An adaptive threshold based hybrid load balancing scheme with sender and receiver initiated approach using random information exchange

Abstract

The primary objective of load balancing for distributed systems is to minimize the job execution time while maximizing the resource utilization. Load balancing on decentralized systems need effective information exchange policy so that with minimum amount of communication the nodes have up to date information about other nodes in the system. Periodic, event-based and on-demand information exchange are some important policies used for the same. All these approaches involve a lot of overhead and even sometime leading toward obsolete data with the nodes if there is a delay in the updation. This work presents an adaptive threshold-based hybrid load balancing scheme with sender and receiver initiated approach (HLBWSR) using random information exchange (RIE). RIE ensures that the information is exchanged in such a way that each node in the system has up-to-date state of the other nodes with much reduced communication overhead. Further, the adaptive threshold ensures that almost an average numbers of jobs are executed by all the nodes in the system. The study of the effect of the use of RIE on sender initiated, receiver initiated and hybrid of sender and receiver initiated load balancing approach establishes the superior performance of HLBWSR among its RIE-based peers. A comparative analysis of HLBWSR, with periodic information exchange strategy, modified estimated load information scheduling algorithm and load balancing on arrival reveals its effectiveness under various test conditions. allocation and execution by transferring tasks from the heavily loaded to the lightly loaded nodes with the aim of improving the performance of the system. A typical load balancing algorithm is defined by three inherent policies [4-6]: Information policyIt deals with the amount of load information that is required by the job placement decision-makers for efficient job transfer. Transfer policyThe current load of the host and the size of the job under consideration determine the conditions under which the job should be transferred from heavily loaded node to lightly loaded node. Placement policyThe processing element with the least number of jobs and closest to the sender node is the best recipient. The affinity for processing element to receive job is the deciding factor during job placement.Load balancing algorithms are broadly characterized as static and dynamic. Static algorithms make use of prior information for load balancing while dynamic load balancing algorithms use runtime system-state information to make load balancing decisions. Dynamic algorithms have the potential to outperform static algorithms as they make use of system-state information during the runtime to enhance the quality of their decisions. However, dynamic algorithms incur more overhead than their static counterparts as they have to collect, store, and analyze state information. This overhead can be overcome if reduction in the response time incurred because of the redistribution outperforms the overhead [3,4,6].According to the degree of centralization, load sch...

Resource discovery for distributed computing systems: A comprehensive survey

“…Harchol et al, [112] discussed some well-known natural algorithms such as ALG-Flooding [246], Swamping [246,247], Random Pointer Jump [246] and presented Name-Dropper [112] to perform a synchronous resource discovery. Table 5, compares these algorithms to other approaches on three performance measures: time complexity, pointer complexity, and message complexity.…”

Load balancing in cloud computing environments based on adaptive starvation threshold

“…In receiver‐initiated algorithms, the load balancing activity is initiated from an underloaded node (receiver) which attempts to get tasks from an overloaded one. The symmetrically‐initiated algorithm is a combination of the two previous approaches. It requires both senders and receivers to look for suitable nodes to send/receive loads to/from.…”

Improving utilization of heterogeneous clusters

“…One of the most important challenges in successfully leveraging the performance of these large systems is to consistently distribute the load among the available resources, proportionally to their computing capacity [4][5][6]. This has been traditionally attempted through dynamic algorithms that are able to adapt to the varying requirements of the workloads.…”