promotes resource sharing and decentralization.
Multi-tier Architecture: An extension of the client-server model that separates different functions (such as presentation, application processing, and database management) into different layers.

3. Communication Protocols

The choice of communication protocols significantly impacts the performance and reliability of a distributed system. Common protocols include:

Remote Procedure Call (RPC): Allows a program to cause a procedure to execute in another address space.
Message Queuing Protocols (e.g., MQTT, AMQP): Provides a mechanism for distributed applications to communicate asynchronously.
HTTP/REST: A stateless communication model often used in web services, which allows clients and servers to exchange data over the internet.

4. Consistency Models

Data consistency is a critical aspect of distributed systems, often dictated by the chosen consistency model such as:

Strong Consistency: Guarantees that all accesses will return the latest data after an update.
Eventual Consistency: Allows for temporary inconsistencies, with the guarantee that all replicas will become consistent eventually.
Causal Consistency: Ensures that operations that are causally related are seen by all processes in the same order.

5. Fault Tolerance and Replication

To ensure reliability, distributed systems often incorporate fault tolerance mechanisms, such as data replication, consensus algorithms (e.g., Paxos, Raft), and failure detection strategies. These methods allow systems to continue functioning despite the presence of hardware or software failures.

Usage and Implementation

Distributed systems find applications across various domains, including:

Cloud Computing: Offers on-demand access to a network of servers, allowing scalable and flexible resource utilization.
Big Data Processing: Frameworks like Hadoop and Spark leverage distributed systems to process large data sets efficiently.
Content Delivery Networks (CDNs): Distribute content geographically to improve access speed and redundancy by caching data across multiple nodes.
Blockchain: A distributed ledger technology that ensures secure peer-to-peer transactions without a central authority.

The implementation of distributed systems requires a deep understanding of both the technical challenges involved and the operational requirements of the applications being developed. Developers must consider aspects such as network latency, data locality, and synchronization to achieve optimal performance.

Challenges in Implementation

Implementing distributed systems introduces several challenges, including:

Network Partitioning: Communication failures that lead to split-brain scenarios can compromise data consistency.
Latency Issues: Network delays can impact system responsiveness, particularly in real-time applications.
Complex Debugging: The distributed nature of the system can complicate troubleshooting and error detection.

Addressing these challenges requires robust designs, continuous monitoring, and efficient resource management.

Real-world Examples

1. Google Distributed Systems

Google has developed a range of distributed systems including:

Google File System (GFS): Designed to provide high-throughput access to large datasets using a distributed file system architecture.
Bigtable: A distributed storage system for managing structured data, designed to scale to petabytes across thousands of servers.
MapReduce: A programming model designed for distributed processing of large data sets across clusters.

2. Amazon Web Services (AWS)

AWS provides cloud computing services that leverage distributed system architectures, including:

Amazon S3 (Simple Storage Service): Allows storage and retrieval of any amount of data at any time, featuring high availability and scalability.
Amazon DynamoDB: A fully managed NoSQL database service that delivers fast and predictable performance with seamless scalability.
AWS Lambda: A serverless compute service that automatically manages the underlying infrastructure, allowing developers to execute code in response to events.

3. Apache Hadoop Ecosystem

Apache Hadoop is a suite of tools designed for distributed storage and processing of large data sets. Its ecosystem includes:

Hadoop Distributed File System (HDFS): A distributed file system that provides high-throughput access to application data.
YARN (Yet Another Resource Negotiator): A resource management layer that allocates system resources to applications running in a Hadoop cluster.
MapReduce: A programming model for processing large data sets in parallel across a Hadoop Cluster.

Criticism and Controversies

Despite their advantages, distributed systems face criticism and several controversies, particularly regarding issues of security, data privacy, and inefficiency:

Security Concerns: The distributed nature of these systems can expose them to a variety of attacks such as Distributed Denial of Service (DDoS), making security a paramount concern.
Data Privacy: The handling of sensitive data across multiple nodes raises concerns about unauthorized access and data breaches.
Complexity and Cost: The implementation and maintenance of distributed systems can be complex and costly, especially for small enterprises without dedicated resources.

