Swarming Overlay Construction Strategies

Abstract

Swarming peer-to-peer systems play an increasingly instrumental role in Internet content distribution. It is therefore important to better understand how these systems behave in practice. Recent research efforts have looked at various protocol parameters and have measured how they affect system performance and robustness. However, the importance of the strategy based on which peers establish connections has been largely overlooked.This work utilizes extensive simulations to examine the default overlay construction strategy in BitTorrent systems. Based on the results, we identify a critical parameter, the maximum allowable number of outgoing connections at each peer, and evaluate its impact on the robustness of the generated overlay. We find that there is no single optimal value for this parameter using the default strategy. We then propose an alternative strategy that allows certain new peer connection requests to replace existing connections. Further experiments with the new strategy demonstrate that it outperforms the default one for all considered metrics by creating an overlay more robust to churn. Additionally, our proposed strategy exhibits optimal behavior for a well-defined value of the maximum number of outgoing connections, thereby removing the need to set this parameter in an ad-hoc manner 1 .

Modeling BitTorrent-like systems with many classes of users

“…This is because a slow seed may not be fast enough in disseminating file blocks into the system, and situations may arise where a user may not have blocks that his/her neighbors are interested in. (A similar situation in a flash crowd scenario may arise if one keeps the seed upload link capacity fixed but starts increasing the number of leechers [Al-Hamra et al 2009]). To investigate this scenario, in Figures 5(ii) and 5(iii), we repeat the same experiment as in Figure 5(i) but with lower upload link capacities for the seed, 250Kbps and 100Kbps respectively.…”

Exploring your neighborhood: Comparing connection strategies in swarming networks through evolving graphs

“…It is motivated by the intuition that choosing random neighbors tends to generate a random graph [3], whose main property is the robustness to handle highly transient populations [10]. However, previous work shows that the strategy clusters peers according to their arrival time in the system, which unfortunately decreases the overall network robustness [4], [5]. In an extreme scenario, the topology could even be partitioned.…”

“…The work on preemption (or preemptive) strategy [4] was one of the first attempts to model connections among peers. Through simulations, the authors captured and compared the evolution of the overlay topology, considering both preemptive and tracker (inactive) strategies for neighborhood management.…”

“…The preemptive strategy [4], an alternative for the inactive connection management, is executed when a peer with full neighborhood receives a connection request. Instead of simply rejecting the request (inactive strategy), the peer disconnects a current neighbor (preempts it) in order to accept the incoming connection request.…”

“…The first strategy, inactive, is the simplest: once a connection is established, it is kept alive until one of the connected peers leaves the system [3]. In the second one, called preemptive strategy, peers may replace ("preempt") current neighbors when receiving new incoming connection requests [4]. The third and last strategy represents the reverse process, that is, pro-actively and periodically a peer disconnects a subset of its neighborhood, allowing new connections to be established [5].…”