*Pod = 컨테이너 그룹
rc로 제어되는 pod는 삭제하면 다시 생성된다
반면 자체적으로 만든 pod는 삭제하면 재생성되지 않음
*멀티 컨테이너 :
- create pod
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | multi.yaml apiVersion: v1 kind: Pod metadata: name: multi labels: app: wordpress spec: containers: - name: wordpress image: reg.cloud.com/wordpress ports: - containerPort: 80 - containerPort: 443 env: - name: WORDPRESS_DB_HOST value: 127.0.0.1 - name: WORDPRESS_DB_USER value: wordpress - name: WORDPRESS_DB_PASSWORD value: password - name: db image: reg.cloud.com/mysql env: - name: MYSQL_ROOT_PASSWORD value: password - name: MYSQL_DATABASE value: wordpress - name: MYSQL_USER value: wordpress - name: MYSQL_PASSWORD value: password [root@host01-4 hk]# k create -f multi.yaml | cs |
- 멀티 컨테이너에서는 컨테이너를 지정해서 들어갈 수 있다(DB랑 web이 하나의 포드로 묶여있기 때문에 좋은 구성은 아니다).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 | [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE multi 2/2 Running 1 1m 10.44.0.2 host01-2.cloud.com nginx-app-7qpbv 1/1 Running 0 1h 10.36.0.1 host01-3.cloud.com nginx-app2-tgqqf 1/1 Running 0 1h 10.36.0.2 host01-3.cloud.com nginx-hk-app 1/1 Running 0 1h 10.44.0.1 host01-2.cloud.com [root@host01-4 hk]# k exec -it multi -c wordpress sh # ip a 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 18: eth0@if19: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1376 qdisc noqueue state UP group default link/ether 32:a0:1d:fb:91:b1 brd ff:ff:ff:ff:ff:ff inet 10.44.0.2/12 brd 10.47.255.255 scope global eth0 valid_lft forever preferred_lft forever inet6 fe80::30a0:1dff:fefb:91b1/64 scope link tentative dadfailed valid_lft forever preferred_lft forever # ^C # hostname multi # [root@host01-4 hk]# k exec -it multi -c db sh # mysql ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: NO) # mysql -uroot -p Enter password: ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: YES) # | cs |
*Label 정보 확인
1 2 3 4 5 6 7 8 9 | [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS multi 2/2 Running 1 5m app=wordpress nginx-app-7qpbv 1/1 Running 0 1h run=nginx-app nginx-app2-tgqqf 1/1 Running 0 1h type=test nginx-hk-app 1/1 Running 0 1h type=web [root@host01-4 hk]# | cs |
*Label 조회/생성/Overwrite
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS multi 2/2 Running 1 5m app=wordpress nginx-app-7qpbv 1/1 Running 0 1h run=nginx-app nginx-app2-tgqqf 1/1 Running 0 1h type=test nginx-hk-app 1/1 Running 0 1h type=web [root@host01-4 hk]# ^C [root@host01-4 hk]# k label po simple-pod loc=seoul Error from server (NotFound): pods "simple-pod" not found [root@host01-4 hk]# k label po nginx-app-7qpbv loc=seoul pod "nginx-app-7qpbv" labeled [root@host01-4 hk]# k label po nginx-app-7qpbv type=was --overwrite pod "nginx-app-7qpbv" labeled [root@host01-4 hk]# | cs |
*Pod 찾기(-l 옵션으로 key=value 방식으로 가져올 수 있음) :
1 2 3 4 5 6 7 8 9 10 11 | [root@host01-4 hk]# k get po -l type=test NAME READY STATUS RESTARTS AGE nginx-app2-tgqqf 1/1 Running 0 1h [root@host01-4 hk]# ^C [root@host01-4 hk]# k get po -l type NAME READY STATUS RESTARTS AGE nginx-app-7qpbv 1/1 Running 0 1h nginx-app2-tgqqf 1/1 Running 0 1h nginx-hk-app 1/1 Running 0 1h [root@host01-4 hk]# | cs |
*Pod 찾기(-L : Label로 찾기)
1 2 3 4 5 6 7 8 | [root@host01-4 hk]# k get po -L app,type,loc NAME READY STATUS RESTARTS AGE APP TYPE LOC multi 2/2 Running 1 11m wordpress nginx-app-7qpbv 1/1 Running 0 1h was seoul nginx-app2-tgqqf 1/1 Running 0 1h test nginx-hk-app 1/1 Running 0 1h web [root@host01-4 hk]# | cs |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 | [root@host01-4 hk]# k get po NAME READY STATUS RESTARTS AGE multi 2/2 Running 1 15m nginx-app-7qpbv 1/1 Running 0 1h nginx-app2-tgqqf 1/1 Running 0 1h nginx-hk-app 1/1 Running 0 1h nginx-hk-app2 0/1 Pending 0 32s [root@host01-4 hk]# k get no NAME STATUS ROLES AGE VERSION host01-2.cloud.com Ready <none> 3h v1.10.3 host01-3.cloud.com Ready <none> 3h v1.10.3 host01-4.cloud.com Ready master 3h v1.10.3 [root@host01-4 hk]# k label no host01-2.cloud.com gpu=true node "host01-2.cloud.com" labeled [root@host01-4 hk]# k get no --show-labels NAME STATUS ROLES AGE VERSION LABELS host01-2.cloud.com Ready <none> 3h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,gpu=true,kubernetes.io/hostname=host01-2.cloud.com host01-3.cloud.com Ready <none> 3h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/hostname=host01-3.cloud.com host01-4.cloud.com Ready master 3h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/hostname=host01-4.cloud.com,node-role.kubernetes.io/master= [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE multi 2/2 Running 1 17m 10.44.0.2 host01-2.cloud.com nginx-app-7qpbv 1/1 Running 0 1h 10.36.0.1 host01-3.cloud.com nginx-app2-tgqqf 1/1 Running 0 1h 10.36.0.2 host01-3.cloud.com nginx-hk-app 1/1 Running 0 1h 10.44.0.1 host01-2.cloud.com nginx-hk-app2 1/1 Running 0 3m 10.44.0.3 host01-2.cloud.com | cs |
