Docker 与 K8S学习笔记（二十）—— 使用Downward API向容器注入Pod信息

2023-10-02 15:14:16

Kubernetes在创建Pod时，会为Pod和容器设置一些额外的信息，比如Pod名称、Pod IP、Node IP、Label、Annotation、资源限制等，我们经常会在应用程序中使用到这些数据，比如利用Pod名称作为应用日志的字段，方便分析日志。为了能在容器内获取这些信息，我们可以使用Downward API机制来实现。

Downward API可以通过环境变量和Volume挂载这两种方式将Pod信息注入容器，我们分别来看一下：

 

一、环境变量方式

我们还是以Busybox为例进行演示，我们将Pod信息和Container信息以环境变量方式注入容器，在容器启动后通过env命令打印出来，我们Yaml文件内容如下：

apiVersion: v1
kind: Pod
metadata:
  name: busybox-pod
spec:
  containers:
  - name: busybox
    image: busybox
    command: ["/bin/sh", "-c", "env | grep VAR_"]
    resources:
      requests:
        memory: "16Mi"
        cpu: "125m"
      limits:
        memory: "32Mi"
        cpu: "250m"
    env:
    - name: VAR_NODE_NAME
      valueFrom:
        fieldRef:
          fieldPath: spec.nodeName
    - name: VAR_POD_NAME
      valueFrom:
        fieldRef:
          fieldPath: metadata.name
    - name: VAR_POD_NAMESPACE
      valueFrom:
        fieldRef:
          fieldPath: metadata.namespace
    - name: VAR_POD_IP
      valueFrom:
        fieldRef:
          fieldPath: status.podIP
    - name: VAR_SERVICE_ACCOUNT
      valueFrom:
        fieldRef:
          fieldPath: spec.serviceAccountName
    - name: VAR_CPU_REQUEST
      valueFrom:
        resourceFieldRef:
          containerName: busybox
          resource: requests.cpu
    - name: VAR_CPU_LIMIT
      valueFrom:
        resourceFieldRef:
          containerName: busybox
          resource: limits.cpu
    - name: VAR_MEM_REQUEST
      valueFrom:
        resourceFieldRef:
          containerName: busybox
          resource: requests.memory
    - name: VAR_MEM_LIMIT
      valueFrom:
        resourceFieldRef:
          containerName: busybox
          resource: limits.memory
  restartPolicy: Never

我们创建Pod并使用kubtctl logs命令打印下输出：

$ sudo kubectl apply -f busy_pod.yaml
pod/busybox-pod created
$ sudo kubectl logs busybox-pod
VAR_MEM_REQUEST=16777216      # 容器内存请求值
VAR_NODE_NAME=ayato           # 节点名称
VAR_SERVICE_ACCOUNT=default   # Pod使用的ServiceAccount名称
VAR_CPU_REQUEST=1             # 容器cpu请求值
VAR_POD_NAME=busybox-pod      # pod名称
VAR_MEM_LIMIT=33554432        # 容器内存限制值
VAR_POD_NAMESPACE=default     # Pod所在命名空间
VAR_POD_IP=172.17.0.6         # Pod ip地址
VAR_CPU_LIMIT=1               # 容器cpu请求值

 

二、Volume挂载方式

我们接下来尝试使用Volume挂载方式，将Pod信息注入容器。还是以Busybox为例，由于Pod信息都是以文件方式注入容器，所以我们修改容器启动后执行命令：我们使用cat不断打印注入的文件，修改后的Yaml文件如下：

apiVersion: v1
kind: Pod
metadata:
  name: busybox-pod
  labels:
    cluster: demo-cluster
    type: tool-pod
  annotations:
    builder: alalazy
spec:
  containers:
  - name: busybox
    image: busybox
    command: ["/bin/sh", "-c"]
    args:
    - while true; do
        if [[ -e /etc/podinfo/labels ]]; then
          echo -en '\n\n'; cat /etc/podinfo/labels; fi;
        if [[ -e /etc/podinfo/annotations ]]; then
          echo -en '\n\n'; cat /etc/podinfo/annotations; fi;
        if [[ -e /etc/podinfo/cpu_limit ]]; then
          echo -en '\n\n'; cat /etc/podinfo/cpu_limit; fi;
        if [[ -e /etc/podinfo/cpu_request ]]; then
          echo -en '\n\n'; cat /etc/podinfo/cpu_request; fi;
        if [[ -e /etc/podinfo/mem_limit ]]; then
          echo -en '\n\n'; cat /etc/podinfo/mem_limit; fi;
        if [[ -e /etc/podinfo/mem_request ]]; then
          echo -en '\n\n'; cat /etc/podinfo/mem_request; fi;
        sleep 5;
      done;
    volumeMounts:
      - name: podinfo
        mountPath: /etc/podinfo
    resources:
      requests:
        memory: "16Mi"
        cpu: "125m"
      limits:
        memory: "32Mi"
        cpu: "250m"
  volumes:
    - name: podinfo
      downwardAPI:
        items:
          - path: "labels"
            fieldRef:
              fieldPath: metadata.labels
          - path: "annotations"
            fieldRef:
              fieldPath: metadata.annotations
          - path: "cpu_limit"
            resourceFieldRef:
              containerName: busybox
              resource: limits.cpu
              divisor: 1m
          - path: "cpu_request"
            resourceFieldRef:
              containerName: busybox
              resource: requests.cpu
              divisor: 1m
          - path: "mem_limit"
            resourceFieldRef:
              containerName: busybox
              resource: limits.memory
              divisor: 1Mi
          - path: "mem_request"
            resourceFieldRef:
              containerName: busybox
              resource: requests.memory
              divisor: 1Mi

