Imporve monitoring 5
This commit is contained in:
@@ -1,160 +1,61 @@
|
||||
"""Main FastAPI application for AI Insights Service."""
|
||||
|
||||
from fastapi import FastAPI, Response
|
||||
from fastapi.middleware.cors import CORSMiddleware
|
||||
from contextlib import asynccontextmanager
|
||||
import structlog
|
||||
import os
|
||||
|
||||
from app.core.config import settings
|
||||
from app.core.database import init_db, close_db
|
||||
from app.api import insights
|
||||
from shared.monitoring.logging import setup_logging
|
||||
from shared.monitoring.metrics import MetricsCollector, add_metrics_middleware
|
||||
from shared.monitoring.system_metrics import SystemMetricsCollector
|
||||
from shared.service_base import StandardFastAPIService
|
||||
|
||||
# OpenTelemetry imports
|
||||
from opentelemetry import trace
|
||||
from opentelemetry.sdk.trace import TracerProvider
|
||||
from opentelemetry.sdk.trace.export import BatchSpanProcessor
|
||||
from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter
|
||||
from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor
|
||||
from opentelemetry.instrumentation.httpx import HTTPXClientInstrumentor
|
||||
from opentelemetry.instrumentation.redis import RedisInstrumentor
|
||||
from opentelemetry.instrumentation.sqlalchemy import SQLAlchemyInstrumentor
|
||||
from opentelemetry.sdk.resources import Resource
|
||||
|
||||
# Configure OpenTelemetry tracing
|
||||
def setup_tracing(service_name: str = "ai-insights"):
|
||||
"""Initialize OpenTelemetry tracing with OTLP exporter for Jaeger"""
|
||||
resource = Resource.create({"service.name": service_name})
|
||||
|
||||
otlp_exporter = OTLPSpanExporter(
|
||||
endpoint=os.getenv("OTEL_EXPORTER_OTLP_ENDPOINT", "http://signoz-otel-collector.bakery-ia.svc.cluster.local:4317"),
|
||||
insecure=True
|
||||
)
|
||||
|
||||
provider = TracerProvider(resource=resource)
|
||||
processor = BatchSpanProcessor(otlp_exporter)
|
||||
provider.add_span_processor(processor)
|
||||
trace.set_tracer_provider(provider)
|
||||
|
||||
return provider
|
||||
|
||||
# Initialize tracing
|
||||
tracer_provider = setup_tracing("ai-insights")
|
||||
|
||||
# Setup logging
|
||||
setup_logging("ai-insights", getattr(settings, 'LOG_LEVEL', 'INFO'))
|
||||
# Initialize logger
|
||||
logger = structlog.get_logger()
|
||||
|
||||
# Setup OpenTelemetry logging export if enabled
|
||||
logger.info(f"OTEL_LOGS_EXPORTER env var: {os.getenv('OTEL_LOGS_EXPORTER', 'not set')}")
|
||||
if os.getenv("OTEL_LOGS_EXPORTER", "").lower() == "otlp":
|
||||
try:
|
||||
logger.info("Attempting to setup OpenTelemetry logging")
|
||||
from shared.monitoring.logs_exporter import setup_otel_logging
|
||||
result = setup_otel_logging("ai-insights", settings.SERVICE_VERSION)
|
||||
if result:
|
||||
logger.info("OpenTelemetry logs export enabled for ai-insights")
|
||||
else:
|
||||
logger.warning("OpenTelemetry logs export setup returned None")
|
||||
except Exception as e:
|
||||
logger.error(f"Failed to setup OpenTelemetry logs export: {e}", exc_info=True)
|
||||
else:
|
||||
logger.info("OpenTelemetry logs export disabled - OTEL_LOGS_EXPORTER not set to otlp")
|
||||
|
||||
class AIInsightsService(StandardFastAPIService):
|
||||
"""AI Insights Service with standardized monitoring setup"""
|
||||
|
||||
async def on_startup(self, app):
|
||||
"""Custom startup logic for AI Insights"""
|
||||
# Initialize database
|
||||
await init_db()
|
||||
logger.info("Database initialized")
|
||||
|
||||
await super().on_startup(app)
|
||||
|
||||
async def on_shutdown(self, app):
|
||||
"""Custom shutdown logic for AI Insights"""
|
||||
await super().on_shutdown(app)
|
||||
|
||||
# Close database
|
||||
await close_db()
|
||||
logger.info("Database connections closed")
|
||||
|
||||
|
||||
@asynccontextmanager
|
||||
async def lifespan(app: FastAPI):
|
||||
"""Lifespan event handler for startup and shutdown."""