Understanding these criticisms is crucial for developers and organizations to address potential pitfalls effectively.

Influence and Impact

Distributed systems have profoundly influenced the landscape of modern computing, driving innovations across various fields:

They have enabled businesses to increase scalability and reliability in their operations.
The rise of cloud computing, driven by distributed systems, has reshaped the IT industry, affecting how organizations manage resources and data.
Innovations in big data technologies, such as Apache Spark and Kafka, are heavily reliant on distributed system paradigms.
The development of blockchain technologies represents a push towards more decentralized, secure, and transparent systems.

The ongoing evolution of distributed systems is expected to contribute further to advancements in computing, facilitating new application possibilities and addressing global challenges.

References

[1] - Microsoft Cloud Computing
[2] - IBM's Overview of Distributed Systems
[3] - DigitalOcean's Guide to Cloud Computing
[4] - Apache Hadoop Official Website
[5] - Introduction to Amazon Web Services
[6] - "The Google File System," by Sanjay Ghemawat et al.

@@ Line 1: / Line 1: @@
-== Introduction ==
+== Distributed Systems ==
-Distributed systems are collections of independent computers that collaborate through the sharing of networked resources to achieve a common goal. They operate as a cohesive unit while maintaining the autonomy of individual components. These systems are designed to handle large-scale, complex applications and can range from small clusters to vast networks of geographically distributed nodes. The significance of distributed systems lies in their ability to improve scalability, reliability, and resource utilization compared to traditional centralized systems.
+A '''distributed system''' is a model in which components located on networked computers communicate and coordinate their actions by passing messages. The components interact with one another in order to achieve a common goal. Distributed systems can be categorized based on their architecture, networking topology, and consistency models, amongst other factors. They are increasingly important in computing, as they facilitate the development of applications that are more scalable, resilient, and accessible.
-== History ==
+== Introduction ==
-The concept of distributed systems emerged in the late 20th century as computer networks began to proliferate. Early forms of distributed computing can be traced back to the 1960s, when researchers sought to connect multiple computers to process tasks in parallel. The development of the ARPANET in the late 1960s, which was funded by the U.S. Department of Defense, laid the groundwork for networked communication and the eventual rise of distributed systems.
+Distributed systems are prevalent in modern computing and form the backbone of many major applications and services. They provide key advantages such as resource sharing, fault tolerance, scalability, and improved performance. In a distributed system, components located on multiple networked computers work together to perform tasks, effectively giving the appearance of a single coherent system to the user. The emergence of cloud computing, web services, and peer-to-peer systems has further propelled the relevance and use of distributed systems.
-Throughout the 1970s and 1980s, key advancements were made in distributed algorithms and protocols, including the development of the Client-Server model, which became the foundation for many subsequent distributed applications. Notable contributions include the implementations of distributed databases and file systems, along with the introduction of communication protocols such as TCP/IP.
+While distributed systems may seem similar to cluster computing or grid computing, they present unique challenges in terms of coordination, data consistency, and security. As technology advances and the demand for effective data management increases, distributed systems will continue to evolve and adapt.
-The 1990s marked a significant milestone with the advent of the World Wide Web, which highlighted the potential of distributed systems to provide services on a global scale. Innovations such as peer-to-peer networks and grid computing emerged during this period, expanding the application of distributed systems beyond traditional boundaries.
+== History ==
-With the rise of cloud computing in the early 2000s, distributed systems gained renewed attention. Companies began leveraging distributed architectures to provide scalable services and applications over the internet. Technologies such as MapReduce, Hadoop, and distributed databases like Amazon DynamoDB and Google Bigtable became crucial components in managing vast amounts of data across distributed environments.
