fix: remove hardcoded service URLs, require ENV config
Build and Push / build (release) Successful in 29s
Build and Push / build (release) Successful in 29s
- Remove hardcoded default URLs (portainer, npm, ollama, etc.) - All external service URLs now required via ENV vars - Remove obsolete ai_client and ai_controller (core-ai proxy) - Clean up health_controller obsolete core-ai references - App fails fast at startup if required config missing 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Claude Opus 4.5
@@ -1,219 +0,0 @@
|
||||
"""
|
||||
AI Metrics Proxy Controller
|
||||
|
||||
Provides proxy endpoints to Core-AI service metrics.
|
||||
Allows external access to AI performance stats via core-api.
|
||||
"""
|
||||
from fastapi import APIRouter, HTTPException
|
||||
from typing import Dict, List, Any
|
||||
from src.clients.ai_client import get_ai_client
|
||||
from src.logging_config import get_logger
|
||||
|
||||
logger = get_logger(__name__)
|
||||
|
||||
# Create router
|
||||
router = APIRouter(
|
||||
prefix="/ai",
|
||||
tags=["AI Metrics"]
|
||||
)
|
||||
|
||||
|
||||
@router.get(
|
||||
"/health",
|
||||
summary="Check Core-AI service health",
|
||||
description="Verify that the Core-AI service is accessible and responding"
|
||||
)
|
||||
async def ai_health_check():
|
||||
"""
|
||||
Check if Core-AI service is healthy
|
||||
|
||||
Returns:
|
||||
Health status and availability
|
||||
"""
|
||||
try:
|
||||
ai_client = get_ai_client()
|
||||
is_healthy = await ai_client.health_check()
|
||||
|
||||
return {
|
||||
"service": "core-ai",
|
||||
"status": "healthy" if is_healthy else "unhealthy",
|
||||
"accessible": is_healthy
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"AI health check failed: {e}")
|
||||
return {
|
||||
"service": "core-ai",
|
||||
"status": "error",
|
||||
"accessible": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
|
||||
@router.get(
|
||||
"/metrics",
|
||||
response_model=Dict[str, Any],
|
||||
summary="Get comprehensive AI performance metrics",
|
||||
description="Returns detailed metrics including agent performance, tool execution stats, memory system metrics, and user activity"
|
||||
)
|
||||
async def get_ai_metrics():
|
||||
"""
|
||||
Proxy endpoint for Core-AI metrics
|
||||
|
||||
Returns comprehensive AI performance data:
|
||||
- Agent request statistics (total, by type, response times)
|
||||
- Response time percentiles (p50, p95, p99)
|
||||
- Tool execution metrics (calls, success rates, durations)
|
||||
- Memory system statistics (cache hits, consolidations)
|
||||
- User activity tracking
|
||||
- Concurrency metrics
|
||||
|
||||
Returns:
|
||||
Dict with all collected metrics
|
||||
|
||||
Raises:
|
||||
HTTPException: If Core-AI is unreachable or returns error
|
||||
"""
|
||||
try:
|
||||
ai_client = get_ai_client()
|
||||
metrics = await ai_client.get_metrics()
|
||||
return metrics
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to fetch AI metrics: {e}")
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail=f"Core-AI service unavailable: {str(e)}"
|
||||
)
|
||||
|
||||
|
||||
@router.get(
|
||||
"/metrics/errors",
|
||||
response_model=Dict[str, Any],
|
||||
summary="Get recent request errors",
|
||||
description="Returns recent AI agent request errors with timestamps and details"
|
||||
)
|
||||
async def get_ai_errors(limit: int = 20):
|
||||
"""
|
||||
Get recent AI request errors
|
||||
|
||||
Args:
|
||||
limit: Maximum number of errors to return (default: 20)
|
||||
|
||||
Returns:
|
||||
Dict with error list and total count
|
||||
|
||||
Example response:
|
||||
{
|
||||
"errors": [
|
||||
{
|
||||
"timestamp": "2025-12-03T19:45:12Z",
|
||||
"agent_type": "pydantic",
|
||||
"error": "Connection timeout",
|
||||
"duration_ms": 5000
|
||||
}
|
||||
],
|
||||
"total": 1
|
||||
}
|
||||
"""
|
||||
try:
|
||||
ai_client = get_ai_client()
|
||||
errors = await ai_client.get_recent_errors(limit=limit)
|
||||
|
||||
return {
|
||||
"errors": errors,
|
||||
"total": len(errors)
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to fetch AI errors: {e}")
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail=f"Core-AI service unavailable: {str(e)}"
|
||||
)
|
||||
|
||||
|
||||
@router.get(
|
||||
"/metrics/tool-failures",
|
||||
response_model=Dict[str, Any],
|
||||
summary="Get recent tool execution failures",
|
||||
description="Returns recent tool execution failures with error details"
|
||||
)
|
||||
async def get_ai_tool_failures(limit: int = 20):
|
||||
"""
|
||||
Get recent tool execution failures
|
||||
|
||||
Args:
|
||||
limit: Maximum number of failures to return (default: 20)
|
||||
|
||||
Returns:
|
||||
Dict with failure list and total count
|
||||
|
||||
Example response:
|
||||
{
|
||||
"failures": [
|
||||
{
|
||||
"timestamp": "2025-12-03T19:50:30Z",
|
||||
"tool_name": "list_containers",
|
||||
"error": "Connection refused",
|
||||
"duration_ms": 150
|
||||
}
|
||||
],
|
||||
"total": 1
|
||||
}
|
||||
"""
|
||||
try:
|
||||
ai_client = get_ai_client()
|
||||
failures = await ai_client.get_tool_failures(limit=limit)
|
||||
|
||||
return {
|
||||
"failures": failures,
|
||||
"total": len(failures)
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to fetch tool failures: {e}")
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail=f"Core-AI service unavailable: {str(e)}"
|
||||
)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/metrics/reset",
|
||||
summary="Reset all AI metrics (admin)",
|
||||
description="Clear all collected metrics. This is an administrative operation that resets all counters and history."
