Adriana Iamnitchi

Mapping the Gnutella network

We studied the topology and protocols of the public Gnutella network. Its substantial user base and open architecture make it a good large-scale, if uncontrolled, testbed. We captured the networks topology, generated traffic, and dynamic behavior to determine its connectivity structure and how well (if at all) Gnutellas overlay network topology maps to the physical Internet infrastructure. Our analysis of the network allowed us to evaluate costs and benefits of the peer-to-peer (P2P) approach and to investigate possible improvements that would allow better scaling and increased reliability in Gnutella and similar networks. A mismatch between Gnutellas overlay network topology and the Internet infrastructure has critical performance implications.

On Death, Taxes, and the Convergence of Peer-to-Peer and Grid Computing

It has been reported [25] that life holds but two certainties, death and taxes. And indeed, it does appear that any society-and in the context of this article, any large-scale distributed system-must address both death (failure) and the establishment and maintenance of infrastructure (which we assert is a major motivation for taxes, so as to justify our title!).

Giggle: A Framework for Constructing Scalable Replica Location Services

In wide area computing systems, it is often desirable to create remote read-only copies (replicas) of files. Replication can be used to reduce access latency, improve data locality, and/or increase robustness, scalability and performance for distributed applications. We define a replica location service (RLS) as a system that maintains and provides access to information about the physical locations of copies. An RLS typically functions as one component of a data grid architecture. This paper makes the following contributions. First, we characterize RLS requirements. Next, we describe a parameterized architectural framework, which we name Giggle (for GIGa-scale Global Location Engine), within which a wide range of RLSs can be defined. We define several concrete instantiations of this framework with different performance characteristics. Finally, we present initial performance results for an RLS prototype, demonstrating that RLS systems can be constructed that meet performance goals.

On Fully Decentralized Resource Discovery in Grid Environments

Computational grids provide mechanisms for sharing and accessing large and heterogeneous collections of remote resources such as computers, online instruments, storage space, data, and applications. Resources are identified based on a set of desired attributes. Resource attributes have various degrees of dynamism, from mostly static attributes, like operating system version, to highly dynamic ones, like network bandwidth or CPU load. In this paper we propose a peer-to-peer architecture for resource discovery in a large and dynamic collection of resources. We evaluate a set of request-forwarding algorithms in a fully decentralized architecture, designed to accommodate heterogeneity (in both sharing policies and resource types) and dynamism. For this, we build a testbed that models two usage characteristics: (1) resource distribution on peers, that varies in the number and the frequency of shared resources; and (2) various requests patterns for resources. We analyzed our resource discovery mechanisms on up to 5000 peers, where each peer provides information about at least one resource. We learned that a decentralized approach is not only desirable from administrative reasons, but it is also supported by promising performance results. Our results also allow us to characterize the correlation between resource discovery performance and sharing characteristics.

A peer-to-peer approach to resource location in grid environments

Computational grids provide mechanisms for sharing and accessing large and heterogeneous collections of remote resources such as computers, online instruments, storage space, data, and applications. Resources are requested by specifying a set of desired attributes. Resource attributes have various degrees of dynamism, from mostly static attributes, such as operating system version, to highly dynamic ones, such as available network bandwidth or CPU load. Another dimension of dynamism is introduced by variable and highly diverse sharing policies: resources are made available to the grid community based on locally defined and potentially changing policies.

Improving Data Availability through Dynamic Model-Driven Replication in Large Peer-to-Peer Communities

Efficient data sharing in global peer-to-peer systems is complicated by erratic node failure, unreliable network connectivity and limited bandwidth. Replicating data on multiple nodes can improve availability and response time. Yet determining when and where to replicate data in order to meet performance goals in large-scale systems with many users and files, dynamic network characteristics, and changing user behavior is difficult. We propose an approach in which peers create replicas automatically in a decentralized fashion, as required to meet availability goals. The aim of our framework is to maintain a threshold level of availability at all times. We identify a set of factors that hinder data availability and propose a model that decides when more replication is necessary. We evaluate the accuracy and performance of the proposed model using simulations. Our preliminary results show that the model is effective in predicting the required number of replicas in the system.

Small-world file-sharing communities

Web caches, content distribution networks, peer-to-peer file sharing networks, distributed file systems, and data grids all have in common that they involve a community of users who generate requests for shared data. In each case, overall system performance can be improved significantly if we can first identify and then exploit interesting structure within a communitys access patterns. To this end, we propose a novel perspective on file sharing that considers the relationships that form among users based on the files in which they are interested. We propose a new structure that captures common user interests in data - the data-sharing graph - and justify its utility with studies on three data-distribution systems: a high-energy physics collaboration, the Web, and the Kazaa peer-to-peer network. We find small-world patterns in the data-sharing graphs of all three communities. We analyze these graphs and propose some probable causes for these emergent small-world patterns. The significance of small-world patterns is twofold: it provides a rigorous support to intuition and, perhaps most importantly, it suggests ways to design mechanisms that exploit these naturally emerging patterns.

Locating Data in (Small-World?) Peer-to-Peer Scientific Collaborations

Data-sharing scientific collaborations have particular characteristics, potentially different from the current peer-to-peer environments. In this paper we advocate the benefits of exploiting emergent patterns in self-configuring networks specialized for scientific data-sharing collaborations. We speculate that a peer-to-peer scientific collaboration network will exhibit small-world topology, as do a large number of social networks for which the same pattern has been documented. We propose a solution for locating data in decentralized, scientific, data-sharing environments that exploits the small-worlds topology. The research challenge we raise is: what protocols should be used to allow a self-configuring peer-to-peer network to form small worlds similar to the way in which the humans that use the network do in their social interactions?

The Globus Replica Location Service: Design and Experience

Distributed computing systems employ replication to improve overall system robustness, scalability, and performance. A replica location service (RLS) offers a mechanism to maintain and provide information about physical locations of replicas. This paper defines a design framework for RLSs that supports a variety of deployment options. We describe the RLS implementation that is distributed with the Globus toolkit and is in production use in several grid deployments. Features of our modular implementation include the use of soft-state protocols to populate a distributed index and Bloom filter compression to reduce overheads for distribution of index information. Our performance evaluation demonstrates that the RLS implementation scales well for individual servers with millions of entries and up to 100 clients. We describe the characteristics of existing RLS deployments and discuss how RLS has been integrated with higher-level data management services.

Prometheus: user-controlled P2P social data management for socially-aware applications

Recent Internet applications, such as online social networks and user-generated content sharing, produce an unprecedented amount of social information, which is further augmented by location or collocation data collected from mobile phones. Unfortunately, this wealth of social information is fragmented across many different proprietary applications. Combined, it could provide a more accurate representation of the social world, and it could enable a whole new set of socially-aware applications. We introduce Prometheus, a peer-to-peer service that collects and manages social information from multiple sources and implements a set of social inference functions while enforcing user-defined access control policies. Prometheus is socially-aware: it allows users to select peers that manage their social information based on social trust and exploits naturally-formed social groups for improved performance. We tested our Prometheus prototype on PlanetLab and built a mobile social application to test the performance of its social inference functions under realtime constraints. We showed that the social-based mapping of users onto peers improves the service response time and high service availability is achieved with low overhead.