+The concept of distributed systems has its roots in the early days of computing when multiple computers were connected via networks to share resources. The genesis of distributed systems can be traced back to the following milestones:
+* During the 1970s, early efforts such as the ARPANET showcased the potential of connecting computers remotely, facilitating communication and collaboration among researchers.
+* By the 1980s, the introduction of distributed file systems and early database management systems allowed organizations to manage data across multiple nodes, albeit with significant limitations in performance and scalability.
+* The 1990s saw the emergence of more sophisticated mechanisms such as remote procedure calls (RPC) and various protocols for inter-process communication, which laid the groundwork for modern distributed systems.
+* The late 1990s and early 2000s witnessed the rise of web-based applications and the shift towards service-oriented architectures enabling distributed computing on a global scale.
+* Recent developments in cloud computing and microservices have further transformed the landscape of distributed systems, allowing for highly scalable and fault-tolerant applications.
 == Design and Architecture ==
-Distributed systems are characterized by specific architectural patterns and design principles that differentiate them from centralized systems. The design focuses on ensuring consistent performance, fault tolerance, and resource management across multiple nodes. Key elements of distributed system architecture include:
+The design of a distributed system can vary greatly based on the intended use case, architecture, and protocols employed. It typically involves several key components and design patterns:
 === 1. Components ===
-Distributed systems typically consist of multiple components, which may include:
-* **Nodes**: Individual computing devices that participate in the system.
-* **Middleware**: Software that acts as an intermediary layer, facilitating communication and data exchange between nodes.
-* **Storage systems**: Solutions that provide distributed data storage and management capabilities.
-=== 2. Communication ===
+Distributed systems consist of multiple autonomous components that work collaboratively. The primary component types include:
-Effective communication is pivotal in distributed systems. Various communication models are used, including:
+* '''Clients''': Users or systems that request services from servers.
-* **Message Passing**: Nodes communicate by sending and receiving messages.
+* '''Servers''': Components that provide services to clients, typically by processing requests and returning results.
-* **Shared Memory**: Nodes share a common memory space, although this requires synchronization mechanisms to ensure data consistency.
+* '''Middleware''': Software that lies between client applications and server resources, aiding communication and data management.
-=== 3. Consistency Models ===
+=== 2. Architectural Models ===
-Maintaining data consistency across distributed nodes is challenging due to the potential for asynchrony and network partitioning. Common consistency models include:
-* **Strong Consistency**: Guarantees that all nodes see the same data at the same time.
-* **Eventual Consistency**: Allows for temporary discrepancies, with the assurance that all updates will propagate to all nodes eventually.
-=== 4. Fault Tolerance ===
+Several architectural models guide the design of distributed systems:
-Distributed systems must be resilient to component failures. Strategies to achieve fault tolerance include:
+* '''Client-Server Architecture''': In this model, clients request resources or services from centralized servers which provide responses. This is the most common distributed system architecture.
-* **Replication**: Duplicating data across multiple nodes to ensure availability in the case of failures.
+* '''Peer-to-Peer (P2P) Architecture''': All nodes have equal responsibilities and can act as both client and server. This model promotes resource sharing and decentralization.
-* **Consensus Algorithms**: Mechanisms such as Paxos and Raft are used to achieve agreement among nodes despite failures.
+* '''Multi-tier Architecture''': An extension of the client-server model that separates different functions (such as presentation, application processing, and database management) into different layers.
-=== 5. Scalability ===
+=== 3. Communication Protocols ===
-Scalability refers to the ability of a system to handle increasing loads by adding more resources. Distributed systems may be designed for:
-* **Vertical Scaling**: Adding more resources (CPU, memory) to existing nodes.
