Year: 2022
Type: article
Abstract: The efficient processing of database applications on computing systems with multi-tiered persistent storage devices needs specialized algorithms to create optimal persistent storage management plans.A correct allocation and deallocation of multi-tiered persistent storage may significantly improve the overall performance of data processing.This paper describes the new algorithms that create allocation and deallocation plans for computing systems with multi-tiered persistent storage devices.One of the main contributions of this paper is an extension and application of a notation of Petri nets to describe the data flows in multi-tiered persistent storage.This work assumes a pipelined data processing model and uses a formalism of extended Petri nets to describe the data flows between the tiers of persistent storage.The algorithms presented in the paper perform linearization of the extended Petri nets to generate the optimal persistent storage allocation/deallocation plans.The paper describes the experiments that validate the data allocation/deallocation plans for multitiered persistent storage and shows the improvements in performance compared with the random data allocation/deallocation plans.
DOI: 10.25046/aj070603
Authors: Nan Noon Noon, Janusz R. Getta, Tianbing Xia
Institution: University of Wollongong
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4309972007
