Scalable Load-Distance Balancing
Israel Cidon2007, Lecture Notes in Computer Science
Sign up for access to the world's latest research
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Abstract
We introduce the problem of load-distance balancing in assigning users of a delay-sensitive networked application to servers. We model the service delay experienced by a user as a sum of a network-incurred delay, which depends on its network distance from the server, and a server-incurred delay, stemming from the load on the server. The problem is to minimize the maximum service delay among all users. We address the challenge of finding a near-optimal assignment in a scalable distributed manner. The key to achieving scalability is using local solutions, whereby each server only communicates with a few close servers. Note, however, that the attainable locality of a solution depends on the workload -when some area in the network is congested, obtaining a near-optimal cost may require offloading users to remote servers, whereas when the network load is uniform, a purely local assignment may suffice. We present algorithms that exploit the opportunity to provide a local solution when possible, and thus have communication costs and stabilization times that vary according to the network congestion. We evaluate our algorithms with a detailed simulation case study of their application in assigning hosts to Internet gateways in an urban wireless mesh network (WMN).
Key takeaways
- Initially, every cluster consists of a single server.
- Under this assignment, a cluster's cost is defined as the maximum service delay among the users in this cluster.
- Two clusters C i and C j (1 ≤ i < j ≤ k) are called neighbors if there exists an l such that server s l belongs to cluster C i and server s l+1 belongs to cluster C j .
- Subsequently, dominating cluster leaders send "propose" messages to invite others to merge with them, and cluster leaders that agree respond with "accept" messages with full assignment information.
- The average cluster size does not exceed 2.5 servers for ε ≥ 0.5.
Related papers
Scalable Load-Distance Balancing in Large Networks
We focus on a setting where users of a real-time networked application need to be assigned to servers, e.g., assignment of hosts to Internet gateways in a wireless mesh network (WMN). The service delay experienced by a user is a sum of the network-incurred delay, which depends on its network distance from the server, and a server-incurred delay, stemming from the load on the server. We introduce the problem of load-distance balancing, which seeks to minimize the maximum service delay among all users. We address the challenge of finding a near-optimal assignment in a distributed manner, without global communication, in a large network. We present a scalable algorithm for doing so, and evaluate our solution with a case study of its application in an urban WMN.
On the Optimization of Load Balancing in Distributed Networks in the Presence of Delay
Distributing the total computational load across available processors is referred to as load balancing in the literature. A typical distributed system consists of a cluster of physically or virtually distant and independent computational elements (CEs), which are linked to one another by some communication medium. The workload has to be distributed over all the available CEs in proportion to their processing speed and availability such that the overall work completion time is minimized. In most practical distributed systems, due to the unknown character of the incoming workload, the nodes exhibit non-deterministic run-time performance. Thus, it is advantageous to perform the load balancing periodically during a run-time so that the run-time variability is minimized. This is referred to in the literature as dynamic load balancing [1]. However, the frequent load balancing requires the periodic communication (and transfer of loads, of course) between the CEs so that the shared knowledge of the load state of the system can be used by individual CEs to judiciously assign an appropriate fraction of the incoming loads to less busy CEs according to some load-balancing policy.
A dynamic load sharing algorithm for massively multiplayer online games
The 11th IEEE International Conference on Networks, 2003. ICON2003., 2003
To support hundreds of thousands of players in massively multiplayer online games, a distributed client-server architecture is widely used in which multiple servers are deployed and each server handles a partition of the virtual world. Because of the unpredictable movements and interactions of avatars, the concentration of avatars in some regions of the virtual world may cause some servers be overloaded. Existing load balancing schemes for distributed virtual environments and multiplayer games try to balance the workload among servers by transferring some workload of an overloaded server to other servers. While load balancing algorithms can minimize the average response time of the system, they may also result in frequent client migrations, which may damage the interactivity of an online game. In this paper, we propose a dynamic load sharing algorithm together with an efficient client migration scheme based on the concept of subscription regions. Simulation study has also been done to verify the effectiveness of our scheme.