-* **Horizontal Scaling**: Adding more nodes to the system, distributing the workload.
-== Usage and Implementation ==
+The choice of communication protocols significantly impacts the performance and reliability of a distributed system. Common protocols include:
+* '''Remote Procedure Call (RPC)''': Allows a program to cause a procedure to execute in another address space.
+* '''Message Queuing Protocols (e.g., MQTT, AMQP)''': Provides a mechanism for distributed applications to communicate asynchronously.
+* '''HTTP/REST''': A stateless communication model often used in web services, which allows clients and servers to exchange data over the internet.
-Distributed systems find application in numerous fields, including cloud computing, data storage and management, web services, and enterprise applications. Below are some prominent implementations and their use cases:
+=== 4. Consistency Models ===
-=== 1. Cloud Computing ===
+Data consistency is a critical aspect of distributed systems, often dictated by the chosen consistency model such as:
-Cloud platforms such as Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform are built on distributed systems principles. They enable users to provision resources, deploy applications, and scale services dynamically across multiple geographical locations.
+* '''Strong Consistency''': Guarantees that all accesses will return the latest data after an update.
+* '''Eventual Consistency''': Allows for temporary inconsistencies, with the guarantee that all replicas will become consistent eventually.
+* '''Causal Consistency''': Ensures that operations that are causally related are seen by all processes in the same order.
-=== 2. Distributed Databases ===
+=== 5. Fault Tolerance and Replication ===
-These databases, such as Apache Cassandra, MongoDB, and CockroachDB, leverage distribution to provide scalability and fault tolerance for data storage. They allow for high availability and can handle large volumes of transactions across distributed nodes.
-=== 3. Microservices Architecture ===
+To ensure reliability, distributed systems often incorporate fault tolerance mechanisms, such as data replication, consensus algorithms (e.g., Paxos, Raft), and failure detection strategies. These methods allow systems to continue functioning despite the presence of hardware or software failures.
-The microservices pattern promotes the development of applications as a suite of small, independent services that communicate over a network. This architecture enhances scalability, as services can be developed, deployed, and scaled independently.
-=== 4. Peer-to-Peer Networks ===
+== Usage and Implementation ==
-In peer-to-peer (P2P) systems, nodes act as both clients and servers, sharing resources directly with each other. P2P applications include file sharing (e.g., BitTorrent) and cryptocurrency networks (e.g., Bitcoin), which capitalize on the decentralized nature of distributed systems.
-=== 5. Big Data Processing ===
+Distributed systems find applications across various domains, including:
-Frameworks such as Apache Hadoop and Apache Spark utilize distributed systems to perform large-scale data processing tasks. These frameworks enable the analysis of massive datasets across clusters of machines, allowing businesses to derive insights and make data-driven decisions.
+* '''Cloud Computing''': Offers on-demand access to a network of servers, allowing scalable and flexible resource utilization.
+* '''Big Data Processing''': Frameworks like Hadoop and Spark leverage distributed systems to process large data sets efficiently.
+* '''Content Delivery Networks (CDNs)''': Distribute content geographically to improve access speed and redundancy by caching data across multiple nodes.
+* '''Blockchain''': A distributed ledger technology that ensures secure peer-to-peer transactions without a central authority.
-== Real-world Examples ==
+The implementation of distributed systems requires a deep understanding of both the technical challenges involved and the operational requirements of the applications being developed. Developers must consider aspects such as network latency, data locality, and synchronization to achieve optimal performance.
-Distributed systems are prevalent in various domains and industries. Here are several notable examples:
+=== Challenges in Implementation ===
-=== 1. Google MapReduce ===
+Implementing distributed systems introduces several challenges, including:
-MapReduce is a programming model designed for processing large datasets with a distributed algorithm on a cluster. Google utilized it to index the web and extract meaningful data, revolutionizing data processing capabilities.
+* '''Network Partitioning''': Communication failures that lead to split-brain scenarios can compromise data consistency.
