Add OMB & K6 experiment files

Author: olefb

test-k6-high-throughput.js appendix/k6/test-k6-high-throughput.jstest-k6-high-throughput.js renamed to appendix/k6/test-k6-high-throughput.js

@@ -0,0 +1,76 @@
+import json
+import matplotlib.pyplot as plt
+import sys
+import os
+
+def load_metrics(filename):
+    """
+    Parses the flat-array JSON format from OpenMessaging Benchmark
+    """
+    if not os.path.exists(filename):
+        print(f"Error: File {filename} not found.")
+        return None
+
+    with open(filename, 'r') as f:
+        data = json.load(f)
+
+    # We generate a 'time' axis based on the number of data points
+    count = len(data.get('publishRate', []))
+    time_axis = list(range(1, count + 1))
+
+    return {
+        'driver': data.get('driver', 'Unknown'),
+        'time': time_axis,
+        'throughput': data.get('publishRate', []),
+        'p99_pub_latency': data.get('publishLatency99pct', []),
+        'p99_e2e_latency': data.get('endToEndLatency99pct', [])
+    }
+
+def plot_comparison(rabbit_file, kafka_file):
+    rabbit_data = load_metrics(rabbit_file)
+    kafka_data = load_metrics(kafka_file)
+
+    if not rabbit_data or not kafka_data:
+        return
+
+    # Create a figure with 2 subplots (Latency and Throughput)
+    fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(12, 10))
+
+    # Plot 1: P99 Publish Latency (Responsiveness)
+    ax1.plot(rabbit_data['time'], rabbit_data['p99_pub_latency'],
+             label='RabbitMQ', marker='o', linestyle='-', color='orange')
+    ax1.plot(kafka_data['time'], kafka_data['p99_pub_latency'],
+             label='Kafka', marker='x', linestyle='--', color='blue')
+
+    ax1.set_title('Hypothesis 1: P99 Publish Latency (Lower is Better)')
+    ax1.set_ylabel('Latency (ms)')
+    ax1.set_xlabel('Benchmark Interval')
+    ax1.grid(True)
+    ax1.legend()
+
+    # Plot 2: Throughput Stability
+    ax2.plot(rabbit_data['time'], rabbit_data['throughput'],
+             label='RabbitMQ', color='orange')
+    ax2.plot(kafka_data['time'], kafka_data['throughput'],
+             label='Kafka', color='blue')
+
+    # Add a line for the Target Rate (1000)
+    ax2.axhline(y=1000, color='r', linestyle=':', label='Target Rate (1000 msg/s)')
+
+    ax2.set_title('Throughput Stability (Target: 1000 msg/s)')
+    ax2.set_ylabel('Messages / Second')
+    ax2.set_xlabel('Benchmark Interval')
+    ax2.grid(True)
+    ax2.legend()
+
+    plt.tight_layout()
+    output_file = 'benchmark_comparison_simple.png'
+    plt.savefig(output_file)
+    print(f"Successfully generated comparison graph: {output_file}")
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        print("Usage: python analyze_results_1000.py <rabbitmq_result.json> <kafka_result.json>")
+        sys.exit(1)
+
+    plot_comparison(sys.argv[1], sys.argv[2])

test-k6.js appendix/k6/test-k6.jstest-k6.js renamed to appendix/k6/test-k6.js

@@ -0,0 +1,75 @@
+import json
+import matplotlib.pyplot as plt
+import sys
+import os
+
+def load_metrics(filename):
+    """
+    Parses the flat-array JSON format from OpenMessaging Benchmark
+    """
+    if not os.path.exists(filename):
+        print(f"Error: File {filename} not found.")
+        return None
+
+    with open(filename, 'r') as f:
+        data = json.load(f)
+
+    # We generate a 'time' axis based on the number of data points
+    count = len(data.get('publishRate', []))
+    time_axis = list(range(1, count + 1))
+
+    return {
+        'driver': data.get('driver', 'Unknown'),
+        'time': time_axis,
+        'throughput': data.get('publishRate', []),
+        'p99_pub_latency': data.get('publishLatency99pct', []),
+        'p99_e2e_latency': data.get('endToEndLatency99pct', [])
+    }
+
+def plot_comparison(rabbit_file, kafka_file):
+    rabbit_data = load_metrics(rabbit_file)
+    kafka_data = load_metrics(kafka_file)
+
+    if not rabbit_data or not kafka_data:
+        return
+
+    # Create a figure with 2 subplots (Latency and Throughput)
+    fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(12, 10))
+
+    # Plot 1: P99 Publish Latency (Responsiveness)
+    ax1.plot(rabbit_data['time'], rabbit_data['p99_pub_latency'],
+             label='RabbitMQ', marker='o', linestyle='-', color='orange')
+    ax1.plot(kafka_data['time'], kafka_data['p99_pub_latency'],
+             label='Kafka', marker='x', linestyle='--', color='blue')
+
+    ax1.set_title('Hypothesis 1: P99 Publish Latency (Lower is Better)')
+    ax1.set_ylabel('Latency (ms)')
+    ax1.set_xlabel('Benchmark Interval')
+    ax1.grid(True)
+    ax1.legend()
+
+    # Plot 2: Throughput Stability
+    ax2.plot(rabbit_data['time'], rabbit_data['throughput'],
+             label='RabbitMQ', color='orange')
+    ax2.plot(kafka_data['time'], kafka_data['throughput'],
+             label='Kafka', color='blue')
+
+    ax2.axhline(y=10000, color='r', linestyle=':', label='Target Rate (10000 msg/s)')
+
+    ax2.set_title('Throughput Stability (Target: 10000 msg/s)')
+    ax2.set_ylabel('Messages / Second')
+    ax2.set_xlabel('Benchmark Interval')
+    ax2.grid(True)
+    ax2.legend()
+
+    plt.tight_layout()
+    output_file = 'benchmark_comparison_stress.png'
+    plt.savefig(output_file)
+    print(f"Successfully generated comparison graph: {output_file}")
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        print("Usage: python analyze_results_10000.py <rabbitmq_result.json> <kafka_result.json>")
+        sys.exit(1)
+
+    plot_comparison(sys.argv[1], sys.argv[2])

appendix/omb/analyze_results_1000.py

@@ -0,0 +1,19 @@
+name: Kafka
+driverClass: io.openmessaging.benchmark.driver.kafka.KafkaBenchmarkDriver
+
+# Kafka Client configuration
+replicationFactor: 1
+topicConfig: |
+  min.insync.replicas=1
+
+commonConfig: |
+  bootstrap.servers=localhost:29092
+
+producerConfig: |
+  acks=1
+  linger.ms=1
+  batch.size=131072
+
+consumerConfig: |
+  auto.offset.reset=earliest
+  enable.auto.commit=false

appendix/omb/analyze_results_10000.py

@@ -0,0 +1,12 @@
+name: RabbitMQ
+driverClass: io.openmessaging.benchmark.driver.rabbitmq.RabbitMqBenchmarkDriver
+
+# RabbitMQ Connection configuration
+amqpUris: amqp://localhost
+
+# Topology configuration
+exchangeType: topic
+exchangeName: benchmark-exchange
+durable: true
+persistentDelivery: true # Matches Kafka's disk persistence
+messagePersistence: true