k8s 收集event事件至Loki
🧭 前言
在生产环境中,Kubernetes 事件(Event)默认存储时间短、查询不便。通过将事件导出并存储在 Grafana Loki 中,我们可以长期保留、可视化和告警 Kubernetes 的运行状态变化,这对故障排查、调度优化、容量规划非常重要。本文档基于 loki-distributed 部署方式,详细介绍如何使用 kubernetes-event-exporter 获取事件,存入 Loki,并在 Grafana 中导入官方 Dashboard(ID:17882),快速形成监控产物。
✅ 前置条件
确保你已有以下环境:
-
Kubernetes 集群正常运行
-
loki-distributed 已部署完成,且可访问 Loki 对外网关
-
Grafana 已部署,并配置了 Loki 作为 Data Source
📁 部署目录及清单
.
├── 00-roles.yaml # RBAC 配置
├── 01-config.yaml # event-exporter 配置
└── 02-deployment.yaml # Deployment 配置
1️⃣ RBAC 权限配置(00-roles.yaml)
apiVersion: v1
kind: Namespace
metadata:name: monitoring
---
apiVersion: v1
kind: ServiceAccount
metadata:name: event-exporternamespace: monitoring
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:name: event-exporter
rules:- apiGroups: ["*"]resources: ["*"]verbs: ["get","watch","list"]- apiGroups: ["coordination.k8s.io"]resources: ["leases"]verbs: ["*"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:name: event-exporter
roleRef:apiGroup: rbac.authorization.k8s.iokind: ClusterRolename: event-exporter
subjects:- kind: ServiceAccountname: event-exporternamespace: monitoring
2️⃣ event-exporter 配置(01-config.yaml)
apiVersion: v1
kind: ConfigMap
metadata:name: event-exporter-confignamespace: monitoring
data:config.yaml: |logLevel: debuglogFormat: jsonmetricsNamePrefix: event_exporter_sinks:loki:type: lokiconfig:url: "http://loki-distributed-gateway.loki.svc.cluster.local:80/loki/api/v1/push"route:routes:- match:- receiver: "loki"
3️⃣ 部署 event-exporter(02-deployment.yaml)
apiVersion: apps/v1
kind: Deployment
metadata:name: event-exporternamespace: monitoring
spec:replicas: 1selector:matchLabels:app: event-exporterversion: v1template:metadata:labels:app: event-exporterversion: v1annotations:prometheus.io/scrape: 'true'prometheus.io/port: '2112'prometheus.io/path: '/metrics'spec:serviceAccountName: event-exportersecurityContext:runAsNonRoot: trueseccompProfile:type: RuntimeDefaultcontainers:- name: event-exporterimage: ghcr.io/resmoio/kubernetes-event-exporter:latestargs:- -conf=/config/config.yamlvolumeMounts:- name: config-volumemountPath: /configsecurityContext:allowPrivilegeEscalation: falsecapabilities:drop: [ALL]volumes:- name: config-volumeconfigMap:name: event-exporter-config
4️⃣ 应用部署
kubectl apply -f 00-roles.yaml
kubectl apply -f 01-config.yaml
kubectl apply -f 02-deployment.yaml
确认 Pod 启动无误:
kubectl -n monitoring get pods -l app=event-exporter
kubectl -n monitoring logs deploy/event-exporterkubectl get pod -n monitoring
NAME READY STATUS RESTARTS AGE
event-exporter-6768f485cb-knx9v 1/1 Running 0 3h32m
5️⃣ Grafana 配置与 Dashboard 导入
-
添加 Loki 数据源
-
Grafana → Settings → Data Sources → Add → Loki
-
URL 填写:http://<loki-gateway-service>:80,测试连接
-
导入 Dashboard
-
Grafana → “+” → Import → 输入 Dashboard ID:17882
-
选择刚配置的 Loki Data Source,完成导入
6️⃣ 验证与查询示例
-
直接在 Grafana → Explore 中查询:
{app="event-exporter"} | json
-
示例查询错误事件数并填补零值:
sum(count_over_time({app="event-exporter"} | json | reason="Failed"[$__interval])) or 0
📚 参考资料
-
官方 GitHub:resmoio/kubernetes-event-exporter
-
Dashboard ID:17882 Kubernetes Event Exporter
-
Loki 文档:loki-distributed 模式说明(Simple Scalable)
✅ 小结
通过本文档,你将实现一个完整方案:监听 Kubernetes Event → event-exporter 将其推入 loki-distributed → Grafana 导入 Dashboard 可视化事件信息。可进一步优化包括消息过滤、报警告警及多集群统一事件收集。如需示例 Helm values 或 Kustomize 器,随时告诉我!