Parallel processing for data retrieval in ODOO enterprise resource planning reporting system / Roua Abdelmuniem Osman Alhag Eisa

Roua Abdelmuniem , Osman Alhag Eisa (2021) Parallel processing for data retrieval in ODOO enterprise resource planning reporting system / Roua Abdelmuniem Osman Alhag Eisa. Masters thesis, Universiti Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (180Kb)
	PDF (Thesis M.A.) Download (1736Kb)

Abstract

Reporting process in Enterprise Resource Planning (ERP) system plays an important role, as different information from different processes can be merged to generate reports. Management can use these reports for providing key value indicators for progress assessment, as well as the identification of poor business performance and the formulation of strategies to eliminate them. Odoo framework, previously known as OpenERP, is the most commonly installed open source ERP system worldwide. During the ERP system lifetime massive data generated from the daily operations, most implemented open source ERP systems such as the Odoo framework are using Relational Database Management System (RDBMS) as data storage, while the amount of the data increases this traditional data analysis, processing and storage technologies are not capable enough to store and/or process a large amount of data effectively and the performance became an issue as the relational database applies sequential data processing. This performance latency has an implication on overall system performance, concurrent users’ sessions, business processing, and report processing which all affect organization processes and decision making to achieve business goals. Report processing time increases while the number of data increases due to data retrieving from the relational database, where the more data are processed; the more time it needs to generate a report. This research aims to solve Odoo’s reporting latency problem, where the proposed solution is to import data from the Odoo database and store it in NoSQL data storage to perform parallel data processing to generate the required report faster than the existing approaches to generating the same report. The applied research methodology comprises several steps which include a literature review that discusses the previous ERP system comparisons, existing reporting approaches and the successful deployment of parallel data processing in various domains. Another step is preliminary experiment conduct to compare the performance of generating sale orders report using the existed approaches, the remain steps discuss the design, development and evaluation of the research proposed solution. The research results find out that the parallel data retrieval used in the developed solution shows performance improvement over sequential data retrieval used in existed approaches. Organizations with a large scale (500000 records and above per table) can get significant reporting performance improvement which has a direct impact on an organization's processes, achieve insights into business data, forecasting, decision support and to meet business goals.

Actions (For repository staff only : Login required)