feat: Performance and Scale Study

[Musaddika]

KRO Performance Framework

This directory contains the KRO Performance Framework, a comprehensive testing infrastructure for measuring KRO (Kubernetes Resource Orchestrator) performance metrics.

Overview

The framework is designed to benchmark and analyze the performance of various KRO components:

1. CRUD Operations: Create, Read, Update, Delete operations on KRO resources
2. CEL Expressions: Common Expression Language evaluation with varying complexity levels
3. ResourceGraph Operations: Graph operations with different sizes and complexities

How to test

Although this project is in the development phase, once some dependencies are resolved, it can be tested by this way:-

Quick Start

To run a quick demo of simulated performance results:

./run_performance_tests.sh --demo

Running Benchmarks

To run all benchmarks:

./kro-perf benchmark --simulation --duration 30s --workers 4 --resources 100

Or you can use the shell script with similar options:

./run_performance_tests.sh --duration 30s --workers 4 --resources 100

Running Specific Benchmarks

Run only CRUD benchmarks:

./kro-perf benchmark --type crud --simulation

Run only CEL expression benchmarks:

./kro-perf benchmark --type cel --complexity all --simulation

Run only ResourceGraph benchmarks:

./kro-perf benchmark --type resourcegraph --graph-complexity all --simulation

Analysis and Visualization

Analyze benchmark results:

./kro-perf analyze --input ./results/all_results.json --output ./results/analysis.json

Generate visualizations:

./kro-perf visualize --input ./results/analysis.json --output ./results/visualizations --charts all

Command Line Options

Benchmark Command

Option	Description	Default
--type	Benchmark type: crud, cel, resourcegraph, all	all
--duration	Duration of each benchmark	60s
--workers	Number of concurrent workers	4
--resources	Number of resources for CRUD tests	100
--namespace	Kubernetes namespace	default
--kubeconfig	Path to kubeconfig file
--simulation	Run in simulation mode	true
--complexity	CEL expression complexity: simple, medium, complex, very-complex, all	all
--graph-complexity	Graph complexity: small, medium, large, all	all
--nodes	Number of nodes for large graph tests	30
--edges	Number of edges for large graph tests	45
--output	Output directory for results	./results
--verbose	Enable verbose output	false

Analyze Command

Option	Description	Default
--input	Input results file to analyze	(required)
--output	Output file for analysis	(required)
--baseline	Baseline results for comparison
--format	Output format: json, yaml, text	json
--compare-to-last	Compare to last run	false

Visualize Command

Option	Description	Default
--input	Input analysis file to visualize	(required)
--output	Output directory for visualizations	(required)
--charts	Chart types to generate: bar, line, heatmap, all	all

[Musaddika]

Hi @a-hilaly,
I have implemented this project. If you could kindly give it a quick review, I’ll be able to proceed further with the next steps. Let me know if any changes are required.
Thanks!