东莞网站设计百年营销推广策划及渠道

东莞网站设计百年,营销推广策划及渠道,殷氏科技网站建设工作室,中企动力做的网站山西太原一、功能描述 VMAlert负责执行告警规则、生成告警事件#xff0c;并将告警推送给 Alertmanager 或 Webhook#xff0c;同时还能计算 Recording Rule 并写入存储。 在 集群模式下#xff0c; VictoriaMetrics 具有独立的写入和读取路径组件 。vminsert写入组件用于执行规则…一、功能描述VMAlert负责执行告警规则、生成告警事件并将告警推送给 Alertmanager 或 Webhook同时还能计算 Recording Rule 并写入存储。在 集群模式下 VictoriaMetrics 具有独立的写入和读取路径组件 。vminsert写入组件用于执行规则表达式读取vmselect组件用于持久化记录规则结果和警报状态。集群模式可以包含多个写入和读取组件。vmselectvminsertvminsertvmselectvmalert配置标志1. 执行 Alerting Rules告警规则从 Prometheus / VictoriaMetrics 数据源查询数据周期性评估告警alert: HighErrorRate expr: rate(http_errors_total[5m]) 10 for: 1m当条件满足 → 触发告警当条件恢复 → 发送恢复2. 发送告警通知Alert notifications支持两种方式发送告警发送给 Alertmanager推荐走 Alertmanager再由 Alertmanager 路由到不同通知渠道Slack、钉钉、飞书、Webhook 等直接发送 Webhook不经过 Alertmanager适合自定义告警系统或你这种需要直接推送到接口的情况。3. 执行 Recording Rules记录规则计算新的时序数据并写入 VictoriaMetricsremote writerecord: job:http_requests:rate5m expr: rate(http_requests_total[5m])Recording rules 的好处查询速度更快降低 VictoriaMetrics 计算压力Dashboards 用录制好的指标更顺畅4. 动态加载规则不需要重启VMAlert 支持ConfigMap 动态挂载规则随时 reload而不重启 Pod5. 提供 UI 查看规则状态当前规则执行情况告警触发状态执行错误数据查询结果二、多个alert实例需要增加下面参数用来去重VM单点模式下insert和select 增加dedup.minScrapeInterval: 10sVM-Cluster 模式下storage 需要增加 dedup.minScrapeInterval: 15sselect 需要增加 dedup.minScrapeInterval: 15s数据延迟配置extraArgs: loggerLevel: INFO rule.evalDelay: 30s #默认值需要和search.latencyOffset 设置为一致 数据有延迟的时候需要配置三、监控看板https://grafana.com/grafana/dashboards/24513-victorialogs-vlagent/四、全量配置apiVersion: operator.victoriametrics.com/v1beta1 kind: VMAlert metadata: name: custom-alert namespace: monitoring spec: replicaCount: 1 imagePullSecrets: - name: uhub-registry datasource: url: http://vm-victoria-metrics-cluster-vmselect.monitoring:8481/select/0/prometheus remoteWrite: url: http://vm-victoria-metrics-cluster-vminsert.monitoring:8480/insert/0/prometheus remoteRead: url: http://vm-victoria-metrics-cluster-vmselect.monitoring:8481/select/0/prometheus maxConnections: 4 image: repository: vmalert tag: v1.128.0 pullPolicy: Always externalLabels: cluster: xxx-prod environment: cn-beijing notifier: url: http://vmalertmanager-custom-alertmanager.monitoring.svc:9093 timeout: 10s maxConnections: 5 #alertmanager: # enable: true #alertmanagerConfigSelector: # matchLabels: # vmamcfg: custom-webhook #notifier: # alertmanager: # enable: true # selector: # matchLabels: # app.kubernetes.io/name: alertmanager resources: limits: cpu: 2 memory: 4Gi requests: cpu: 100m memory: 128Mi #ExternalURL: https://victoria-alert.xxx.cn #external: # url: https://victoria-alert.xxx.cn extraArgs: loggerLevel: INFO rule.evalDelay: 30s #默认值需要和search.latencyOffset 设置为一致 数据有延迟的时候需要配置 external.url: https://victoria-alert.xxx.cn # 可以将rules存到s3 # rule: s3://bucket/dir/alert.rules serviceMonitor: enabled: true interval: 30s tolerations: - effect: NoSchedule key: service operator: Equal value: sre-victoria-metrics affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: business operator: In values: - sre selectAllByDefault: true evaluationInterval: 30s scrapeConfigSelector: {} serviceScrapeSelector: {} ruleNamespaceSelector: {} podScrapeSelector: {} nodeScrapeSelector: {} staticScrapeSelector: {} probeSelector: {} ruleSelector: {}参考文档https://docs.victoriametrics.com/victoriametrics/vmalert/https://docs.victoriametrics.com/operator/resources/