This paper introduces a Content-Addressable Network (CAN), a distributed infrastructure that provides hash table-like functionality on Internet-like scales. The CAN is scalable, fault-tolerant, and self-organizing. It is designed to address the limitations of existing peer-to-peer systems, such as Napster and Gnutella, which suffer from scalability and robustness issues. The CAN uses a virtual d-dimensional Cartesian coordinate space to store and retrieve (key, value) pairs. Each node in the CAN maintains a zone of the coordinate space and a list of neighboring zones. When a key is inserted, it is mapped to a point in the coordinate space, and the corresponding value is stored at the node that owns the zone containing that point. To retrieve a value, the key is mapped to the same point, and the value is retrieved from the node that owns the zone containing that point. The CAN is completely distributed, requiring no centralized control, and is scalable, as nodes maintain only a small amount of state that is independent of the number of nodes in the system. The CAN is also fault-tolerant, as nodes can route around failures. The design of the CAN is based on the idea of a hash table, where the basic operations are insertion, lookup, and deletion of (key, value) pairs. The CAN is used in peer-to-peer systems for file sharing, as well as in large-scale storage management systems. The CAN is also used in wide-area name resolution services. The design of the CAN is evaluated through simulation, and the results show that the CAN is scalable, robust, and has low latency. The CAN is also compared to other routing algorithms, such as the Distance Vector (DV) and Link State (LS) algorithms used in IP routing. The CAN is found to be more scalable and robust than these algorithms. The CAN is also compared to the Plaxton algorithm, which is used in web caching environments. The CAN is found to be more suitable for peer-to-peer systems, as it is self-configuring and can handle a large number of hosts. The CAN is also more suitable for environments where nodes are potentially flaky. The CAN is designed to be implemented entirely at the application level. The CAN is evaluated through simulation, and the results show that the CAN is scalable, robust, and has low latency. The CAN is also compared to other routing algorithms, such as the Distance Vector (DV) and Link State (LS) algorithms used in IP routing. The CAN is found to be more scalable and robust than these algorithms. The CAN is also compared to the Plaxton algorithm, which is used in web caching environments. The CAN is found to be more suitable for peer-to-peer systems, as it is self-configuring and can handle a large number of hosts. The CAN is also more suitable for environments where nodes are potentially flaky. The CAN is designed to be implemented entirely at the application level.This paper introduces a Content-Addressable Network (CAN), a distributed infrastructure that provides hash table-like functionality on Internet-like scales. The CAN is scalable, fault-tolerant, and self-organizing. It is designed to address the limitations of existing peer-to-peer systems, such as Napster and Gnutella, which suffer from scalability and robustness issues. The CAN uses a virtual d-dimensional Cartesian coordinate space to store and retrieve (key, value) pairs. Each node in the CAN maintains a zone of the coordinate space and a list of neighboring zones. When a key is inserted, it is mapped to a point in the coordinate space, and the corresponding value is stored at the node that owns the zone containing that point. To retrieve a value, the key is mapped to the same point, and the value is retrieved from the node that owns the zone containing that point. The CAN is completely distributed, requiring no centralized control, and is scalable, as nodes maintain only a small amount of state that is independent of the number of nodes in the system. The CAN is also fault-tolerant, as nodes can route around failures. The design of the CAN is based on the idea of a hash table, where the basic operations are insertion, lookup, and deletion of (key, value) pairs. The CAN is used in peer-to-peer systems for file sharing, as well as in large-scale storage management systems. The CAN is also used in wide-area name resolution services. The design of the CAN is evaluated through simulation, and the results show that the CAN is scalable, robust, and has low latency. The CAN is also compared to other routing algorithms, such as the Distance Vector (DV) and Link State (LS) algorithms used in IP routing. The CAN is found to be more scalable and robust than these algorithms. The CAN is also compared to the Plaxton algorithm, which is used in web caching environments. The CAN is found to be more suitable for peer-to-peer systems, as it is self-configuring and can handle a large number of hosts. The CAN is also more suitable for environments where nodes are potentially flaky. The CAN is designed to be implemented entirely at the application level. The CAN is evaluated through simulation, and the results show that the CAN is scalable, robust, and has low latency. The CAN is also compared to other routing algorithms, such as the Distance Vector (DV) and Link State (LS) algorithms used in IP routing. The CAN is found to be more scalable and robust than these algorithms. The CAN is also compared to the Plaxton algorithm, which is used in web caching environments. The CAN is found to be more suitable for peer-to-peer systems, as it is self-configuring and can handle a large number of hosts. The CAN is also more suitable for environments where nodes are potentially flaky. The CAN is designed to be implemented entirely at the application level.