文档首页 / HelmChart 一键部署

HelmChart 一键部署

概述

Helm 是 Kubernetes 的包管理工具，Chart 是 Helm 的打包格式。通过 HelmChart，可以一键部署完整的监控体系（Prometheus + Grafana + Alertmanager + Exporters）。

环境准备

# 安装 Helm 3


curl -fsSL https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash


验证安装


helm version


添加常用仓库


helm repo add bitnami https://charts.bitnami.com/bitnami


helm repo add prometheus-community https://prometheus-community.github.io/helm-charts


helm repo add grafana https://grafana.github.io/helm-charts


helm repo update

kube-prometheus-stack（推荐）

kube-prometheus-stack 是生产级监控方案，一键部署 Prometheus + Grafana + Alertmanager + Exporter。

# 添加仓库


helm repo add prometheus-community https://prometheus-community.github.io/helm-charts


helm repo update


安装


helm install prometheus prometheus-community/kube-prometheus-stack \


--namespace monitoring \


--create-namespace \


--set prometheus.prometheusSpec.retention=15d \


--set prometheus.prometheusSpec.replicas=2 \


--set prometheus.prometheusSpec.storageSpec.volumeClaimTemplate.spec.resources.requests.storage=50Gi \


--set alertmanager.alertmanagerSpec.replicas=2 \


--set grafana.adminPassword="YourStrongPassword" \


--set grafana.persistence.enabled=true \


--set grafana.persistence.size=10Gi


验证


kubectl get pods -n monitoring

values.yaml 核心配置

# 自定义 values-prometheus.yaml


prometheus:


prometheusSpec:


# 数据保留时间


retention: 30d


retentionSize: "20Gi"


# 高可用：2 副本


replicas: 2


# 资源限制


resources:


requests:


cpu: 500m


memory: 2Gi


limits:


cpu: 2


memory: 4Gi


# 持久化存储


storageSpec:


volumeClaimTemplate:


spec:


storageClassName: gp3


resources:


requests:


storage: 100Gi


# 告警规则启用


ruleSelector: true


ruleSelectorMixins: true


# 抓取配置


serviceMonitorSelector: {}


serviceMonitorNamespaceSelector: {}


# 额外 scrape 配置


extraScrapeConfigs: |


    

  • job_name: 'blackbox-http'
    • 



metrics_path: /probe


params:


module: [http_2xx]


static_configs:


    

  • targets:
    • 

  • https://example.com
    • 



relabel_configs:


    

  • source_labels: [__address__]
    • 



target_label: __param_target


    

  • target_label: __address__
    • 



replacement: blackbox-exporter:9115


alertmanager:


alertmanagerSpec:


replicas: 2


# 持久化


storage:


volumeClaimTemplate:


spec:


storageClassName: gp3


resources:


requests:


storage: 5Gi


# 通知配置


config:


global:


resolve_timeout: 5m


route:


group_by: ['job', 'severity']


group_wait: 10s


group_interval: 10s


receiver: 'dingtalk'


routes:


    

  • match:
    • 



severity: critical


receiver: 'dingtalk'


continue: true


receivers:


    

  • name: 'dingtalk'
    • 



webhook_configs:


    

  • url: 'http://dingtalk-hook:5000/dingtalk'
    • 



send_resolved: true


grafana:


adminPassword: "ChangeMe123!"


# 持久化


persistence:


enabled: true


size: 10Gi


storageClassName: gp3


# 额外数据源


datasources:


datasources.yaml:


apiVersion: 1


datasources:


    

  • name: Prometheus
    • 



type: prometheus


access: proxy


url: http://prometheus-operated:9090


isDefault: true


# 预设 Dashboard


dashboards:


default:


kubernetes-cluster:


gnetId: 7249


revision: 2


kubernetes-pods:


gnetId: 6336


revision: 5


# 通过 sidecar 自动导入 Dashboard ConfigMap


sidecar:


dashboards:


enabled: true


label: grafana_dashboard


searchNamespace: ALL


Prometheus Operator CRD 配置


prometheusOperator:


admissionWebhooks:


enabled: true


certManager:


enabled: false


# TLS 证书生成


admissionWebhooks.patch:


enabled: true

ServiceMonitor（自动服务发现）

在 Prometheus Operator 架构中，通过 ServiceMonitor 自动发现监控目标：

# myapp-servicemonitor.yaml


apiVersion: monitoring.coreos.com/v1


kind: ServiceMonitor


metadata:


name: myapp-monitor


namespace: monitoring


labels:


release: prometheus  # 必须匹配 prometheus-operator 的 rule selector


spec:


jobLabel: myapp


selector:


matchLabels:


app: myapp


namespaceSelector:


matchNames:


