애플리케이션 볼륨 사용법

빈 디렉터리(emptyDir)

호스트 경로(hostPath)

NFS 볼륨

퍼시스턴트볼륨(PV)와 퍼시스턴트볼륨클레임(PVC)

스토리지클래스(StorageClass)

볼륨클레임템플릿(volumeClaimTemplates)

볼륨

빈 디렉터리(emptyDir)

•

하나의 Pod 내에서 컨테이너끼리 공유하는 공간이다.

apiVersion: v1
kind: Pod
metadata:
  name: pod-emptydir
  labels:
    app: nginx
spec:
  containers:
  - name: web-page  // 실제 서비스가 올라가는 컨테이너
    image: nginx
    volumeMounts:
    - mountPath: /usr/share/nginx/html
      name: empty-directory

  - name: html-builder // 페이지가 만들어지는 컨테이너
    image: alpine
    volumeMounts:
    - mountPath: /html-dir
      name: empty-directory
    command: ["/bin/sh", "-c"]
    args:
      - echo "This page created on $(date +%Y-%m-%d)" > /html-dir/index.html;
        sleep infinity;

  volumes:
  - name: empty-directory  // 위의 두 개의 컨테이너를 공유
    emptyDir: {}

NAME           READY   STATUS             RESTARTS   AGE    IP               NODE     NOMINATED NODE   READINESS GATES
net            0/1     ImagePullBackOff   0          5d1h   172.16.221.158   w1-k8s   <none>           <none>
pod-emptydir   2/2     Running            0          20s    172.16.132.53    w3-k8s   <none>           <none>
YAML
복사

호스트 경로(hostPath)

•

노드에 있는 디렉터리를 볼륨으로 사용한다.

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: ds-hostpath
  labels:
    app: ds-hostpath
spec:
  selector:
    matchLabels:
      app: ds-hostpath
  template:
    metadata:
      labels:
        app: ds-hostpath
    spec:
      containers:
      - name: host-mon
        image: sysnet4admin/sleepy
        volumeMounts:
        - mountPath: /host-log
          name: hostpath-directory
      volumes:
      - name: hostpath-directory
        hostPath:
          path: /var/log
YAML
복사

NFS 볼륨

# NFS 서비스 구동
[root@m-k8s ~]# ~/_Lecture_k8s_starter.kit/ch5/5.2/nfs-exporter.sh log
Created symlink from /etc/systemd/system/multi-user.target.wants/nfs-server.service to /usr/lib/systemd/system/nfs-server.service.

# NFS 설정파일 확인(log폴더에 192.168.1.0으로 붙도록 마운트)
[root@m-k8s ~]# cat /etc/exports
/nfs_shared/log 192.168.1.0/24(rw,sync,no_root_squash)

# 접속 시 접속 로그를 /audit에 저장하는 Pod 배포
[root@m-k8s nfs_shared]# kubectl apply -f ~/_Lecture_k8s_starter.kit/ch5/5.2/dpy-chk-log.yaml
deployment.apps/dpy-chk-log created

# 내용을 살펴보면 volumes를 앞에서 올린 nfs_shared/log로 사용하도록 설정되어 있으며 /audit과 마운트되도록 함
apiVersion: apps/v1
kind: Deployment
metadata:
  name: deploy-nfs
  labels:
    app: deploy-nfs
spec:
  replicas: 3
  selector:
    matchLabels:
      app: deploy-nfs
  template:
    metadata:
      labels:
        app: deploy-nfs
    spec:
      containers:
      - name: chk-log
        image: sysnet4admin/chk-log
        volumeMounts:
        - name: nfs-vol
          mountPath: /audit
      volumes:
      - name: nfs-vol
        nfs:
          server: 192.168.1.10
          path: /nfs_shared/nfs-vol

# pods 확인
[root@m-k8s nfs_shared]# kubectl get pods -o wide
NAME                           READY   STATUS    RESTARTS   AGE     IP               NODE     NOMINATED NODE   READINESS GATES
dpy-chk-log-655668ffb8-7m4jm   1/1     Running   0          9m57s   172.16.132.14    w3-k8s   <none>           <none>
dpy-chk-log-655668ffb8-c4wwt   1/1     Running   0          9m57s   172.16.221.138   w1-k8s   <none>           <none>
dpy-chk-log-655668ffb8-r84j2   1/1     Running   0          9m57s   172.16.103.142   w2-k8s   <none>           <none>

# pod에 접속
[root@m-k8s nfs_shared]# curl 172.16.132.14
pod_n: dpy-chk-log-655668ffb8-7m4jm | ip_dest: 172.16.132.14

# pod에 shell로 접속해서 audit 내 로그 확인
[root@m-k8s nfs_shared]# kubectl exec dpy-chk-log-655668ffb8-7m4jm -it -- /bin/bash
root@dpy-chk-log-655668ffb8-7m4jm:/# cat /audit/audit_dpy-chk-log-655668ffb8-7m4jm.log
01/Nov/2022:00:29:48 +0900  172.16.132.14  GET
root@dpy-chk-log-655668ffb8-7m4jm:/# exit
exit

# 배포한 Pod 삭제
[root@m-k8s log]# kubectl delete -f ~/_Lecture_k8s_starter.kit/ch5/5.2/dpy-chk-log.yaml
deployment.apps "dpy-chk-log" deleted

[root@m-k8s log]# kubectl get pods
No resources found in default namespace.

