Apache Airflow

本文档介绍如何配置 Google Kubernetes Engine 部署,以便使用 Google Cloud Managed Service for Prometheus 从 StatsD 收集指标。本页面介绍如何完成以下任务:

  • 设置 StatsD 以报告指标。
  • 为 Managed Service for Prometheus 配置 PodMonitoring 资源以收集导出的指标。
  • 在 Cloud Monitoring 中访问信息中心以查看指标。
  • 配置提醒规则以监控指标。

以下说明仅在您将代管式收集功能与 Managed Service for Prometheus 搭配使用时适用。 如果您使用的是自行部署的收集功能,请参阅 Airflow 文档了解安装信息。

这些说明仅作为示例提供,应该适用于大多数 Kubernetes 环境。如果由于安全或组织政策的限制,您在安装应用或导出器时遇到问题,我们建议您查看开源文档以获取支持。

如需了解 Airflow,请参阅 Airflow

前提条件

如需使用 Managed Service for Prometheus 和代管式收集功能从 StatsD 收集指标,您的部署必须满足以下要求:

  • 您的集群必须运行 Google Kubernetes Engine 1.21.4-gke.300 或更高版本。
  • 您必须运行 Managed Service for Prometheus,并启用代管式收集功能。如需了解详情,请参阅代管式收集功能使用入门

官方 Airflow Helm 图表包含一个 StatsD 部署,用于自动公开 Prometheus 格式指标。

要验证 StatsD 是否在预期的端点上发出指标,请执行以下操作:

  1. 使用以下命令设置端口转发:
    kubectl -n NAMESPACE_NAME port-forward deploy/AIRFLOW_RELEASE_NAME-statsd 9102
    
  2. 使用浏览器或另一个终端会话中的 curl 实用程序访问端点 localhost:9102/metrics

定义 PodMonitoring 资源

对于目标发现,Managed Service for Prometheus Operator 需要与同一命名空间中的 Airflow 导出器对应的 PodMonitoring 资源。

您可以使用以下 PodMonitoring 配置:

# Copyright 2022 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: airflow
  labels:
    app.kubernetes.io/name: airflow
    app.kubernetes.io/part-of: google-cloud-managed-prometheus
spec:
  endpoints:
  - port: 9102
    scheme: http
    interval: 30s
    path: /metrics
  selector:
    matchLabels:
      tier: airflow
      component: statsd
      release: airflow
确保 portmatchLabels 字段的值与要监控的 StatsD pod 的值相匹配。使用 Helm 部署 Airflow 时,系统会默认设置此处显示的标签和值。

如需从本地文件应用配置更改,请运行以下命令:

kubectl apply -n NAMESPACE_NAME -f FILE_NAME

您还可以使用 Terraform 管理您的配置。

定义规则和提醒

您可以使用以下 Rules 配置来定义有关 Airflow 指标提醒:

# Copyright 2022 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

apiVersion: monitoring.googleapis.com/v1
kind: Rules
metadata:
  name: airflow-rules
  labels:
    app.kubernetes.io/component: rules
    app.kubernetes.io/name: airflow-rules
    app.kubernetes.io/part-of: google-cloud-managed-prometheus
spec:
  groups:
  - name: airflow
    interval: 30s
    rules:
    - alert: AirflowDAGImportErrors
      annotations:
        description: |-
          Airflow dag import errors
            VALUE = {{ $value }}
            LABELS: {{ $labels }}
        summary: Airflow dag import errors (instance {{ $labels.instance }})
      expr: airflow_dag_processing_import_errors > 0
      for: 5m
      labels:
        severity: critical

如需从本地文件应用配置更改,请运行以下命令:

kubectl apply -n NAMESPACE_NAME -f FILE_NAME

您还可以使用 Terraform 管理您的配置。

如需详细了解如何将规则应用于您的集群,请参阅代管式规则评估和提醒

您可以根据您的应用调整提醒阈值。

验证配置

您可以使用 Metrics Explorer 验证您是否正确配置了 Airflow 导出器。Cloud Monitoring 可能需要一两分钟时间来注入您的指标。

要验证指标是否已注入,请执行以下操作:

  1. 在 Google Cloud 控制台的导航面板中,选择 Monitoring,然后选择  Metrics Explorer

    进入 Metrics Explorer

  2. 在查询构建器窗格的工具栏中,选择名为  MQL PromQL 的按钮。
  3. 验证已在语言切换开关中选择 PromQL。语言切换开关位于同一工具栏中,用于设置查询的格式。
  4. 输入并运行以下查询:
    up{job="airflow", cluster="CLUSTER_NAME", namespace="NAMESPACE_NAME"}

查看信息中心

Cloud Monitoring 集成包括 Airflow Prometheus 概览信息中心。当您配置集成时,系统会自动安装信息中心。您还可以在不安装集成的情况下查看信息中心的静态预览。

如需查看已安装的信息中心,请执行以下操作:

  1. 在 Google Cloud 控制台的导航面板中,选择 Monitoring,然后选择  信息中心

    前往信息中心

  2. 选择信息中心列表标签页。
  3. 选择集成类别。
  4. 点击信息中心的名称,例如 Airflow Prometheus 概览

如需查看信息中心的静态预览,请执行以下操作:

  1. 在 Google Cloud 控制台的导航面板中,选择 Monitoring,然后选择  集成

    前往集成

  2. 点击 Kubernetes Engine 部署平台过滤条件。
  3. 找到 Apache Airflow 集成,然后点击查看详情
  4. 选择信息中心标签页。

问题排查

如需了解如何排查指标注入问题,请参阅排查注入端问题中的从导出器收集的问题