FESTIval

FESTIval evaluates spatial indices to enable quantitative, reproducible experimental comparisons of index performance in spatial database systems and Geographic Information Systems (GIS), including disk-based and flash-aware storage contexts.

Key Features:

• Support for Diverse Spatial Indices: Supports a wide range of spatial indices, including disk-based indices and flash-aware indices such as eFIND-based indices.
• User-Defined Workloads: Allows specification and execution of custom workloads for range queries and point queries.
• Comprehensive Data Schema: Implements a schema that stores detailed spatial index configurations alongside statistical results from executed workloads to enable analysis and reproducibility.
• Extensible Dataset Support: Handles any spatial dataset with spatial indices and provides default spatial datasets for experiments.
• Query Generation Algorithms: Includes algorithms to generate query windows for range queries and points for point queries.

Scientific Applications:

• Performance Evaluation: Quantitatively compares spatial index performance under different storage conditions and workload configurations in spatial database systems and GIS.
• Benchmarking Flash-Aware Indices: Validates and benchmarks flash-aware spatial index proposals, including eFIND-based indices.
• Index Selection and Optimization: Supports selection and tuning of spatial index types and parameter configurations to improve query efficiency.
• Reproducible Experiments: Enables reproducible experimental setups by recording configurations and statistical outputs for subsequent analysis.

Methodology:

Executes user-defined workloads, records index configurations and statistical results in a data schema, and uses algorithms to generate query windows for range queries and points for point queries.