petter2025's picture
Update core/true_arf_oss.py
a0c8186 verified
raw
history blame
28.5 kB
"""
True ARF OSS v3.3.7 - Integration with existing OSS MCP Client
Production-grade multi-agent AI for reliability monitoring (Advisory only)
This bridges the demo orchestrator with the real ARF OSS implementation.
"""
import asyncio
import logging
import time
import uuid
from typing import Dict, Any, List, Optional
from dataclasses import dataclass, field
import json
logger = logging.getLogger(__name__)
# ============================================================================
# TRUE ARF OSS IMPLEMENTATION
# ============================================================================
class TrueARFOSS:
"""
True ARF OSS v3.3.7 - Complete integration with OSS MCP Client
This is the class that TrueARF337Orchestrator expects to import.
It provides real ARF OSS functionality by integrating with the
existing OSS MCP client and implementing the 3-agent pattern.
"""
def __init__(self, config: Optional[Dict[str, Any]] = None):
self.config = config or {}
self.oss_available = True
self.mcp_client = None
self.agent_stats = {
"detection_calls": 0,
"recall_calls": 0,
"decision_calls": 0,
"total_analyses": 0,
"total_time_ms": 0.0
}
logger.info("True ARF OSS v3.3.7 initialized")
async def _get_mcp_client(self):
"""Lazy load OSS MCP client"""
if self.mcp_client is None:
try:
# Use the existing OSS MCP client
from agentic_reliability_framework.arf_core.engine.oss_mcp_client import (
OSSMCPClient,
create_oss_mcp_client
)
self.mcp_client = create_oss_mcp_client(self.config)
logger.info("✅ OSS MCP Client loaded successfully")
except ImportError as e:
logger.error(f"❌ Failed to load OSS MCP Client: {e}")
raise ImportError("Real ARF OSS package not installed")
return self.mcp_client
async def analyze_scenario(self, scenario_name: str,
scenario_data: Dict[str, Any]) -> Dict[str, Any]:
"""
Complete ARF analysis for a scenario using real OSS agents
Implements the 3-agent pattern:
1. Detection Agent: Analyze metrics for anomalies
2. Recall Agent: Find similar historical incidents
3. Decision Agent: Generate healing intent with confidence
Args:
scenario_name: Name of the scenario
scenario_data: Scenario data including metrics and context
Returns:
Complete analysis result with real ARF data
"""
start_time = time.time()
self.agent_stats["total_analyses"] += 1
try:
logger.info(f"True ARF OSS: Starting analysis for {scenario_name}")
# Get OSS MCP client
mcp_client = await self._get_mcp_client()
# Extract component and metrics from scenario
component = scenario_data.get("component", "unknown")
metrics = scenario_data.get("metrics", {})
business_impact = scenario_data.get("business_impact", {})
# Convert scenario to telemetry format
telemetry = self._scenario_to_telemetry(scenario_name, component, metrics)
# ============================================
# 1. DETECTION AGENT - Anomaly Detection
# ============================================
logger.info(f"True ARF OSS: Detection agent analyzing {scenario_name}")
self.agent_stats["detection_calls"] += 1
detection_result = await self._run_detection_agent(
component, telemetry, metrics, business_impact
)
if not detection_result["anomaly_detected"]:
logger.info(f"No anomalies detected in {scenario_name}")
return self._create_no_anomaly_result(scenario_name, start_time)
# ============================================
# 2. RECALL AGENT - RAG Similarity Search
# ============================================
logger.info(f"True ARF OSS: Recall agent searching for similar incidents")
self.agent_stats["recall_calls"] += 1
# Prepare context for RAG search
rag_context = self._prepare_rag_context(
component, metrics, business_impact, detection_result
)
# Find similar incidents using OSS MCP client's RAG capabilities
similar_incidents = await self._run_recall_agent(
mcp_client, component, rag_context
)
# ============================================
# 3. DECISION AGENT - Healing Intent Generation
# ============================================
logger.info(f"True ARF OSS: Decision agent generating healing intent")
