A proximity-aware load balancing in peer-to-peer-based volunteer computing systems

Abstract

One of the main challenges in peer-to-peer based volunteer computing systems is efficient resource discovery algorithm. Load balancing is a part of resource discovery algorithm and aims to minimize the overall response time of the system. This paper introduces an analytical model based on distributed parallel queues to optimize the average response time of the system in a distributed manner. The proposed resource discovery algorithm consists of two phases. In the first phase, it selects peers in a load-balanced manner based on QoS constraints of request. In the second phase, a proximity-aware feature is applied to select the peer with minimum communication overhead among selected peers in the first phase. Two dispatching strategies are proposed for the load balancing based on stochastic analysis of routing in the distributed parallel queues. These policies adopt probabilistic and deterministic sequences to redirect requests to the capable peers in the system. Simulation results show that the proposed resource discovery algorithm improves the response time of user's requests by a factor of 1.8 under a moderate load.

Cloud-aware data intensive workflow scheduling on volunteer computing systems

“…This step consists of selecting a random target sub-workflow except max weight sub-workflow and moving a random vertex from a sub-workflow with its weight is more than target sub-workflow. The tabu search algorithm replaces the current solution with a best non-recently visited neighboring solution (lines [25][26][27][28]. This algorithm uses tabu list to forbid the recently visited solutions in order to prevent cycling (line 13,22 …”

“…The selection of the volunteer resources in the reporting nodes is done based on the QoS constraints and a load balancing policy [25]. This analytical model is a knowledge-free approach based on the routing in parallel queues [12].…”

“…In this model, the objective function is to find the optimal arrival rate to each logical cluster such that the incoming requests are distributed in a balanced manner among its logical clusters. The optimal arrival rate to each logical cluster j ( j  ) based on this model is computed by the following equation [25]:…”

Seamless LTE connectivity in high‐speed trains

“…In general, in load‐balancing problems, the arrival rate must be optimized based on the server capacity . Focusing on the effect of HO on service interruption while neglecting HO cyclostationarity (Section 2.1), we consider N independent discrete‐time Markov chains with q max +1 states to model N queues, one for each train carriage, where is the rate of Poisson‐like traffic generated by all passengers inside the i th carriage.…”

AVOCLOUDY: a simulator of volunteer clouds

“…For example, in France at INRIA (Clouds@Home), in Italy at the universities of Bologna (peer‐to‐peer cloud system) and Messina (Cloud@Home), and in general within the European Community (NaDa—NanoDatacenter , EDGI—European Desktop Grid initiative ), the CERN's volunteer cloud , in the USA within the National Science Foundation, which supports the BOINC project (to date with five research grants ), HTCondor (to date with four grants ), Seattle (to date with three grants ), SETI@home (with support from the National Science Foundation and NASA ), and the Federal University of Campina Grande in Brasil with the OurGrid initiative . Other works have modeled resource discovery algorithms in volunteer clouds relying on queueing theory and explored the opportunity for distributing Matlab simulations in a university volunteer infrastructure .…”