+* '''Latency Issues''': Network delays can impact system responsiveness, particularly in real-time applications.
+* '''Complex Debugging''': The distributed nature of the system can complicate troubleshooting and error detection.
-=== 2. Amazon DynamoDB ===
+Addressing these challenges requires robust designs, continuous monitoring, and efficient resource management.
-DynamoDB is a fully managed NoSQL database service designed to handle high-traffic workloads while providing low latency data access. It scales automatically and offers high availability across multiple regions.
-=== 3. Bitcoin Blockchain ===
+== Real-world Examples ==
-The Bitcoin blockchain operates as a distributed ledger that ensures transactions are securely recorded across a network of nodes. It employs a consensus algorithm known as proof-of-work to validate transactions and maintain the integrity of the ledger.
-=== 4. Apache Kafka ===
+=== 1. Google Distributed Systems ===
-Kafka is a distributed streaming platform that enables the building of real-time data pipelines and streaming applications. It is designed to handle high-throughput data feeds, making it a fundamental component in microservices architectures.
-=== 5. Kubernetes ===
+Google has developed a range of distributed systems including:
-Kubernetes is an orchestration platform for automating the deployment, scaling, and management of containerized applications. It operates in a distributed manner, allowing developers to manage clusters of machines efficiently.
+* '''Google File System (GFS)''': Designed to provide high-throughput access to large datasets using a distributed file system architecture.
+* '''Bigtable''': A distributed storage system for managing structured data, designed to scale to petabytes across thousands of servers.
+* '''MapReduce''': A programming model designed for distributed processing of large data sets across clusters.
-== Criticism and Controversies ==
+=== 2. Amazon Web Services (AWS) ===
-While distributed systems offer numerous advantages, they are not without challenges and criticisms. Concerns include:
+AWS provides cloud computing services that leverage distributed system architectures, including:
+* '''Amazon S3 (Simple Storage Service)''': Allows storage and retrieval of any amount of data at any time, featuring high availability and scalability.
+* '''Amazon DynamoDB''': A fully managed NoSQL database service that delivers fast and predictable performance with seamless scalability.
+* '''AWS Lambda''': A serverless compute service that automatically manages the underlying infrastructure, allowing developers to execute code in response to events.
-=== 1. Complexity ===
+=== 3. Apache Hadoop Ecosystem ===
-The design, implementation, and maintenance of distributed systems can be significantly more complex than centralized systems. Debugging and troubleshooting issues can be particularly challenging due to the involvement of multiple components and potential network-related problems.
-=== 2. Security Issues ===
+Apache Hadoop is a suite of tools designed for distributed storage and processing of large data sets. Its ecosystem includes:
-Distributed systems can introduce vulnerabilities, especially when nodes communicate over insecure networks. Ensuring data security, privacy, and integrity across distributed components is an ongoing challenge that must be addressed through robust security mechanisms.
+* '''Hadoop Distributed File System (HDFS)''': A distributed file system that provides high-throughput access to application data.
+* '''YARN (Yet Another Resource Negotiator)''': A resource management layer that allocates system resources to applications running in a Hadoop cluster.
+* '''MapReduce''': A programming model for processing large data sets in parallel across a Hadoop Cluster.
-=== 3. Performance Overheads ===
+== Criticism and Controversies ==
-Communication between distributed nodes can introduce latency, impacting overall system performance. Optimizing data exchange and ensuring efficient communication protocols are critical to mitigating these challenges.
-=== 4. Data Consistency Challenges ===
+Despite their advantages, distributed systems face criticism and several controversies, particularly regarding issues of security, data privacy, and inefficiency:
-Achieving strong consistency in distributed systems can lead to trade-offs with availability and performance, notably in the presence of network partitions. The CAP theorem postulates that it's impossible to achieve all three properties (Consistency, Availability, Partition tolerance) simultaneously, necessitating design decisions that can impact system behavior.
+* '''Security Concerns''': The distributed nature of these systems can expose them to a variety of attacks such as Distributed Denial of Service (DDoS), making security a paramount concern.