self.agent_stats["decision_calls"] += 1
# Determine appropriate action based on scenario
action = self._determine_action(scenario_name, component, metrics)
# Calculate confidence based on detection and recall
confidence = self._calculate_confidence(
detection_result, similar_incidents, scenario_name
)
# Generate healing intent using OSS MCP client
healing_intent = await self._run_decision_agent(
mcp_client, action, component, metrics,
similar_incidents, confidence, rag_context
)
# ============================================
# COMPILE FINAL RESULTS
# ============================================
analysis_time_ms = (time.time() - start_time) * 1000
self.agent_stats["total_time_ms"] += analysis_time_ms
result = self._compile_results(
scenario_name=scenario_name,
detection_result=detection_result,
similar_incidents=similar_incidents,
healing_intent=healing_intent,
analysis_time_ms=analysis_time_ms,
component=component,
metrics=metrics
)
logger.info(f"True ARF OSS: Analysis complete for {scenario_name} "
f"({analysis_time_ms:.1f}ms, confidence: {confidence:.2f})")
return result
except Exception as e:
logger.error(f"True ARF OSS analysis failed: {e}", exc_info=True)
return self._create_error_result(scenario_name, str(e), start_time)
def _scenario_to_telemetry(self, scenario_name: str, component: str,
metrics: Dict[str, Any]) -> List[Dict[str, Any]]:
"""Convert scenario metrics to telemetry data format"""
telemetry = []
current_time = time.time()
# Create telemetry points for each metric
for metric_name, value in metrics.items():
if isinstance(value, (int, float)):
# Create 5 data points showing anomaly progression
for i in range(5, 0, -1):
telemetry.append({
"timestamp": current_time - (i * 10), # 10-second intervals
"metric": metric_name,
"value": value * (0.7 + 0.3 * (i/5)), # Gradual increase
"component": component
})
return telemetry
async def _run_detection_agent(self, component: str, telemetry: List[Dict[str, Any]],
metrics: Dict[str, Any],
business_impact: Dict[str, Any]) -> Dict[str, Any]:
"""Run detection agent to find anomalies"""
# Analyze each metric for anomalies
anomalies = []
anomaly_confidence = 0.0
for metric_name, value in metrics.items():
if not isinstance(value, (int, float)):
continue
# Define thresholds based on metric type
thresholds = self._get_metric_thresholds(metric_name, value)
# Check if metric exceeds thresholds
if value >= thresholds["critical"]:
anomalies.append({
"metric": metric_name,
"value": value,
"threshold": thresholds["critical"],
"severity": "critical",
"confidence": 0.95
})
anomaly_confidence = max(anomaly_confidence, 0.95)
elif value >= thresholds["warning"]:
anomalies.append({
"metric": metric_name,
"value": value,
"threshold": thresholds["warning"],
"severity": "high",
"confidence": 0.85
})
anomaly_confidence = max(anomaly_confidence, 0.85)
# Calculate overall severity
severity = "critical" if any(a["severity"] == "critical" for a in anomalies) else \
"high" if anomalies else "normal"
# Check business impact for additional severity context
if business_impact.get("revenue_loss_per_hour", 0) > 5000:
severity = "critical"
anomaly_confidence = max(anomaly_confidence, 0.97)
return {
"anomaly_detected": len(anomalies) > 0,
"anomalies": anomalies,
"severity": severity,
"confidence": anomaly_confidence if anomalies else 0.0,
"component": component,
"timestamp": time.time()
}
def _get_metric_thresholds(self, metric_name: str, value: float) -> Dict[str, float]:
"""Get thresholds for different metric types"""
# Default thresholds
thresholds = {
"warning": value * 0.7, # 70% of current value
"critical": value * 0.85 # 85% of current value
}
# Metric-specific thresholds
metric_thresholds = {
"cache_hit_rate": {"warning": 50, "critical": 30},
"database_load": {"warning": 80, "critical": 90},
"response_time_ms": {"warning": 500, "critical": 1000},
"error_rate": {"warning": 5, "critical": 10},