    

  • production
    • 



endpoints:


    

  • port: web
    • 



path: /metrics


interval: 15s


scrapeTimeout: 10s

PodMonitor（直接监控 Pod）：

apiVersion: monitoring.coreos.com/v1


kind: PodMonitor


metadata:


name: myapp-pod-monitor


namespace: monitoring


spec:


selector:


matchLabels:


app: myapp


podMetricsEndpoints:


    

  • port: metrics
    • 



path: /metrics


interval: 15s

PrometheusRule（自定义告警规则）

# myapp-alertrules.yaml


apiVersion: monitoring.coreos.com/v1


kind: PrometheusRule


metadata:


name: myapp-alerts


namespace: production


labels:


release: prometheus


spec:


groups:


    

  • name: myapp.rules
    • 



interval: 30s


rules:


    

  • alert: MyAppHighErrorRate
    • 



expr: |


sum(rate(http_requests_total{status=~"5.."}[5m])) /


sum(rate(http_requests_total[5m])) > 0.05


for: 5m


labels:


severity: critical


annotations:


summary: "服务 {{ $labels.instance }} 5xx 错误率超过 5%"


description: "当前错误率：{{ $value | printf \"%.2f\" }}%"

升级与回滚

# 升级到新版本


helm upgrade prometheus prometheus-community/kube-prometheus-stack \


--namespace monitoring \


-f values-prometheus.yaml


查看发布历史


helm history prometheus -n monitoring


回滚到上一版本


helm rollback prometheus -n monitoring


回滚到指定版本


helm rollback prometheus 3 -n monitoring


完全卸载


helm uninstall prometheus -n monitoring


kubectl delete crd prometheuses.monitoring.coreos.com


kubectl delete crd prometheusrules.monitoring.coreos.com

持久化配置

# 使用 AWS EBS gp3 存储


prometheus:


prometheusSpec:


storageSpec:


volumeClaimTemplate:


spec:


storageClassName: gp3


resources:


requests:


storage: 100Gi


alertmanager:


alertmanagerSpec:


storage:


volumeClaimTemplate:


spec:


storageClassName: gp3


resources:


requests:


storage: 5Gi


grafana:


persistence:


enabled: true


storageClassName: gp3


size: 10Gi

高可用部署

# values-ha.yaml


prometheus:


prometheusSpec:


replicas: 2


replicaExternalLabelName: "__replica__"


externalLabels:


cluster: prod


# 启用 Thanos Sidecar 实现长期存储


thanos:


version: v0.31.0


objectStorage:


configFile: /etc/thanos/object-storage-config.yaml


alertmanager:


alertmanagerSpec:


replicas: 2


# 使用 gossip 协议同步状态


gossip:


enabled: true


config:


global:


# 所有集群共享同一个 alertmanager 集群


cluster: "alertmanager-monitoring-0.alertmanager-operated:9093,alertmanager-monitoring-1.alertmanager-operated:9093"

常用 Helm 命令速查

# 查看集群已安装的 release


helm list -A


查看 release 配置


helm get values prometheus -n monitoring


下载 Chart 到本地


helm pull prometheus-community/kube-prometheus-stack --untar


模板渲染（不实际安装）


helm template prometheus ./prometheus-chart -f values.yaml


调试模板


helm install --dry-run --debug prometheus ./prometheus-chart -f values.yaml


导出默认 values


helm show values prometheus-community/kube-prometheus-stack > defaults.yaml

生产环境最佳实践

  • 命名空间隔离： 所有监控组件部署在 monitoring 独立命名空间
  • 资源配额： 设置 LimitRange 防止单组件耗尽集群资源
  • 存储类选择： 生产使用 EBS/GCE PD 的 gp3/pd-standard 存储类
  • 高可用： Prometheus 2 副本 + Alertmanager 2 副本
  • 告警静默： 通过 Alertmanager Web UI 配置静默规则
  • 备份： 定期备份 Prometheus TSDB 和 Alertmanager 状态
  • 升级顺序： 先升级 CRD，再升级 Operator，最后升级 Helm Release