feat: 完善服务监控与下游重投

This commit is contained in:
hectorzhao
2026-08-14 17:21:29 +08:00
parent d30d9ea4d0
commit 1ef4380422
50 changed files with 1279 additions and 82 deletions
+4 -1
View File
@@ -7,9 +7,10 @@
```bash
cd /opt/cmpp-platform
bash tools/monitoring/install-prometheus-monitoring.sh
bash tools/monitoring/install-service-exporters.sh
```
脚本安装Prometheus与Node Exporter备份已有Prometheus配置及本脚本曾写入的systemd override、校验规则、写入新override,并仅重启这两个监控服务。备份目录会在脚本结束时打印。它不会重启API、Gateway、数据库、Redis、MinIO或Nginx。9090和9100固定监听`127.0.0.1`
第一个脚本安装Prometheus与Node Exporter备份已有配置并校验规则;第二个脚本安装PostgreSQL、Redis和Nginx Exporter,开启MinIO回环原生指标。API指标仅监听`127.0.0.1:9464`Gateway指标复用回环控制端口`8090`。9090、9100、9187、9121、9113和9464均不得对公网开放。两个脚本均会打印恢复资产路径
安装后将下列配置写入`/etc/cmpp-platform/cmpp-platform.env`,再按正常发布窗口重启API
@@ -26,6 +27,8 @@ promtool check rules /etc/prometheus/cmpp-alerts.yml
curl -fsS http://127.0.0.1:9090/-/ready
curl -fsS 'http://127.0.0.1:9090/api/v1/query?query=up'
ss -lnt | grep -E ':(9090|9100)'
curl -fsS http://127.0.0.1:9464/metrics
curl -fsS http://127.0.0.1:8090/metrics
```
完整架构、PromQL口径、故障语义和验收标准见`docs/prometheus-system-monitoring-design-20260814.md`