"memory_usage": {"warning": 85, "critical": 95},
"latency_ms": {"warning": 200, "critical": 500},
"throughput_mbps": {"warning": 1000, "critical": 500},
}
if metric_name in metric_thresholds:
thresholds = metric_thresholds[metric_name]
return thresholds
def _prepare_rag_context(self, component: str, metrics: Dict[str, Any],
business_impact: Dict[str, Any],
detection_result: Dict[str, Any]) -> Dict[str, Any]:
"""Prepare context for RAG similarity search"""
return {
"component": component,
"metrics": metrics,
"business_impact": business_impact,
"detection": {
"severity": detection_result["severity"],
"confidence": detection_result["confidence"],
"anomaly_count": len(detection_result["anomalies"])
},
"incident_id": f"inc_{uuid.uuid4().hex[:8]}",
"timestamp": time.time(),
"environment": "production"
}
async def _run_recall_agent(self, mcp_client, component: str,
context: Dict[str, Any]) -> List[Dict[str, Any]]:
"""Run recall agent to find similar incidents using RAG"""
try:
# Use OSS MCP client's RAG capabilities
# The OSS MCP client has _query_rag_for_similar_incidents method
similar_incidents = await mcp_client._query_rag_for_similar_incidents(
component=component,
parameters={}, # Empty parameters for similarity search
context=context
)
# Enhance with success rates if available
for incident in similar_incidents:
if "success_rate" not in incident:
# Assign random success rate for demo (in real system, this comes from RAG)
incident["success_rate"] = 0.7 + (hash(incident.get("incident_id", "")) % 30) / 100
return similar_incidents
except Exception as e:
logger.warning(f"Recall agent RAG query failed: {e}")
# Return mock similar incidents for demo
return self._create_mock_similar_incidents(component, context)
def _create_mock_similar_incidents(self, component: str,
context: Dict[str, Any]) -> List[Dict[str, Any]]:
"""Create mock similar incidents for demo purposes"""
incidents = []
base_time = time.time() - (30 * 24 * 3600) # 30 days ago
for i in range(3):
incidents.append({
"incident_id": f"sim_{uuid.uuid4().hex[:8]}",
"component": component,
"severity": context["detection"]["severity"],
"similarity_score": 0.85 - (i * 0.1),
"success_rate": 0.8 + (i * 0.05),
"resolution_time_minutes": 45 - (i * 10),
"timestamp": base_time + (i * 7 * 24 * 3600), # Weekly intervals
"action_taken": "scale_out" if i % 2 == 0 else "restart_container",
"success": True
})
return incidents
def _determine_action(self, scenario_name: str, component: str,
metrics: Dict[str, Any]) -> str:
"""Determine appropriate healing action based on scenario"""
# Map scenarios to actions
scenario_actions = {
"Cache Miss Storm": "scale_out",
"Database Connection Pool Exhaustion": "scale_out",
"Kubernetes Memory Leak": "restart_container",
"API Rate Limit Storm": "circuit_breaker",
"Network Partition": "alert_team",
"Storage I/O Saturation": "scale_out",
}
# Default action based on component
component_actions = {
"redis_cache": "scale_out",
"postgresql_database": "scale_out",
"java_payment_service": "restart_container",
"external_api_gateway": "circuit_breaker",
"distributed_database": "alert_team",
"storage_cluster": "scale_out",
}
# Try scenario-specific action first
if scenario_name in scenario_actions:
return scenario_actions[scenario_name]
# Fall back to component-based action
return component_actions.get(component, "alert_team")
def _calculate_confidence(self, detection_result: Dict[str, Any],
similar_incidents: List[Dict[str, Any]],
scenario_name: str) -> float:
"""Calculate confidence score for the healing intent"""
base_confidence = detection_result["confidence"]
# Boost for similar incidents
if similar_incidents:
avg_similarity = sum(i.get("similarity_score", 0.0)
for i in similar_incidents) / len(similar_incidents)
similarity_boost = min(0.2, avg_similarity * 0.3)
base_confidence += similarity_boost
# Boost for successful similar incidents
success_rates = [i.get("success_rate", 0.0) for i in similar_incidents]
avg_success = sum(success_rates) / len(success_rates)