Distributed algorithms for QoS load balancing
Distributed Computing, 2011
We consider a dynamic load balancing scenario in which users allocate resources in a non-cooperative and selfish fashion. The perceived performance of a resource for a user decreases with the number of users that allocate the resource. In our dynamic, concurrent model, users may reallocate resources in a round-based fashion. As opposed to various settings analyzed in the literature, we assume that users have quality of service (QoS) demands. A user has zero utility when falling short of a certain minimum performance threshold and having positive utility otherwise.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. IEEE TRANSACTIONS ON COMPUTERS Resource Allocation in a Client/Server System for Massive Multi
2014
The creation of a Massive Multi-Player On-line Game (MMOG) has significant costs, such as maintenance of server rooms, server administration, and customer service. The capacity of servers in a client/server MMOG is hard to scale and cannot adjust quickly to peaks in demand while maintaining the required response time. To handle these peaks in demand, we propose to employ users ’ computers as secondary servers. The introduction of users ’ computers as secondary servers allows the performance of the MMOG to support an increase in users. Here, we consider two cases. First, for the minimization of the response times from the server, we develop and implement five static heuristics to implement a secondary server scheme that reduces the time taken to compute the state of the MMOG. Second, for our study on fairness, the goal of the heuristics is to provide a “fair ” environment for all the users (in terms of similar response times), and to be “robust ” against the uncertainty of the number...
A small world based overlay network for improving dynamic load-balancing
Journal of Systems and Software, 2015
Load-balancing algorithms play a key role in improving the performance of distributed-computing-systems that consist of heterogeneous nodes with different capacities. The performance of load-balancing algorithms and its convergence-rate deteriorate as the number-of-nodes in the system, the network-diameter, and the communication-overhead increase. Moreover, the load-balancing technical-factors significantly affect the performance of rebalancing the load among nodes. Therefore, we propose an approach that improves the performance of load-balancing algorithms by considering the load-balancing technical-factors and the structure of the network that executes the algorithm. We present the design of an overlay network, namely, functional small world (FSW) that facilitates efficient load-balancing in heterogeneous systems. The FSW achieves the efficiency by reducing the number-of-nodes that exchange their information, decreasing the network diameter, minimizing the communication-overhead, and decreasing the time-delay results from the tasks remigration process. We propose an improved load-balancing algorithm that will be effectively executed within the constructed FSW, where nodes consider the capacity and calculate the average effective-load. We compared our approach with two significant diffusion methods presented in the literature. The simulation results indicate that our approach considerably outperformed the original neighborhood approach and the nearest neighbor approach in terms of response time, throughput, communication overhead, and movements cost.
Load Balancing for Response Time
Journal of Algorithms, 2000
A centralized scheduler must assign tasks to servers, processing on-line a sequence of task arrivals and departures. Each task runs for an unknown length of time, but comes with a weight that measures resource utilization per unit time. The response time of a server is the sum of the weights of the tasks assigned to it. The goal is to minimize the maximum response time, i.e., load, of any server. Previous papers on online load balancing have generally concentrated only on keeping the current maximum load on an on-line server bounded by some function of the maximum o-line load ever seen. Our goal is to keep the current maximum load on an on-line server bounded by a function of the current o-line load. Thus our algorithms are not skewed by transient peaks, and provide bounded response time at all point in the run. To achieve this, the scheduler must occasionally reassign tasks, in an attempt to decrease the maximum load. We study several variants of load balancing, including identical machines, related machines, restricted assignment tasks, and virtual circuit routing. In each case, only a limited amount of reassignment is used but the load is kept substantially lower than possible without reassignment.
Server selection with delay constraints for online games
2010 IEEE Globecom Workshops, 2010
Improving latency is the key to a successful online game-playing experience. With the use of multiple servers along with a well-provisioned network it is possible to reduce the latency. Given a network of servers, game clients, and a desired delay bound, we have designed algorithms to determine the subnetwork of servers whose cardinality is minimal. We have considered the cases wherein the subnetwork architecture is a client-server and a peer-to-peer. We have also provided exhaustive empirical evaluations of our algorithms and compared their performance with the optimum. Experimental results show that our polynomial-time algorithms could find good solutions quickly.
Deploying a Zone-Based Massively Multiplayer Online Game on a Congested Network
International Journal of Information Systems and Supply Chain Management, 2012
A massively multiplayer online game often deploys dozens or hundreds of servers to support millions of players around the world. A slow response time stemming from an ill-designed network infrastructure could render the game noncompetitive. To establish an efficient infrastructure, the authors focus on selecting host facilities on backbone network nodes, assigning client clusters to these facilities, and determining the required capacity for each host site. An exact solution approach is obtained from solving a minimum cost set-covering problem. The efficiency of the solution approach, in comparison with a greedy heuristic, is also reported.
