OpenShift：アプリケーションの構成と運用

oc explain によるliveness Probe設定値の確認コマンド

# oc explain pod.spec.containers.livenessProbe

リスト1：ディレクトリ内のファイルリスト

# ls
Dockerfile  main.go

リスト2：Dockerfile

FROM golang:alpine

WORKDIR /app
COPY main.go .

CMD go run main.go

リスト3：検証用アプリケーションのコード

package main

import (
    "fmt"
    "net/http"
    "time"
)

func rootHandler(w http.ResponseWriter, r *http.Request) {
    fmt.Fprintf(w, "Hello, World")
}

func healthHandler(w http.ResponseWriter, r *http.Request) {
    fmt.Printf("%v: request from: %v\n", time.Now(), r.RemoteAddr)
}

func main() {
    http.HandleFunc("/", rootHandler)
    http.HandleFunc("/health", healthHandler)
    http.ListenAndServe(":8080", nil)
}

ビルドとデプロイの手順

$ oc new-build --name=go-httpd --strategy=docker --binary
$ oc start-build go-httpd --from-dir=. --follow
$ oc new-app go-httpd

oc set probeコマンドの例

$ oc set probe dc/go-httpd --liveness --get-url=http://:8080/health

Podのログを確認

# oc logs go-httpd-4-vxwjp
2019-01-06 07:47:45.798994545 +0000 UTC m=+4.165327161: request from: 10.128.0.1:54808
2019-01-06 07:47:55.897736149 +0000 UTC m=+14.264069102: request from: 10.128.0.1:54878
2019-01-06 07:48:05.799532904 +0000 UTC m=+24.165865675: request from: 10.128.0.1:54988
2019-01-06 07:48:15.7983669 +0000 UTC m=+34.164699506: request from: 10.128.0.1:55078
2019-01-06 07:48:25.798541175 +0000 UTC m=+44.164873661: request from: 10.128.0.1:55142
2019-01-06 07:48:35.798272671 +0000 UTC m=+54.164605630: request from: 10.128.0.1:55244

goプロセスを停止してみる

$ oc exec go-httpd-4-vxwjp -- pkill -STOP go

eventログを確認

$ oc get event -w
0s        0s        1         go-httpd-4-k7kc7.157c67383af6b0b8   Pod       spec.containers{go-httpd}   Warning   Unhealthy   kubelet, knakayam-ose311-all   Liveness probe failed: Get http://10.128.1.138:8080/health: net/http: request canceled (Client.Timeout exceeded while awaiting headers)
0s        10s       2         go-httpd-4-k7kc7.157c67383af6b0b8   Pod       spec.containers{go-httpd}   Warning   Unhealthy   kubelet, knakayam-ose311-all   Liveness probe failed: Get http://10.128.1.138:8080/health: net/http: request canceled (Client.Timeout exceeded while awaiting headers)
0s        20s       3         go-httpd-4-k7kc7.157c67383af6b0b8   Pod       spec.containers{go-httpd}   Warning   Unhealthy   kubelet, knakayam-ose311-all   Liveness probe failed: Get http://10.128.1.138:8080/health: net/http: request canceled (Client.Timeout exceeded while awaiting headers)
0s        0s        1         go-httpd-4-k7kc7.157c673d04c21ac6   Pod       spec.containers{go-httpd}   Normal    Killing   kubelet, knakayam-ose311-all   Killing container with id docker://go-httpd:Container failed liveness probe.. Container will be killed and recreated.
0s        58s       2         go-httpd-4-k7kc7.157c672f9672fdb6   Pod       spec.containers{go-httpd}   Normal    Pulling   kubelet, knakayam-ose311-all   pulling image "docker-registry.default.svc:5000/ch05/go-httpd@sha256:2643d3821c64bfd0a51d5f7753e4fa3f03cbe703e33058f354a3b7210da1a3a8"
0s        58s       2         go-httpd-4-k7kc7.157c672fb212efda   Pod       spec.containers{go-httpd}   Normal    Pulled    kubelet, knakayam-ose311-all   Successfully pulled image "docker-registry.default.svc:5000/ch05/go-httpd@sha256:2643d3821c64bfd0a51d5f7753e4fa3f03cbe703e33058f354a3b7210da1a3a8"
0s        58s       2         go-httpd-4-k7kc7.157c672fb75df174   Pod       spec.containers{go-httpd}   Normal    Created   kubelet, knakayam-ose311-all   Created container
0s        58s       2         go-httpd-4-k7kc7.157c672fccab6180   Pod       spec.containers{go-httpd}   Normal    Started   kubelet, knakayam-ose311-all   Started container

リスト4：init Containerの設定例

  initContainers:
  - name: init-myservice
    image: busybox
    command: ['sh', '-c', 'until nslookup myservice; do echo waiting for myservice; sleep 2; done;']
  containers:
    ...略 ...

リスト5：NFSをバックエンドに使うPVの定義

cat <<EOF > ./pv1.yaml
apiVersion: "v1"
kind: "PersistentVolume"
metadata:
  name: "pv0001"
spec:
  capacity:
    storage: "1Gi"
  accessModes:
    - "ReadWriteMany"
  nfs:
    path: "/nfs"
    server: "192.168.133.4"  # ホスト名/IPはご利用の環境に合わせて変更する必要があります。
  persistentVolumeReclaimPolicy: "Recycle"
EOF

PVの作成コマンド

oc create -f pv1.yaml

リスト6：PVCの定義

cat <<EOF > ./pvc.yaml
apiVersion: "v1"
kind: "PersistentVolumeClaim"
metadata:
  name: "claim1"
spec:
  accessModes:
    - "ReadWriteMany"
  resources:
    requests:
      storage: "1Gi"
EOF

PVCの作成コマンド

oc create -f pvc.yaml

PVC確認コマンド

$ oc get pvc
NAME      STATUS    VOLUME    CAPACITY   ACCESS MODES   STORAGECLASS   AGE
claim1    Bound     pv0001    1Gi        RWX                           2s

oc set volumeコマンドによるDeploymentConfigの更新

$ oc set volume dc/go-httpd --add -t pvc --name=test-vol --claim-name=claim1 --mount-path=/mnt

oc set volumeコマンドによるマウントポイントの確認

$ oc set volume pod/go-httpd-5-5892n
deploymentconfigs/go-httpd
  pvc/claim1 (allocated 1GiB) as test-vol
    mounted at /mnt

oc get eventコマンドによるイベントログの確認

$ oc get event

永続ストレージ上のファイルは残ることを確認

# oc rsh go-httpd-5-mnkzk touch /mnt/test-file

# oc delete pod go-httpd-5-mnkzk
pod "go-httpd-5-mnkzk" deleted

# oc rsh go-httpd-5-z62nj ls /mnt
test-file