Post Processing: Add a timeseries report generator

[harriscr]

Purpose of Change

This PR implements a comprehensive time-series reporting capability for the CBT (Ceph Benchmarking Tool) post-processing framework. The implementation adds the ability to parse FIO time-series data, generate time-series plots (IOPS, bandwidth, and latency over time), and create reports in both GitHub markdown and PDF formats.

Key Features

• FIO Time-Series Parser: Parses FIO time-series log data into a common intermediate format
• Time-Series Plotters: Specialized plotters for IOPS, bandwidth, and latency metrics over time
• Timestamp Alignment: Aligns timestamps across multiple data sources for accurate comparison
• Report Generation: Creates time-series reports in markdown and PDF formats
• Common Intermediate Format: Analogous to the existing hockey-stick curve format, enabling a single plotter to handle time-series data from multiple benchmark tools

Design Approach

The implementation follows existing design patterns in the codebase:

• Uses axis plotter sub-classes for each plot type, allowing multiple data lines on a single plot
• Extends the common intermediate format pattern previously implemented for hockey-stick curves
• Maintains consistency with the existing post-processing architecture
• Modular design enables future extension to other benchmark tools beyond FIO

Automated Testing

• New Unit Tests: Added comprehensive unit tests covering all new functionality
• Extended Coverage: Enhanced test coverage for modified existing files
• Code Quality Checks: All changes pass:
  • Ruff (linting)
  • Pylint (static analysis)
  • Mypy (type checking)
  • Black (code formatting)

Test Files Added/Modified:

• test_time_series_report_generator.py (418 lines)
• test_time_series_plotter.py (536 lines)
• test_time_series_output_formatter.py (248 lines)
• test_time_series_formatter.py (388 lines)
• test_fio_log_parser.py (294 lines)
• test_timestamp_aligner.py (354 lines)
• test_fio_benchmarks.py (947 lines)
• test_run_result.py (462 lines)
• Plus updates to 15+ existing test files

Manual Testing

A time series report was generated from a CBT RBD FIO benchmark run:
timeseries_performance_report_260602_141547.pdf

A regression test was also run, producing a comparison report and a simple report to verify the new code still produces identical reports

Additional Notes

• IBM Bob v1.02 and v1.03 were used to assist with these changes
• New tool added: tools/generate_timeseries_performance_report.py for easy report generation
• Updated requirements.txt with necessary dependencies
• Comprehensive documentation maintained throughout the codebase

Related Documentation

• Implementation follows patterns described in existing post_processing README files
• New parsers documented in post_processing/parsers/ directory
• Plotter usage documented in post_processing/plotter/README.md