The Impact of the Implementation Cost of Replication in Data Grid Job Scheduling

Abstract

Data Grids deal with geographically-distributed large-scale data-intensive applications. Schemes scheduled for data grids attempt to not only improve data access time, but also aim to improve the ratio of data availability to a node, where the data requests are generated. Data replication techniques manage large data by storing a number of data files efficiently. In this paper, we propose centralized dynamic scheduling strategy-replica placement strategies (CDSS-RPS). CDSS-RPS schedule the data and task so that it minimizes the implementation cost and data transfer time. CDSS-RPS consists of two algorithms, namely (a) centralized dynamic scheduling (CDS) and (b) replica placement strategy (RPS). CDS considers the computing capacity of a node and finds an appropriate location for the job. RPS attempts to improve file access time by using replication on the basis of number of accesses, storage capacity of a computing node, and response time of a requested file. Extensive simulations are carried out to demonstrate the effectiveness of the proposed strategy. Simulation results demonstrate that the replication and scheduling strategies improve the implementation cost and average access time significantly.

Minimizing data access latency in data grids by neighborhood‐based data replication and job scheduling

“…The PRS algorithm considers the number of accesses and storage capacity of computing nodes for replica placement. 18 A series of previous works propose replication and scheduling algorithms for specific architectures. Hierarchical architecture is trendy among researchers.…”

“…Parameter RFC is obtained from Equation (17) by adding the time taken to transfer the file and the waiting time in the storage disk queue. The DTC in Equation 17is the data transfer cost which is given by Equation (18) where S f is the size of requested file f, band i,k is the bandwidth of the connection links between requesting site i and replica site k, DS k is the disk speed of site k, and lat i,k is the latency between the links caused by different factors like the distance and traffic of routers. The latency is estimated by the time taken to transfer an empty packet from site i to the site k. DWQ k is the waiting time in the storage disk queue of the site k, and is calculated using Equation (19) where S fz is the size of the requested file waiting in the storage queue of site k, and n is the number of file-access requests waiting in the storage disk queue.…”

“…• We propose two new job scheduling algorithms, called neighborhood-based job-scheduling strategy (NJS) and parallel neighborhood-based job-scheduling strategy (PNJS). The NJS and PNJS algorithms improve the previous algorithms like Random, 12 shortest Queue, 13 Queue access cost, 13 DataPresent, 14 ESS, 15 CSS, 16,17 CDS, 18 and proposed job scheduling algorithm in Meddeber and Yagoubi. 19 The NJS algorithm employs a hierarchical search procedure to reduce the search time.…”

A novel cost-based replica server placement for optimal service quality in cloud-based content delivery network

“…The outcome exhibits satisfactory predictive accuracy with sufficient optimization of channel capacity. The study towards cost estimation is carried out by Nazir et al [20] while performing job scheduling over grid system. The work introduces a unique dynamic scheduling policy that is centralized towards replica placement.…”