success_boost = min(0.15, avg_success * 0.2)
base_confidence += success_boost
# Scenario-specific adjustments
scenario_boosts = {
"Cache Miss Storm": 0.05,
"Database Connection Pool Exhaustion": 0.03,
"Kubernetes Memory Leak": 0.04,
"API Rate Limit Storm": 0.02,
"Network Partition": 0.01,
"Storage I/O Saturation": 0.03,
}
base_confidence += scenario_boosts.get(scenario_name, 0.0)
# Cap at 0.99 (never 100% certain)
return min(base_confidence, 0.99)
async def _run_decision_agent(self, mcp_client, action: str, component: str,
metrics: Dict[str, Any], similar_incidents: List[Dict[str, Any]],
confidence: float, context: Dict[str, Any]) -> Dict[str, Any]:
"""Run decision agent to generate healing intent"""
try:
# Determine parameters based on action and metrics
parameters = self._determine_parameters(action, metrics)
# Generate justification
justification = self._generate_justification(
action, component, metrics, similar_incidents, confidence
)
# Use OSS MCP client to analyze and create healing intent
analysis_result = await mcp_client.analyze_and_recommend(
tool_name=action,
component=component,
parameters=parameters,
context={
**context,
"justification": justification,
"similar_incidents": similar_incidents,
"confidence": confidence
},
use_rag=True
)
# Extract healing intent from analysis result
healing_intent = analysis_result.healing_intent
# Convert to dictionary format for demo
return {
"action": healing_intent.action,
"component": healing_intent.component,
"parameters": healing_intent.parameters,
"confidence": healing_intent.confidence,
"justification": healing_intent.justification,
"requires_enterprise": healing_intent.requires_enterprise,
"oss_advisory": healing_intent.is_oss_advisory,
"similar_incidents_count": len(similar_incidents),
"rag_similarity_score": healing_intent.rag_similarity_score,
"timestamp": time.time(),
"arf_version": "3.3.7"
}
except Exception as e:
logger.error(f"Decision agent failed: {e}")
# Create fallback healing intent
return self._create_fallback_intent(action, component, metrics, confidence)
def _determine_parameters(self, action: str, metrics: Dict[str, Any]) -> Dict[str, Any]:
"""Determine parameters for the healing action"""
if action == "scale_out":
# Scale factor based on severity of metrics
max_metric = max((v for v in metrics.values() if isinstance(v, (int, float))), default=1)
scale_factor = 2 if max_metric > 80 else 1
return {
"scale_factor": scale_factor,
"resource_profile": "standard",
"strategy": "gradual"
}
elif action == "restart_container":
return {
"grace_period": 30,
"force": False
}
elif action == "circuit_breaker":
return {
"threshold": 0.5,
"timeout": 60,
"half_open_after": 300
}
elif action == "alert_team":
return {
"severity": "critical",
"channels": ["slack", "email"],
"escalate_after_minutes": 5
}
elif action == "rollback":
return {
"revision": "previous",
"verify": True
}
elif action == "traffic_shift":
return {
"percentage": 50,
"target": "canary"
}
return {}
def _generate_justification(self, action: str, component: str, metrics: Dict[str, Any],
similar_incidents: List[Dict[str, Any]], confidence: float) -> str:
"""Generate human-readable justification"""
if similar_incidents:
similar_count = len(similar_incidents)
avg_success = sum(i.get("success_rate", 0.0) for i in similar_incidents) / similar_count
return (
f"Detected anomalies in {component} with {confidence:.0%} confidence. "
f"Found {similar_count} similar historical incidents with {avg_success:.0%} average success rate. "
f"Recommended {action} based on pattern matching and historical effectiveness."
)
else:
critical_metrics = []
for metric, value in metrics.items():
if isinstance(value, (int, float)) and value > 80: # Threshold
critical_metrics.append(f"{metric}: {value}")
return (
f"Detected anomalies in {component} with {confidence:.0%} confidence. "
f"Critical metrics: {', '.join(critical_metrics[:3])}. "
f"Recommended {action} based on anomaly characteristics and component type."
