View in EDS

A workload-driven and SLA-aware model for relational to document store schema transformation.

Saved in:
Bibliographic Details
Title: A workload-driven and SLA-aware model for relational to document store schema transformation.
Authors: Soni, Kanika1 (AUTHOR), Sachdeva, Shelly1 (AUTHOR) shellysachdeva@nitdelhi.ac.in
Source: Knowledge & Information Systems. 4/24/2026, Vol. 68 Issue 1, p1-71. 71p.
Subjects: Service level agreements, Nonrelational databases, Scalability, Model validation, Information services
Abstract: With the rise of big data, numerous large-scale applications have shifted away from relational databases (RDB) to NoSQL stores owing to enhanced performance and flexibility. However, migrating from RDB to NoSQL stores involves tedious schema transformation. This is because existing NoSQL schema design approaches mostly rely on rules of thumb or guidelines for manually choosing a good schema. We have proposed a novel workload-driven and service level agreement (SLA)-aware model for relational to document store schema transformation. The proposed model is simultaneously automatic, workload-driven, and SLA-aware. The proposed model has three phases: model input, midway transformations, and model output. The proposed model begins with an extended entity-relationship schema along with workload information and SLA specifications as inputs. The paper proposes two algorithms for the midway transformations phase: (i) entity and relationship transformations, and (ii) workload-driven and SLA-aware refinements and transformations. The third phase includes generating the document store schema as model output. To validate the effectiveness of our research, we conducted an experimental evaluation using a case study in the e-commerce sector. The performance of the proposed model (P) is compared with the four existing workload-driven models, namely, GAF(G), UAF (U), QPG (Q), and Hypergraph (H), for relational to document store schema transformation. Our results show that the proposed model consistently improves query execution time, reduces read and write latency, and enhances aggregation pipeline performance. Finally, our results illustrate that our proposed model requires less storage space, is highly scalable, and demonstrates improved throughput and latency compared to existing models. [ABSTRACT FROM AUTHOR]
Copyright of Knowledge & Information Systems is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:With the rise of big data, numerous large-scale applications have shifted away from relational databases (RDB) to NoSQL stores owing to enhanced performance and flexibility. However, migrating from RDB to NoSQL stores involves tedious schema transformation. This is because existing NoSQL schema design approaches mostly rely on rules of thumb or guidelines for manually choosing a good schema. We have proposed a novel workload-driven and service level agreement (SLA)-aware model for relational to document store schema transformation. The proposed model is simultaneously automatic, workload-driven, and SLA-aware. The proposed model has three phases: model input, midway transformations, and model output. The proposed model begins with an extended entity-relationship schema along with workload information and SLA specifications as inputs. The paper proposes two algorithms for the midway transformations phase: (i) entity and relationship transformations, and (ii) workload-driven and SLA-aware refinements and transformations. The third phase includes generating the document store schema as model output. To validate the effectiveness of our research, we conducted an experimental evaluation using a case study in the e-commerce sector. The performance of the proposed model (P) is compared with the four existing workload-driven models, namely, GAF(G), UAF (U), QPG (Q), and Hypergraph (H), for relational to document store schema transformation. Our results show that the proposed model consistently improves query execution time, reduces read and write latency, and enhances aggregation pipeline performance. Finally, our results illustrate that our proposed model requires less storage space, is highly scalable, and demonstrates improved throughput and latency compared to existing models. [ABSTRACT FROM AUTHOR]
ISSN:02191377
DOI:10.1007/s10115-026-02759-5