+* '''Data Privacy''': The handling of sensitive data across multiple nodes raises concerns about unauthorized access and data breaches.
+* '''Complexity and Cost''': The implementation and maintenance of distributed systems can be complex and costly, especially for small enterprises without dedicated resources.
-=== 5. Vendor Lock-In ===
+Understanding these criticisms is crucial for developers and organizations to address potential pitfalls effectively.
-As organizations adopt cloud-based distributed solutions, they may become dependent on specific vendors, leading to potential lock-in situations where migration to alternative platforms becomes costly and complicated.
 == Influence and Impact ==
-The evolution of distributed systems has had a profound influence on various fields, shaping technology, infrastructure, and practices both in industry and academia. The impact includes:
+Distributed systems have profoundly influenced the landscape of modern computing, driving innovations across various fields:
+* They have enabled businesses to increase scalability and reliability in their operations.
-=== 1. Evolution of Software Development ===
+* The rise of cloud computing, driven by distributed systems, has reshaped the IT industry, affecting how organizations manage resources and data.
-The adoption of distributed architectures has driven the transition from monolithic application development to more modular and agile approaches. The emergence of microservices architecture aligns with contemporary development paradigms that emphasize automation, continuous integration, and deployment.
+* Innovations in big data technologies, such as Apache Spark and Kafka, are heavily reliant on distributed system paradigms.
+* The development of blockchain technologies represents a push towards more decentralized, secure, and transparent systems.
-=== 2. Growth of Cloud Computing ===
-The establishment of distributed systems has been instrumental in the rise of cloud computing paradigms. Organizations can leverage cloud-based resources to achieve scalability and flexibility, leading to cost savings and improved operational efficiency.
-=== 3. Transforming Data Analytics ===
-Distributed systems have transformed the landscape of data analytics by enabling large-scale processing and analysis of big data, allowing organizations to harness insights from vast datasets that were previously infeasible to manage.
-=== 4. Innovations in Networking and Infrastructure ===
-The design principles of distributed systems have influenced advancements in network infrastructure, leading to the proliferation of content delivery networks (CDNs), edge computing, and enhanced network protocols that support efficient communication across distributed environments.
-=== 5. Academic Research ===
+The ongoing evolution of distributed systems is expected to contribute further to advancements in computing, facilitating new application possibilities and addressing global challenges.
-Distributed systems continue to be a vibrant area of academic research, contributing to advancements in algorithms, protocols, and methodologies that address key challenges such as fault tolerance, consensus, and performance optimization.
 == See also ==
 * [[Cloud Computing]]
+* [[Cluster Computing]]
+* [[Grid Computing]]
 * [[Microservices]]
-* [[Peer-to-Peer Networks]]
+* [[Byzantine Fault Tolerance]]
-* [[Blockchain]]
+* [[Distributed Ledger Technology]]
-* [[Distributed Databases]]
+* [[Peer-to-Peer Networking]]
-* [[Grid Computing]]
-* [[Consensus Algorithms]]
-* [[CAP Theorem]]
-* [[Fault Tolerance]]
 == References ==
-* [https://www.darwinsys.com/ Distributed Systems Overview]
+* [https://www.microsoft.com/cloud-computing] - Microsoft Cloud Computing
-* [https://researchgate.net/publication/319208064 Distributed Systems: Principles and Paradigms]
+* [https://www.ibm.com/cloud/learn/distributed-systems] - IBM's Overview of Distributed Systems
-* [https://aws.amazon.com/architecture/distributed-systems/ AWS Distributed Services]
+* [https://www.digitalocean.com/community/tutorials/what-is-cloud-computing] - DigitalOcean's Guide to Cloud Computing
-* [https://towardsdatascience.com/a-comprehensive-guide-to-distributed-systems-for-practitioners-290b9d3f9b4c Comprehensive Guide to Distributed Systems]
+* [https://hadoop.apache.org/] - Apache Hadoop Official Website
-* [https://cassandra.apache.org/_/index.html Apache Cassandra Official Documentation]
+* [https://aws.amazon.com/what-is-aws/] - Introduction to Amazon Web Services
+* [https://research.google/pubs/archive/87533.pdf] - "The Google File System," by Sanjay Ghemawat et al.
 [[Category:Distributed computing]]
 [[Category:Computer science]]
-[[Category:Systems architecture]]
+[[Category:Systems theory]]