)
def _create_fallback_intent(self, action: str, component: str,
metrics: Dict[str, Any], confidence: float) -> Dict[str, Any]:
"""Create fallback healing intent when decision agent fails"""
return {
"action": action,
"component": component,
"parameters": {"fallback": True},
"confidence": confidence * 0.8, # Reduced confidence for fallback
"justification": f"Fallback recommendation for {component} anomalies",
"requires_enterprise": True,
"oss_advisory": True,
"similar_incidents_count": 0,
"rag_similarity_score": None,
"timestamp": time.time(),
"arf_version": "3.3.7"
}
def _compile_results(self, scenario_name: str, detection_result: Dict[str, Any],
similar_incidents: List[Dict[str, Any]], healing_intent: Dict[str, Any],
analysis_time_ms: float, component: str, metrics: Dict[str, Any]) -> Dict[str, Any]:
"""Compile all analysis results into final format"""
return {
"status": "success",
"scenario": scenario_name,
"analysis": {
"detection": detection_result,
"recall": similar_incidents,
"decision": healing_intent
},
"capabilities": {
"execution_allowed": False,
"mcp_modes": ["advisory"],
"oss_boundary": "advisory_only",
"requires_enterprise": True,
},
"agents_used": ["Detection", "Recall", "Decision"],
"analysis_time_ms": analysis_time_ms,
"arf_version": "3.3.7",
"oss_edition": True,
"demo_display": {
"real_arf_version": "3.3.7",
"true_oss_used": True,
"enterprise_simulated": False,
"agent_details": {
"detection_confidence": detection_result["confidence"],
"similar_incidents_count": len(similar_incidents),
"decision_confidence": healing_intent["confidence"],
"healing_action": healing_intent["action"],
}
}
}
def _create_no_anomaly_result(self, scenario_name: str, start_time: float) -> Dict[str, Any]:
"""Create result when no anomalies are detected"""
analysis_time_ms = (time.time() - start_time) * 1000
return {
"status": "success",
"scenario": scenario_name,
"result": "no_anomalies_detected",
"analysis_time_ms": analysis_time_ms,
"arf_version": "3.3.7",
"oss_edition": True,
"demo_display": {
"real_arf_version": "3.3.7",
"true_oss_used": True,
"no_anomalies": True
}
}
def _create_error_result(self, scenario_name: str, error: str,
start_time: float) -> Dict[str, Any]:
"""Create error result"""
analysis_time_ms = (time.time() - start_time) * 1000
return {
"status": "error",
"error": error,
"scenario": scenario_name,
"analysis_time_ms": analysis_time_ms,
"arf_version": "3.3.7",
"oss_edition": True,
"demo_display": {
"real_arf_version": "3.3.7",
"true_oss_used": True,
"error": error[:100]
}
}
def get_agent_stats(self) -> Dict[str, Any]:
"""Get statistics from all agents"""
return {
**self.agent_stats,
"oss_available": self.oss_available,
"arf_version": "3.3.7",
"avg_analysis_time_ms": (
self.agent_stats["total_time_ms"] / self.agent_stats["total_analyses"]
if self.agent_stats["total_analyses"] > 0 else 0
)
}
# ============================================================================
# FACTORY FUNCTION
# ============================================================================
async def get_true_arf_oss(config: Optional[Dict[str, Any]] = None) -> TrueARFOSS:
"""
Factory function for TrueARFOSS
This is the function that TrueARF337Orchestrator expects to call.
Args:
config: Optional configuration
Returns:
TrueARFOSS instance
"""
return TrueARFOSS(config)
# ============================================================================
# SIMPLE MOCK FOR BACKWARDS COMPATIBILITY
# ============================================================================
async def get_mock_true_arf_oss(config: Optional[Dict[str, Any]] = None) -> TrueARFOSS:
"""
Mock version for when dependencies are missing
"""
logger.warning("Using mock TrueARFOSS - real implementation not available")
class MockTrueARFOSS:
def __init__(self, config):
self.config = config or {}
self.oss_available = False
async def analyze_scenario(self, scenario_name, scenario_data):
return {
"status": "mock",
"scenario": scenario_name,
"message": "Mock analysis - install true ARF OSS v3.3.7 for real analysis",
"demo_display": {
"real_arf_version": "mock",
"true_oss_used": False,
"enterprise_simulated": False,
}
}
return MockTrueARFOSS(config)
# ============================================================================
# MAIN ENTRY POINT
# ============================================================================
if __name__ == "__main__":
# Test the implementation
import asyncio
async def test():
# Create test scenario
scenario = {
"component": "redis_cache",
"metrics": {
"cache_hit_rate": 18.5,
"database_load": 92,
"response_time_ms": 1850,
},
"business_impact": {
"revenue_loss_per_hour": 8500
}
}
arf = await get_true_arf_oss()
result = await arf.analyze_scenario("Test Cache Miss Storm", scenario)
print("Test Result:", json.dumps(result, indent=2, default=str))
asyncio.run(test())