# 배포한 Pod을 삭제해도 마스터노드에 볼륨 내 파일이 그대로 남아있음
[root@m-k8s log]# cat /nfs_shared/log/audit_dpy-chk-log-655668ffb8-7m4jm.log
01/Nov/2022:00:29:48 +0900  172.16.132.14  GET
YAML
복사

퍼시스턴트볼륨(PV)와 퍼시스턴트볼륨클레임(PVC)

•

NFS의 볼륨을 보완한다.

•

관리자가 PV를 생성하면 개발자가 PV를 사용하기 위해 PVC를 통해서 Claim한다.

[정적 프로비저닝]

# 관리자가 생성하는 PV
apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-nfs
spec:
  capacity:
    storage: 100Mi
  accessModes:
    - ReadWriteMany
  persistentVolumeReclaimPolicy: Retain
  nfs:
    server: 192.168.1.10
    path: /nfs_shared/pvc-vol
YAML
복사

접근 모드	설명
RWO	하나의 노드에서만 볼륨을 읽고 쓸 수 있게 마운트
ROX	여러 개의 노드가 읽도록 마운트
RWX	여러 개의 노드가 읽고 쓸 수 있도록 마운트

반환 정책	설명
Retain(보존)	PVC 삭제 시에도 PV를 보존
Delete(삭제)	PVC 삭제 시 PV를 함께 삭제
Recycle(재활용)	더이상 사용되지 않음

# 개발자가 생성하는 PVC
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pvc-nfs
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 10Mi
YAML
복사

# Deploy
apiVersion: apps/v1
kind: Deployment
metadata:
  name: deploy-pvc
  labels:
    app: deploy-pvc
spec:
  replicas: 3
  selector:
    matchLabels:
      app: deploy-pvc
  template:
    metadata:
      labels:
        app: deploy-pvc
    spec:
      containers:
      - name: chk-log
        image: sysnet4admin/chk-log
        volumeMounts:
        - name: pvc-vol
          mountPath: /audit
      volumes:
      - name: pvc-vol
        persistentVolumeClaim:
          claimName: pvc-nfs
YAML
복사

# 1. NFS 볼륨을 생성한다.
[root@m-k8s 5.5]# ./nfs-exporter.sh pvc-vol
Created symlink from /etc/systemd/system/multi-user.target.wants/nfs-server.service to /usr/lib/systemd/system/nfs-server.service.

[root@m-k8s 5.5]# cat /etc/exports
/nfs_shared/pvc-vol 192.168.1.0/24(rw,sync,no_root_squash)

# 2. PV를 생성한다.
[root@m-k8s 5.5]# k apply -f persistentvolume-nfs.yaml
persistentvolume/pv-nfs created

[root@m-k8s 5.5]# k get pv
NAME     CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS      CLAIM   STORAGECLASS   REASON   AGE
pv-nfs   100Mi      RWX            Retain           Available                                   24s

# 3. PVC를 생성한다.
[root@m-k8s 5.5]# k apply -f persistentvolumeclaim-nfs.yaml
persistentvolumeclaim/pvc-nfs created

[root@m-k8s 5.5]# k get pvc
NAME      STATUS   VOLUME   CAPACITY   ACCESS MODES   STORAGECLASS   AGE
pvc-nfs   Bound    pv-nfs   100Mi      RWX                           3s

# 4. Deploy를 배포한다.
[root@m-k8s 5.5]# k apply -f deploy-pvc.yaml
deployment.apps/deploy-pvc unchanged

# 5. curl로 로그를 남긴다.
[root@m-k8s 5.5]# k get po -o wide
NAME                          READY   STATUS              RESTARTS   AGE   IP               NODE     NOMINATED NODE   READINESS GATES
deploy-pvc-57684fb7ff-mv9zp   0/1     ContainerCreating   0          35s   <none>           w3-k8s   <none>           <none>
deploy-pvc-57684fb7ff-n4rws   1/1     Running             0          35s   172.16.103.167   w2-k8s   <none>           <none>
deploy-pvc-57684fb7ff-wvcxz   1/1     Running             0          35s   172.16.221.155   w1-k8s   <none>           <none>
net                           1/1     Running             0          9d    172.16.132.23    w3-k8s   <none>           <none>

