在 Google Kubernetes Engine 上部署 Google 构建的 OpenTelemetry 收集器

# Copyright 2024 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
#
#     http://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.

exporters:
  googlecloud:
    log:
      default_log_name: opentelemetry-collector
    user_agent: Google-Cloud-OTLP manifests:0.4.0 OpenTelemetry Collector Built By Google/0.128.0 (linux/amd64)
  googlemanagedprometheus:
    user_agent: Google-Cloud-OTLP manifests:0.4.0 OpenTelemetry Collector Built By Google/0.128.0 (linux/amd64)
  # The otlphttp exporter is used to send traces to Google Cloud Trace using OTLP http/proto
  # The otlp exporter could also be used to send them using OTLP grpc
  otlphttp:
    encoding: proto
    endpoint: https://telemetry.googleapis.com
    # Use the googleclientauth extension to authenticate with Google credentials
    auth:
      authenticator: googleclientauth


extensions:
  health_check:
    endpoint: ${env:MY_POD_IP}:13133
  googleclientauth:


processors:
  filter/self-metrics:
    metrics:
      include:
        match_type: strict
        metric_names:
        - otelcol_process_uptime
        - otelcol_process_memory_rss
        - otelcol_grpc_io_client_completed_rpcs
        - otelcol_googlecloudmonitoring_point_count
  batch:
    send_batch_max_size: 200
    send_batch_size: 200
    timeout: 5s

  k8sattributes:
    extract:
      metadata:
      - k8s.namespace.name
      - k8s.deployment.name
      - k8s.statefulset.name
      - k8s.daemonset.name
      - k8s.cronjob.name
      - k8s.job.name
      - k8s.replicaset.name
      - k8s.node.name
      - k8s.pod.name
      - k8s.pod.uid
      - k8s.pod.start_time
    passthrough: false
    pod_association:
    - sources:
      - from: resource_attribute
        name: k8s.pod.ip
    - sources:
      - from: resource_attribute
        name: k8s.pod.uid
    - sources:
      - from: connection

  memory_limiter:
    check_interval: 1s
    limit_percentage: 65
    spike_limit_percentage: 20

  metricstransform/self-metrics:
    transforms:
    - action: update
      include: otelcol_process_uptime
      operations:
      - action: add_label
        new_label: version
        new_value: Google-Cloud-OTLP manifests:0.4.0 OpenTelemetry Collector Built By Google/0.128.0 (linux/amd64)

  resourcedetection:
    detectors: [gcp]
    timeout: 10s

  transform/collision:
    metric_statements:
    - context: datapoint
      statements:
      - set(attributes["exported_location"], attributes["location"])
      - delete_key(attributes, "location")
      - set(attributes["exported_cluster"], attributes["cluster"])
      - delete_key(attributes, "cluster")
      - set(attributes["exported_namespace"], attributes["namespace"])
      - delete_key(attributes, "namespace")
      - set(attributes["exported_job"], attributes["job"])
      - delete_key(attributes, "job")
      - set(attributes["exported_instance"], attributes["instance"])
      - delete_key(attributes, "instance")
      - set(attributes["exported_project_id"], attributes["project_id"])
      - delete_key(attributes, "project_id")

  # The relative ordering of statements between ReplicaSet & Deployment and Job & CronJob are important.
  # The ordering of these controllers is decided based on the k8s controller documentation available at
  # https://kubernetes.io/docs/concepts/workloads/controllers.
  # The relative ordering of the other controllers in this list is inconsequential since they directly
  # create pods.
  transform/aco-gke:
    metric_statements:
    - context: datapoint
      statements:
      - set(attributes["top_level_controller_type"], "ReplicaSet") where resource.attributes["k8s.replicaset.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.replicaset.name"]) where resource.attributes["k8s.replicaset.name"] != nil
      - set(attributes["top_level_controller_type"], "Deployment") where resource.attributes["k8s.deployment.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.deployment.name"]) where resource.attributes["k8s.deployment.name"] != nil
      - set(attributes["top_level_controller_type"], "DaemonSet") where resource.attributes["k8s.daemonset.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.daemonset.name"]) where resource.attributes["k8s.daemonset.name"] != nil
      - set(attributes["top_level_controller_type"], "StatefulSet") where resource.attributes["k8s.statefulset.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.statefulset.name"]) where resource.attributes["k8s.statefulset.name"] != nil
      - set(attributes["top_level_controller_type"], "Job") where resource.attributes["k8s.job.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.job.name"]) where resource.attributes["k8s.job.name"] != nil
      - set(attributes["top_level_controller_type"], "CronJob") where resource.attributes["k8s.cronjob.name"] != nil
      - set(attributes["top_level_controller_name"], resource.attributes["k8s.cronjob.name"]) where resource.attributes["k8s.cronjob.name"] != nil

  # When sending telemetry to the GCP OTLP endpoint, the gcp.project_id resource attribute is required to be set to your project ID.
  resource/gcp_project_id:
    attributes:
    - key: gcp.project_id
      # MAKE SURE YOU REPLACE THIS WITH YOUR PROJECT ID
      value: ${GOOGLE_CLOUD_PROJECT}
      action: insert
  # The metricstarttime processor is important to include if you are using the prometheus receiver to ensure the start time is set properly.
  # It is a no-op otherwise.
  metricstarttime:
    strategy: subtract_initial_point

receivers:
  # This collector is configured to accept OTLP metrics, logs, and traces, and is designed to receive OTLP from workloads running in the cluster.
  otlp:
    protocols:
      grpc:
        endpoint: ${env:MY_POD_IP}:4317
      http:
        cors:
          allowed_origins:
          - http://*
          - https://*
        endpoint: ${env:MY_POD_IP}:4318
  otlp/self-metrics:
    protocols:
      grpc:
        endpoint: ${env:MY_POD_IP}:14317

service:
  extensions:
  - health_check
  - googleclientauth
  pipelines:
    logs:
      exporters:
      - googlecloud
      processors:
      - k8sattributes
      - resourcedetection
      - memory_limiter
      - batch
      receivers:
      - otlp
    metrics/otlp:
      exporters:
      - googlemanagedprometheus
      processors:
      - k8sattributes
      - memory_limiter
      - metricstarttime
      - resourcedetection
      - transform/collision
      - transform/aco-gke
      - batch
      receivers:
      - otlp
    metrics/self-metrics:
      exporters:
      - googlemanagedprometheus
      processors:
      - filter/self-metrics
      - metricstransform/self-metrics
      - k8sattributes
      - memory_limiter
      - resourcedetection
      - batch
      receivers:
      - otlp/self-metrics
    traces:
      exporters:
      - otlphttp
      processors:
      - k8sattributes
      - memory_limiter
      - resource/gcp_project_id
      - resourcedetection
      - batch
      receivers:
      - otlp
  telemetry:
    logs:
      encoding: json
    metrics:
      readers:
      - periodic:
          exporter:
            otlp:
              protocol: grpc
              endpoint: ${env:MY_POD_IP}:14317