This paper presents the design and evaluation of Pastry, a scalable, distributed object location and routing scheme for wide-area peer-to-peer applications. Pastry provides application-level routing and object location in a potentially very large overlay network of nodes connected via the Internet. It can be used to support a wide range of peer-to-peer applications like global data storage, global data sharing, and naming. An insert operation in Pastry stores an object at a user-defined number of diverse nodes within the Pastry network. A lookup operation reliably retrieves a copy of the requested object if one exists. Moreover, a lookup is usually routed to the node nearest the client issuing the lookup (by some measure of proximity), among the nodes storing the requested object. Pastry is completely decentralized, scalable, and self-configuring; it automatically adapts to the arrival, departure and failure of nodes. Experimental results obtained with a prototype implementation on a simulated network of 100,000 nodes confirm Pastry's scalability, its ability to self-configure and adapt to node failures, and its good network locality properties.

This paper presents the design and evaluation of Pastry, a scalable, distributed object location and routing substrate for wide-area peer-to-peer applications. Pastry performs application-level routing and object location in a potentially very large overlay network of nodes connected via the Internet. It can be used to support a variety of peer-to-peer applications, including global data storage, data sharing, group communication and naming. Each node in the Pastry network has a unique identifier (nodeId). When presented with a message and a key, a Pastry node efficiently routes the message to the node with a nodeId that is numerically closest to the key, among all currently live Pastry nodes. Each Pastry node keeps track of its immediate neighbors in the nodeId space, and notifies applications of new node arrivals, node failures and recoveries. Pastry takes into account network locality; it seeks to minimize the distance messages travel, according to a to scalar proximity metric like the number of IP routing hops. Pastry is completely decentralized, scalable, and self-organizing; it automatically adapts to the arrival, departure and failure of nodes. Experimental results obtained with a prototype implementation on an emulated network of up to 100,000 nodes confirm Pastry’s scalability and efficiency, its ability to self-organize and adapt to node failures, and its good network locality properties.

This paper presents and evaluates the storage management and caching in PAST, a large-scale peer-to-peer persistent storage utility. PAST is based on a self-organizing, Internetbased overlay network of storage nodes that cooperatively route file queries, store multiple replicas of files, and cache additional copies of popular files. In the PAST system, storage nodes and files are each assigned uniformly distributed identifiers, and replicas of a file are stored at nodes whose identifier matches most closely the file’s identifier. This statistical assignment of files to storage nodes approximately balances the number of files stored on each node. However, non-uniform storage node capacities and file sizes require more explicit storage load balancing to permit graceful behavior under high global storage utilization; likewise, non-uniform popularity of files requires caching to minimize fetch distance and to balance the query load. We present and evaluate PAST, with an emphasis on its storage management and caching system. Extensive tracedriven experiments show that the system minimizes fetch distance, that it balances the query load for popular files, and that it displays graceful degradation of performance as the global storage utilization increases beyond 95%.

This paper sketches the design, and presents a scalability analysis and evaluation of NLS, a scalable naming and location service. NLS resolves textual names to the nearest of a set of replicated objects associated with that name, and is designed to scale to the dimensions of a world-wide service. Applications include resolving Web URIs to the nearest cached or replicated object that provides the associated content. The key design goals of NLS are scalability, performance, availability and ease of administration. NLS is based on a dynamically configured, distributed search tree, with a fat-tree based topology at the global level and spanning trees at the local level. Analysis and preliminary empirical results obtained with a prototype implementation indicate that the system scales as expected. 1