我们创建此Pod，并通过kubectl logs查看输出：

$ sudo kubectl apply -f busy_pod.yaml
pod/busybox-pod created
$ sudo kubectl logs busybox-pod
cluster="demo-cluster"
type="tool-pod"

builder="alalazy"
kubectl.kubernetes.io/last-applied-configuration="{\"apiVersion\":\"v1\",\"kind\":\"Pod\",\"metadata\":{\"annotations\":{\"builder\":\"alalazy\"},\"labels\":{\"cluster\":\"demo-cluster\",\"type\":\"tool-pod\"},\"name\":\"busybox-pod\",\"namespace\":\"default\"},\"spec\":{\"containers\":[{\"args\":[\"while true; do if [[ -e /etc/podinfo/labels ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/labels; fi; if [[ -e /etc/podinfo/annotations ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/annotations; fi; if [[ -e /etc/podinfo/cpu_limit ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/cpu_limit; fi; if [[ -e /etc/podinfo/cpu_request ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/cpu_request; fi; if [[ -e /etc/podinfo/mem_limit ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/mem_limit; fi; if [[ -e /etc/podinfo/mem_request ]]; then echo -en '\\\\n\\\\n'; cat /etc/podinfo/mem_request; fi; sleep 5; done;\"],\"command\":[\"/bin/sh\",\"-c\"],\"image\":\"busybox\",\"name\":\"busybox\",\"resources\":{\"limits\":{\"cpu\":\"250m\",\"memory\":\"32Mi\"},\"requests\":{\"cpu\":\"125m\",\"memory\":\"16Mi\"}},\"volumeMounts\":[{\"mountPath\":\"/etc/podinfo\",\"name\":\"podinfo\"}]}],\"volumes\":[{\"downwardAPI\":{\"items\":[{\"fieldRef\":{\"fieldPath\":\"metadata.labels\"},\"path\":\"labels\"},{\"fieldRef\":{\"fieldPath\":\"metadata.annotations\"},\"path\":\"annotations\"},{\"path\":\"cpu_limit\",\"resourceFieldRef\":{\"containerName\":\"busybox\",\"divisor\":\"1m\",\"resource\":\"limits.cpu\"}},{\"path\":\"cpu_request\",\"resourceFieldRef\":{\"containerName\":\"busybox\",\"divisor\":\"1m\",\"resource\":\"requests.cpu\"}},{\"path\":\"mem_limit\",\"resourceFieldRef\":{\"containerName\":\"busybox\",\"divisor\":\"1Mi\",\"resource\":\"limits.memory\"}},{\"path\":\"mem_request\",\"resourceFieldRef\":{\"containerName\":\"busybox\",\"divisor\":\"1Mi\",\"resource\":\"requests.memory\"}}]},\"name\":\"podinfo\"}]}}\n"
kubernetes.io/config.seen="2022-01-15T05:33:53.379386410Z"
kubernetes.io/config.source="api"

250

125

32

16

我们进入容器查看下挂载的文件：

$  sudo kubectl exec -it busybox-pod -- sh
/ # cd /etc/podinfo/
/etc/podinfo # ls
annotations  cpu_limit    cpu_request  labels       mem_limit    mem_request

 

三、Downward API的能力

我们可以通过Downward API向容器注入如下信息：

1）可通过fieldRef获得的信息：

  • metadata.name：Pod 名称

  • metadata.namespace：Pod 名字空间

  • metadata.uid：Pod 的 UID

  • metadata.labels['<KEY>']：Pod标签 <KEY> 的值 (例如, metadata.labels['mylabel']）

  • metadata.annotations['<KEY>']：Pod 的注解 <KEY> 的值（例如, metadata.annotations['myannotation']）

  • metadata.labels：获取所有标签

  • metadata.annotations：获取所有注解

  • status.podIP：节点 IP

  • spec.serviceAccountName：Pod 服务帐号名称, 版本要求v1.4.0-alpha.3

  • spec.nodeName：节点名称, 版本要求 v1.4.0-alpha.3

  • status.hostIP：节点 IP, 版本要求 v1.7.0-alpha.1

2）可通过 resourceFieldRef 获得的信息：

  • 容器的 CPU 约束值

  • 容器的 CPU 请求值

  • 容器的内存约束值

  • 容器的内存请求值

  • 容器的巨页限制值（前提是启用了DownwardAPIHugePages 特性门控）

  • 容器的巨页请求值（前提是启用了DownwardAPIHugePages 特性门控）

  • 容器的临时存储约束值

  • 容器的临时存储请求值

上一篇：我们可以定向调度某个pod在某个node上进行创建


下一篇：k8s(三)：命令行工具kubectl与核心技术Pod