Study on Cloud Service Mode of Agricultural Information Institutions

Xiaorong Yang, Nengfu Xie, Dan Wang, Lihua JiangIFIP 2014 · DOI · OpenAlex

Abstract

Agricultural information institutions manage extensive specialized databases but often struggle with fragmented hardware and software environments. This paper proposes a specialized three-layer cloud service framework—comprising Infrastructure as a Service (IaaS), an Automation Management System, and a Service Layer (PaaS, DaaS, SaaS)—to modernize these institutions. By transitioning to this structured model, institutions can enhance resource utilization, reduce maintenance costs, and provide streamlined data access. A case study of the National Science and Technology Library (NSTL) in China demonstrates the model's effectiveness in managing over 180 million records and supporting dozens of service stations.

Introduction

Agricultural information institutions, including large-scale libraries and research centers, are custodians of massive datasets. However, these resources are frequently trapped in "legacy" silos. The authors identify three primary challenges: low resource utilization of existing hardware, high maintenance costs due to disparate systems, and inefficient service delivery that hinders external integration. To address these, a cloud-based architectural shift is proposed to move from fragmented environments to a unified service model.

The Three-Layer Cloud Service Model

The core of the proposal is a three-layer architecture designed to unify disparate resources into a cohesive service platform. This model integrates infrastructure, management, and diverse service delivery modes.

  1. Infrastructure Layer (IaaS): This foundation utilizes virtualization technology to transform physical servers, storage, and network equipment into a flexible "virtual resource pool."
  2. Automation Management System Layer: Serving as the middleware, this layer handles critical administrative tasks such as user registration, service scheduling, unified authentication, load management, and billing.
  3. Service Layer: This top tier provides specialized functional modes:
  4. PaaS (Platform as a Service): An integrated development environment for building and testing applications.
  5. DaaS (Data as a Service): A specialized mode providing open standard interfaces for external users to embed agricultural data into their own platforms.
  6. SaaS (Software as a Service): Web-based tools for intelligence analysis and knowledge retrieval.
The proposed three-layer cloud service framework for agricultural information institutions, illustrating the relationship between IaaS, the management system, and the service modes (PaaS, DaaS, SaaS).

Case Study: National Science and Technology Library (NSTL)

The model was validated using the holdings of the NSTL in China, which includes over 40 specialized databases and more than 180 million individual records. The implementation demonstrated that a cloud-based approach could effectively manage large-scale scientific and agricultural data across a distributed network.

Key Insights

A significant contribution of this model is the emphasis on Data as a Service (DaaS). While general cloud models often focus on SaaS, agricultural institutions prioritize the data itself as the primary product. DaaS allows smaller, underfunded institutions to leverage high-value datasets without the burden of hosting them locally. Furthermore, the authors suggest that SaaS should be flexible, offering "software only," "software + data," or "collaborative sharing" modes to meet diverse user needs.

Results and Impact

The deployment at NSTL resulted in several operational improvements:

  • Operational Efficiency: Rapid deployment of new applications via virtual machines ensured system continuity and reduced manual effort.
  • Scalability: The PaaS layer successfully supported 9 member institutions and 50 service stations across China, centralizing management while decentralizing access.
  • Democratized Access: Standardized DaaS interfaces allowed resource-poor organizations to access NSTL’s vast record collection more conveniently.

Limitations and Future Work

A major technical hurdle remains: hardware homogeneity. Current cloud platforms often require identical physical equipment, which can lead to "brand lock-in" and prevent the reuse of legacy hardware. Future work should focus on a "pilot first" transition strategy and the cultivation of specialized technical personnel to manage evolving cloud architectures and prevent obsolescence.

Summary updated

BibTeX
@inproceedings{yang2014study,
  author = {Xiaorong Yang and Nengfu Xie and Dan Wang and Lihua Jiang},
  title = {{Study on Cloud Service Mode of Agricultural Information Institutions}},
  booktitle = {IFIP},
  year = {2014},
  doi = {10.1007/978-3-642-54344-9\_57},
}
Figure