|
||||
)
|
||||
async def reset_ai_metrics():
|
||||
"""
|
||||
Reset all AI metrics (admin operation)
|
||||
|
||||
Clears all collected metrics including:
|
||||
- Request history
|
||||
- Tool execution stats
|
||||
- Memory system metrics
|
||||
- Error logs
|
||||
|
||||
Returns:
|
||||
Success confirmation
|
||||
|
||||
Note:
|
||||
This is an administrative operation that should be used carefully.
|
||||
All historical data will be lost.
|
||||
"""
|
||||
try:
|
||||
ai_client = get_ai_client()
|
||||
await ai_client.reset_metrics()
|
||||
|
||||
logger.info("AI metrics reset successfully")
|
||||
return {
|
||||
"success": True,
|
||||
"message": "AI metrics reset successfully"
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to reset AI metrics: {e}")
|
||||
raise HTTPException(
|
||||
status_code=503,
|
||||
detail=f"Core-AI service unavailable: {str(e)}"
|
||||
)
|
||||
@@ -11,9 +11,6 @@ from src.config import get_settings
|
||||
from src.logging_config import get_logger
|
||||
from src.models.ollama_client import get_ollama_client
|
||||
|
||||
# Note: Agent functionality moved to separate core-ai service (Dec 2025)
|
||||
# This service (core-api) only provides infrastructure management and tools
|
||||
AGENT_AVAILABLE = False
|
||||
|
||||
|
||||
logger = get_logger(__name__)
|
||||
@@ -134,11 +131,6 @@ class HealthController(BaseController):
|
||||
ollama_error = str(e)
|
||||
logger.warning(f"Ollama health check failed: {ollama_error}")
|
||||
|
||||
# Note: Agent functionality moved to separate core-ai service
|
||||
# This service only needs Ollama for embeddings (infrastructure tools)
|
||||
# Agent health is checked separately in core-ai service
|
||||
|
||||
# Determine overall status (only Ollama required for core-api)
|
||||
is_healthy = ollama_healthy
|
||||
|
||||
elapsed_ms = int((time.time() - start_time) * 1000)
|
||||
@@ -157,8 +149,7 @@ class HealthController(BaseController):
|
||||
"available": False
|
||||
},
|
||||
"error": ollama_error
|
||||
},
|
||||
"note": "AI agent functionality available in separate core-ai service (port 8086)"
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@@ -204,20 +195,7 @@ class HealthController(BaseController):
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
# 2. Agent Stack - Moved to separate core-ai service
|
||||
diagnostics["components"]["agent"] = {
|
||||
"status": "N/A",
|
||||
"note": "AI agent functionality moved to separate core-ai service (port 8086)",
|
||||
"check_url": "http://core-ai:8086/health"
|
||||
}
|
||||
|
||||
# 3. Memory System (Qdrant) - Moved to core-ai service
|
||||
diagnostics["components"]["qdrant"] = {
|
||||
"status": "N/A",
|
||||
"note": "Memory system managed by core-ai service (port 8086)"
|
||||
}
|
||||
|
||||
# 4. Configuration
|
||||
# 2. Configuration
|
||||
diagnostics["configuration"] = {
|
||||
"agent_fallback_enabled": settings.agent_fallback_enabled,
|
||||
"memory_tier1_max_turns": settings.memory_tier1_max_turns,
|
||||
|
||||
Reference in New Issue
Block a user