|
||||
# Startup
|
||||
logger.info("Starting AI Insights Service", service=settings.SERVICE_NAME, version=settings.SERVICE_VERSION)
|
||||
await init_db()
|
||||
logger.info("Database initialized")
|
||||
|
||||
# Initialize system metrics collection
|
||||
system_metrics = SystemMetricsCollector("ai-insights")
|
||||
logger.info("System metrics collection started")
|
||||
|
||||
# Note: Metrics are exported via OpenTelemetry OTLP to SigNoz - no metrics server needed
|
||||
logger.info("Metrics export configured via OpenTelemetry OTLP")
|
||||
|
||||
yield
|
||||
|
||||
# Shutdown
|
||||
logger.info("Shutting down AI Insights Service")
|
||||
await close_db()
|
||||
logger.info("Database connections closed")
|
||||
|
||||
|
||||
# Create FastAPI app
|
||||
app = FastAPI(
|
||||
title="AI Insights Service",
|
||||
# Create service instance
|
||||
service = AIInsightsService(
|
||||
service_name="ai-insights",
|
||||
app_name="AI Insights Service",
|
||||
description="Intelligent insights and recommendations for bakery operations",
|
||||
version=settings.SERVICE_VERSION,
|
||||
lifespan=lifespan
|
||||
log_level=getattr(settings, 'LOG_LEVEL', 'INFO'),
|
||||
cors_origins=getattr(settings, 'ALLOWED_ORIGINS', ["*"]),
|
||||
api_prefix=settings.API_V1_PREFIX,
|
||||
enable_metrics=True,
|
||||
enable_health_checks=True,
|
||||
enable_tracing=True,
|
||||
enable_cors=True
|
||||
)
|
||||
|
||||
# Instrument FastAPI with OpenTelemetry
|
||||
FastAPIInstrumentor.instrument_app(app)
|
||||
# Create FastAPI app
|
||||
app = service.create_app()
|
||||
|
||||
# Instrument httpx for outgoing requests
|
||||
HTTPXClientInstrumentor().instrument()
|
||||
|
||||
# Instrument Redis
|
||||
RedisInstrumentor().instrument()
|
||||
|
||||
# Instrument SQLAlchemy
|
||||
SQLAlchemyInstrumentor().instrument()
|
||||
|
||||
# Initialize metrics collector
|
||||
metrics_collector = MetricsCollector("ai-insights")
|
||||
|
||||
# Add metrics middleware to track HTTP requests
|
||||
add_metrics_middleware(app, metrics_collector)
|
||||
|
||||
# CORS middleware
|
||||
app.add_middleware(
|
||||
CORSMiddleware,
|
||||
allow_origins=settings.ALLOWED_ORIGINS,
|
||||
allow_credentials=True,
|
||||
allow_methods=["*"],
|
||||
allow_headers=["*"],
|
||||
)
|
||||
|
||||
# Include routers
|
||||
app.include_router(
|
||||
# Add service-specific routers
|
||||
service.add_router(
|
||||
insights.router,
|
||||
prefix=settings.API_V1_PREFIX,
|
||||
tags=["insights"]
|
||||
)
|
||||
|
||||
|
||||
@app.get("/")
|
||||
async def root():
|
||||
"""Root endpoint."""
|
||||
return {
|
||||
"service": settings.SERVICE_NAME,
|
||||
"version": settings.SERVICE_VERSION,
|
||||
"status": "running"
|
||||
}
|
||||
|
||||
|
||||
@app.get("/health")
|
||||
async def health_check():
|
||||
"""Health check endpoint."""
|
||||
return {
|
||||
"status": "healthy",
|
||||
"service": settings.SERVICE_NAME,
|
||||
"version": settings.SERVICE_VERSION
|
||||
}
|
||||
|
||||
|
||||
# Note: Metrics are exported via OpenTelemetry OTLP to SigNoz
|
||||
# The /metrics endpoint is not needed as metrics are pushed automatically
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
import uvicorn
|
||||
|
||||
|
||||
Reference in New Issue
Block a user