Scalable Load Balancing: Distributed Approaches and the Packing Model
2019
Periodical load balancing heuristics are employed in parallel iterative applications to assure the effective use of high performance computing platforms. Work stealing is one of the most widely used load balancing techniques, but it is not the most friendly for iterative applications. Optimal mapping of tasks to machines, while minimizing overall makespan, is regarded as an NP-Hard problem; so suboptimal heuristics are used to schedule these tasks in feasible time. Among the existing approaches, distributed load balancers are the most scalable for iterative applications and have much to profit from work stealing. In this work, we propose the discretization of application workload for load balancing, as well as two distributed load balancers: PackDrop, which is based on constrained work diffusion; and PackSteal, which is based on work stealing. Our algorithms group tasks in batches before migration, creating packs of homogeneous load to make scheduling decisions in an informed and ti...
Related papers
Resource Allocation in a Client/Server System for Massive Multi-Player Online Games
IEEE Transactions on Computers, 2014
The creation of a Massive Multi-Player On-line Game (MMOG) has significant costs, such as maintenance of server rooms, server administration, and customer service. The capacity of servers in a client/server MMOG is hard to scale and cannot adjust quickly to peaks in demand while maintaining the required response time. To handle these peaks in demand, we propose to employ users' computers as secondary servers. The introduction of users' computers as secondary servers allows the performance of the MMOG to support an increase in users. Here, we consider two cases: first, for the minimization of the response times from the server, we develop and implement five static heuristics to implement a secondary server scheme that reduces the time taken to compute the state of the MMOG. Second, for our study on fairness, the goal of the heuristics is to provide a "fair" environment for all the users (in terms of similar response times), and to be "robust" against the uncertainty of the number of new players that may join a given system configuration. The number of heterogeneous secondary servers, conversion of a player to a secondary server, and assignment of players to secondary servers are determined by the heuristics implemented in this study.
Scaling Internet services by dynamic allocation of connections
Integrated Network Management VI. Distributed Management for the Networked Millennium. Proceedings of the Sixth IFIP/IEEE International Symposium on Integrated Network Management. (Cat. No.99EX302)
Network Dispatcher ND is a software tool that routes" TCP connections to multiple TCP servers that share their workload. It exports a set of virtual IP addresses that are concealed and shared by the servers. It implements a novel dynamic load-sharing algorithm for allocation of TCP connections among servers according to their real-time load and responsiveness. ND forwards packets to the servers without performing any TCP IP header translations, consequently outgoing server-to-client packets are not handled, and can follow a separate network route to the clients. Its allocation method was proven to be e cient in live tests, supporting Internet sites that served millions of TCP connections per hour. This paper describes the load management features of ND.
Distributed and Dynamic Resource Allocation for Delay Sensitive Network Services
IEEE GLOBECOM 2008 - 2008 IEEE Global Telecommunications Conference, 2008
In this paper, we present a distributed algorithm to dynamically allocate the available resources of a service-oriented network to delay sensitive network services. We use a utilitybased framework to differentiate services based on both their relative profitability and quality-of-service requirements. Our performance metric is the end-to-end delay that a service class experiences in the network. We use network calculus to obtain a deterministic upper bound of this delay and we incorporate this information into our optimization problem formulation. We leverage a moving average control scheme to capture traffic shifts in real time, which makes our solution to react adaptively to traffic dynamics. Finally, we evaluate our system using real traces of instant messaging service traffic.
load balance
Obtaining maximum throughput across a network or a mesh through optimal load balancing is known to be an NP-hard problem. Designing efficient load balancing algorithms for networks in the wireless domain becomes an especially challenging task due to the limited bandwidth available. In this paper we present heuristic algorithms for load balancing and maximum throughput scheduling in Wireless Mesh Networks with stationary nodes. The goals are to (a) improve the network throughput through admissibly optimal distribution of the network traffic across the wireless links, (b) ensure that the scheme is secure, and (c) ensure fairness to all nodes in the network for bandwidth allocation. The main consideration is the routing of non-local traffic between the nodes and the destination via multiple Internet gateways. Our schemes split an individual node's traffic to the Internet across multiple gateways that are accessible from it. Simulation results show that this approach results in marked increase in average network throughput in moderate to heavy traffic scenarios. We also prove that in our algorithm it is very difficult for an adversary to block a fraction of a node's available paths, making it extremely hard to compromise all traffic from a node. Simulation results also show that our scheme is admissibly fair in bandwidth allocation even to nodes with longest paths to the gateway nodes.
