README.md

COMP4801_FYP_EVALUATION_Post_Benchmark_Processor

FYP: 22013

FYP Team

Student: SRIVASTAVA Dhruv (3035667792)

Supervisor: Dr. Heming Cui

Description

Post Benchmark Processor for HTAP Graph Database Benchmark.

Usage Instructions

Run the Post Benchmark Processor after executing the benchmark with freshness- score mode enabled.

Fill in the configuration details in build_config.py and execute it to generate a configuration file for the Post-Benchmark Processor. Finally, execute the processor by running:

python PBP000_main_processor.py <JSON configuration file>

The output will appear in the target output directory as two folders:

• Freshness_Score containing freshness score of each analytical query per analytical query driver thread in a dedicated file named thread_<x>.csv, with <x> following the thread ID of the analytical thread. The folder will also contain a combined_freshness_scores.csv file containing freshness score of each analytical query in every analytical query driver thread, sorted in ascending order of the end_time of each analytical query.

• Query_Latency containing the latencies of each query sent out by the hybrid query driver for each query type put in a dedicated file for each query type named after the query type TQ_*.csv/AQ_*.csv. The queries are sorted in ascending order of the end_time in each file.

Module Components Description

File Name	Description
PBP000_main_processor.py	Script to execute all post-benchmark processors
PBP001_freshness_score_processor.py	Contains the Freshness Score Processor
PBP002_query_latency_processor.py	Contains the Query Latency Processor
PBP003_query_data.py	Contains class definitions to encapsulate query statistics for processing in the 2 processors
build_config.py	Helper script to generate the configuration file for the post benchmark processor