Astronomy education has long faced challenges with cross-platform deployment complexities of data processing tools and the difficulty in meeting diverse teaching scenario requirements. To address these issues, this study proposes a private cloud-based virtual laboratory solution. By integrating KVM virtualization and automated orchestration technologies, we developed a pre-configured environment generation system supporting multiple Linux distributions (Ubuntu/CentOS/Scientific Linux). The system employs a hierarchical data management architecture, combining template-embedded datasets, shared storage, and self-managed download zones, while supporting three core teaching scenarios: live classroom demonstrations, coursework, and exam assessments.

From 2015 to 2024, this system was implemented in 5 courses, including Multi-Wavelength Astronomical Observation and Data Processing at the University of Chinese Academy of Sciences and Practical Astronomical Observation at Sun Yat-sen University.

Empirical results demonstrate:
1) Student environment setup time was reduced from an average of 4 weeks to 30 minutes;
2) The system supports concurrent operation of 120 virtual machines during peak usage periods;
3) Pre-configured templates exhibit cross-program reusability, with graduate-level course templates being adapted for doctoral dissertation research.

This study demonstrates that the proposed design effectively reduces technical barriers in computational workflows and provides a scalable framework for teaching innovation in data-driven disciplines.

Keywords: Virtual Observatory, Astronomy Education, Cloud Computing, Pre-Configured Environments, Teaching Scenarios.