[root@m-k8s 5.5]# curl 172.16.103.167
pod_n: deploy-pvc-57684fb7ff-n4rws | ip_dest: 172.16.103.167

# 6. pod에 접속해서 볼륨에 생성된 로그를 확인해본다.
[root@m-k8s 5.5]# k exec deploy-pvc-57684fb7ff-n4rws -it -- /bin/bash
root@deploy-pvc-57684fb7ff-n4rws:/# ls audit
audit_deploy-pvc-57684fb7ff-n4rws.log
root@deploy-pvc-57684fb7ff-n4rws:/# exit
exit
[root@m-k8s 5.5]# k exec deploy-pvc-57684fb7ff-mv9zp -it -- /bin/bash
root@deploy-pvc-57684fb7ff-mv9zp:/# ls audit
audit_deploy-pvc-57684fb7ff-n4rws.log
YAML
복사

스토리지클래스(StorageClass)

•

PVC → SC → PV

[NFS Provisioner를 이용한 동적 프로비저닝]

볼륨클레임템플릿(volumeClaimTemplates)

•

statefulset에서만 사용할 수 있는 볼륨 Type이다.

•

SC와도 연동되어 동작한다.

볼륨

•

영속적인 데이터를 보존하기 위해서 Pod를 만들 때 Volume을 붙여서 사용함

# NFS 서비스 구동
[root@m-k8s ~]# ~/_Lecture_k8s_starter.kit/ch5/5.2/nfs-exporter.sh log
Created symlink from /etc/systemd/system/multi-user.target.wants/nfs-server.service to /usr/lib/systemd/system/nfs-server.service.

# NFS 설정파일 확인(log폴더에 192.168.1.0으로 붙도록 마운트)
[root@m-k8s ~]# cat /etc/exports
/nfs_shared/log 192.168.1.0/24(rw,sync,no_root_squash)

[root@m-k8s ~]# cd /nfs_shared
[root@m-k8s nfs_shared]# ls
log

# 접속 시 접속 로그를 /audit에 저장하는 Pod 배포
[root@m-k8s nfs_shared]# kubectl apply -f ~/_Lecture_k8s_starter.kit/ch5/5.2/dpy-chk-log.yaml
deployment.apps/dpy-chk-log created

# 내용을 살펴보면 volumes를 앞에서 올린 nfs_shared/log로 사용하도록 설정되어 있으며 /audit과 마운트되도록 함
[root@m-k8s nfs_shared]# cat ~/_Lecture_k8s_starter.kit/ch5/5.2/dpy-chk-log.yaml
spec:
      containers:
      - name: dpy-chk-log
        image: sysnet4admin/chk-log
        volumeMounts:
        - name: nfs-vol
          mountPath: /audit
      volumes:
      - name: nfs-vol
        nfs:
          server: 192.168.1.10
          path: /nfs_shared/log

# pods 확인
[root@m-k8s nfs_shared]# kubectl get pods -o wide
NAME                           READY   STATUS    RESTARTS   AGE     IP               NODE     NOMINATED NODE   READINESS GATES
dpy-chk-log-655668ffb8-7m4jm   1/1     Running   0          9m57s   172.16.132.14    w3-k8s   <none>           <none>
dpy-chk-log-655668ffb8-c4wwt   1/1     Running   0          9m57s   172.16.221.138   w1-k8s   <none>           <none>
dpy-chk-log-655668ffb8-r84j2   1/1     Running   0          9m57s   172.16.103.142   w2-k8s   <none>           <none>

# pod에 접속
[root@m-k8s nfs_shared]# curl 172.16.132.14
pod_n: dpy-chk-log-655668ffb8-7m4jm | ip_dest: 172.16.132.14

# pod에 shell로 접속해서 audit 내 로그 확인
[root@m-k8s nfs_shared]# kubectl exec dpy-chk-log-655668ffb8-7m4jm -it -- /bin/bash
root@dpy-chk-log-655668ffb8-7m4jm:/# cat /audit/audit_dpy-chk-log-655668ffb8-7m4jm.log
01/Nov/2022:00:29:48 +0900  172.16.132.14  GET
root@dpy-chk-log-655668ffb8-7m4jm:/# exit
exit

# 배포한 Pod 삭제
[root@m-k8s log]# kubectl delete -f ~/_Lecture_k8s_starter.kit/ch5/5.2/dpy-chk-log.yaml
deployment.apps "dpy-chk-log" deleted

[root@m-k8s log]# kubectl get pods
No resources found in default namespace.

# 배포한 Pod을 삭제해도 마스터노드에 볼륨 내 파일이 그대로 남아있음
[root@m-k8s log]# cat /nfs_shared/log/audit_dpy-chk-log-655668ffb8-7m4jm.log
01/Nov/2022:00:29:48 +0900  172.16.132.14  GET
Shell
복사