Optimal Client-Server Assignment for Internet Distributed Systems
IEEE Transactions on Parallel and Distributed Systems, 2013
We investigate an underlying mathematical model and algorithms for optimizing the performance of a class of distributed systems over the Internet. Such a system consists of a large number of clients who communicate with each other indirectly via a number of intermediate servers. Optimizing the overall performance of such a system then can be formulated as a client-server assignment problem whose aim is to assign the clients to the servers in such a way to satisfy some prespecified requirements on the communication cost and load balancing. We show that 1) the total communication load and load balancing are two opposing metrics, and consequently, their tradeoff is inherent in this class of distributed systems; 2) in general, finding the optimal client-server assignment for some pre-specified requirements on the total load and load balancing is NP-hard, and therefore; 3) we propose a heuristic via relaxed convex optimization for finding the approximate s olution. Our simulation results indicate that the proposed algorithm produces superior performance than other heuristics, including the popular Normalized Cuts algorithm.
A Prolific Scheme for Load Balancing Relying on Task Completion Time
International Journal of Electrical and Computer Engineering (IJECE), 2018
In networks with lot of computation, load balancing gains increasing significance. To offer various resources, services and applications, the ultimate aim is to facilitate the sharing of services and resources on the network over the Internet. A key issue to be focused and addressed in networks with large amount of computation is load balancing. Load is the number of tasks"t" performed by a computation system. The load can be categorized as network load and CPU load. For an efficient load balancing strategy, the process of assigning the load between the nodes should enhance the resource utilization and minimize the computation time. This can be accomplished by a uniform distribution of load of to all the nodes. A Load balancing method should guarantee that, each node in a network performs almost equal amount of work pertinent to their capacity and availability of resources. Relying on task subtraction, this work has presented a pioneering algorithm termed as E-TS (Efficient-Task Subtraction). This algorithm has selected appropriate nodes for each task. The proposed algorithm has improved the utilization of computing resources and has preserved the neutrality in assigning the load to the nodes in the network.
Load Balancing for Optimal Sharing of Network Bandwidth
2014
Peers participating in a DHT are able to balance their loads in the virtual servers. In decentralized load balance algorithm in DHT the peers which are participating should be Asymmetric which introduces another load imbalance problem. In our paper, the symmetric load balancing algorithm where each peers independently reallocates. Our proposal exhibits analytical performance in terms of load balance factor and the algorithmic convergence rate and it will not introduce any load imbalance problem due to algorithmic workload.
Load balancing in distributed computing over wireless LAN: effects of network delay
2005
A complete analytical solution to one-shot load balancing in a two-node distributed system is presented and verified experimentally. The model takes into account uncertainty in network delays and the variability in the processors' speeds. The system consisting of two Transmeta processors connected through a wireless LAN, utilizing an 802.11b access point, is used as the testbed. The statistics of the network delays are estimated experimentally and used in the analytical model to show the effect of the balancing instant and gain on the overall completion times of various workloads. The analytical results are verified by comparing them to actual experiments and Monte-Carlo simulation. The effect of the communication delay on the choice of the gain and the balancing instant is clearly observed. Finally, the model is used to determine the optimal gain and balancing instant for different initial load distributions.
Dynamic Multi-User Load Balancing in Distributed Systems
21st IEEE International Parallel and Distributed Processing Symposium (IPDPS 2007), Long Beach, California, USA, pp. 1-10, 26-30 March 2007, 2007
In this paper, we review two existing static load balancing schemes based on M/M/1 queues. We then use these schemes to propose two dynamic load balancing schemes for multi-user (multi-class) jobs in heterogeneous distributed systems. These two dynamic load balancing schemes differ in their objective. One tries to minimize the expected response time of the entire system while the other tries to minimize the expected response time of the individual users. The performance of the dynamic schemes is compared with that of the static schemes using simulations with various loads and parameters. The results show that, at low communication overheads, the dynamic schemes show superior performance over the static schemes. But as the overheads increase, the dynamic schemes (as expected) yield similar performance to that of the static schemes.
Scalable approaches to load sharing in the presence of multicasting
Computer Communications, 1995
This work examines policies based on multicasting for load sharing in a local area network environment.
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.