Files
taiji-AI-PAD/config/prometheus.yml
T
xiaohei 9b5f60fa91 config: 配置Prometheus和Grafana监控
配置内容:
1. 更新Prometheus配置
   - 添加服务标签
   - 优化抓取间隔和超时
   - 配置NATS监控端点
   - 添加外部标签

2. 创建Grafana数据源配置
   - 配置Prometheus数据源
   - 设置为默认数据源
   - 配置查询超时和间隔

3. 创建Grafana仪表板配置
   - 配置自动发现仪表板
   - 支持UI更新

4. 创建监控仪表板
   - taiji-monitoring.json
   - 包含HTTP请求、Agent、工具调用、MCP请求等监控面板
   - 10个监控面板,覆盖系统关键指标

监控指标:
- HTTP请求总数和耗时
- 活跃Agent数
- Agent注册统计
- 工具调用统计
- MCP请求统计
- WebSocket连接数
- 系统连接状态(Redis、NATS、Database)
- 函数注册表大小
- 函数工具调用统计
2025-12-23 07:42:56 +00:00

79 lines
2.0 KiB
YAML

global:
scrape_interval: 15s
evaluation_interval: 15s
external_labels:
cluster: 'taiji-ai-pad'
environment: 'production'
# 告警规则配置(可选,未来可添加)
# rule_files:
# - "alert_rules.yml"
# Alertmanager配置(可选,未来可添加)
# alerting:
# alertmanagers:
# - static_configs:
# - targets: ['alertmanager:9093']
scrape_configs:
# MCP Server
- job_name: 'mcp-server'
static_configs:
- targets: ['mcp-server:8000']
labels:
service: 'mcp-server'
component: 'api'
metrics_path: '/metrics'
scrape_interval: 10s
scrape_timeout: 5s
# Data Ingestion Service
- job_name: 'data-ingestion'
static_configs:
- targets: ['data-ingestion:8000']
labels:
service: 'data-ingestion'
component: 'api'
metrics_path: '/metrics'
scrape_interval: 10s
scrape_timeout: 5s
# LiteLLM Gateway (如果有metrics端点)
- job_name: 'litellm-gateway'
static_configs:
- targets: ['litellm-gateway:4000']
labels:
service: 'litellm-gateway'
component: 'gateway'
metrics_path: '/metrics'
scrape_interval: 10s
scrape_timeout: 5s
# 如果端点不存在,Prometheus会记录错误但不会影响其他服务
# PostgreSQL (需要postgres_exporter,当前未部署)
# - job_name: 'postgres'
# static_configs:
# - targets: ['postgres-exporter:9187']
# labels:
# service: 'postgres'
# component: 'database'
# Redis (需要redis_exporter,当前未部署)
# - job_name: 'redis'
# static_configs:
# - targets: ['redis-exporter:9121']
# labels:
# service: 'redis'
# component: 'cache'
# NATS (NATS自带监控端点)
- job_name: 'nats'
static_configs:
- targets: ['nats:8222']
labels:
service: 'nats'
component: 'message-queue'
metrics_path: '/varz'
scrape_interval: 30s
scrape_timeout: 10s