日志架构演进：从集中式到分布式的Kubernetes日志策略

# 集中式日志记录（Kubernetes 集群之前）
apiVersion: logging.k8s.io/v1beta1
kind: ClusterLogging
metadata:
  name: cluster-logging-intake
spec:
  fluentd:
    baseImage: k8s.gcr.io/fluentd-elasticsearch
    resources:
      limits:
        memory: 200Mi
        cpu: 100m
      requests:
        memory: 100Mi
        cpu: 50m
  logfile:
    storage: 1Gi
 
# 分布式日志记录（Kubernetes 集群内）
apiVersion: logging.k8s.io/v1beta1
kind: ClusterLogging
metadata:
  name: cluster-logging-distributed
spec:
  logStore: elasticsearch
  elasticsearch:
    node:
      resources:
        limits:
          memory: 1Gi
          cpu: 500m
        requests:
          memory: 1Gi
          cpu: 500m
    config:
      cluster.name: ${ELASTICSEARCH_CLUSTER_NAME}
      node.name: "${POD_NAME}.${POD_NAMESPACE}.svc"
      network.host: 0.0.0.0
      discovery.seed_hosts: ${ELASTICSEARCH_SERVICE_ENDPOINT}
 
# 分布式日志记录（Kubernetes 集群外）
apiVersion: logging.k8s.io/v1beta1
kind: ClusterLogging
metadata:
  name: cluster-logging-distributed-external
spec:
  logStore: elasticsearch
  elasticsearchConfig:
    clusterName: ${ELASTICSEARCH_CLUSTER_NAME}
    host: ${ELASTICSEARCH_HOST}
    port: ${ELASTICSEARCH_PORT}

这个代码实例展示了如何使用Kubernetes的ClusterLogging资源来定义集中式和分布式日志记录。它演示了如何为Fluentd日志收集器设置资源限制，以及如何为Elasticsearch设置配置选项。代码还展示了如何引用环境变量以便在不同环境中灵活配置。