* 전체 삭제
1 2 3 4 5 6 7 | [root@host01-4 hk]# k delete all --all pod "multi" deleted service "kubernetes" deleted [root@host01-4 hk]# k get rc No resources found. [root@host01-4 hk]# | cs |
*create rc(2replicas) + describe
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 | apiVersion: v1 kind: ReplicationController metadata: labels: name: simple-rc spec: replicas: 2 template: metadata: labels: type: test spec: containers: - image: reg.cloud.com/nginx name: nginx-app ports: - containerPort: 80 protocol: TCP [root@host01-4 hk]# k create -f rc.yaml replicationcontroller "simple-rc" created [root@host01-4 hk]# get rc -bash: get: command not found [root@host01-4 hk]# k get rc NAME DESIRED CURRENT READY AGE simple-rc 2 2 2 5s [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE simple-rc-26m58 1/1 Running 0 21s 10.44.0.1 host01-2.cloud.com simple-rc-7htdc 1/1 Running 0 21s 10.36.0.1 host01-3.cloud.com [root@host01-4 hk]# vi rc.yaml [root@host01-4 hk]# k describe rc simple-rc Name: simple-rc Namespace: default Selector: type=test Labels: type=test Annotations: <none> Replicas: 2 current / 2 desired Pods Status: 2 Running / 0 Waiting / 0 Succeeded / 0 Failed Pod Template: Labels: type=test Containers: nginx-app: Image: reg.cloud.com/nginx Port: 80/TCP Host Port: 0/TCP Environment: <none> Mounts: <none> Volumes: <none> Events: Type Reason Age From Message ---- ------ ---- ---- ------- Normal SuccessfulCreate 1m replication-controller Created pod: simple-rc-7htdc Normal SuccessfulCreate 1m replication-controller Created pod: simple-rc-26m58 [root@host01-4 hk]# | cs |
*Label 조회
1 2 3 4 5 6 | [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS simple-rc-26m58 1/1 Running 0 2m type=test simple-rc-7htdc 1/1 Running 0 2m type=test [root@host01-4 hk]# | cs |
*Label 값 수정 :
1 2 3 4 5 6 7 8 | root@host01-4 hk]# k label po simple-rc-26m58 loc=se pod "simple-rc-26m58" labeled [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS simple-rc-26m58 1/1 Running 0 3m loc=se,type=test simple-rc-7htdc 1/1 Running 0 3m type=test [root@host01-4 hk]# | cs |
*Label overwrite(하나의 pod가 추가 생성된다)
1 2 3 4 5 6 7 8 9 10 11 12 13 | [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS simple-rc-26m58 1/1 Running 0 4m loc=se,type=test simple-rc-7htdc 1/1 Running 0 4m type=test [root@host01-4 hk]# k label po simple-rc-26m58 type=op --overwrite pod "simple-rc-26m58" labeled [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS simple-rc-26m58 1/1 Running 0 5m loc=se,type=op simple-rc-7htdc 1/1 Running 0 5m type=test simple-rc-bts7r 1/1 Running 0 3s type=test [root@host01-4 hk]# | cs |
*RC replicas Scale out :
1 2 3 | [root@host01-4 hk]# k scale rc simple-rc --replicas=3 replicationcontroller "simple-rc" scaled | cs |
*RC를 yaml 파일로 띄워서 수정(replicas부분을 2로 변경 시 바로 2개로 수정되어 뜬다- 1가 죽음)
1 2 3 4 5 6 7 8 9 | [root@host01-4 hk]# k edit rc simple-rc replicationcontroller "simple-rc" edited [root@host01-4 hk]# k get po --show-labels NAME READY STATUS RESTARTS AGE LABELS simple-rc-26m58 1/1 Running 0 8m loc=se,type=op simple-rc-7htdc 1/1 Running 0 8m type=test simple-rc-9cmhk 0/1 Terminating 0 51s type=test simple-rc-bts7r 1/1 Running 0 3m type=test | cs |
*disk를 ssd로 해서 띄우기
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 | apiVersion: apps/v1beta2 kind: DaemonSet metadata: name: simple-ds spec: selector: matchLabels: type: app #selector는 template.labels.type과 template: metadata: labels: type: app spec: nodeSelector: disk: ssd # 추가됨 containers: - image: reg.cloud.com/nginx name: nginx-app ports: - containerPort: 80 protocol: TCP [root@host01-4 hk]# vi rc.yaml [root@host01-4 hk]# k create -f rc.yaml daemonset.apps "simple-ds" created [root@host01-4 hk]# | cs |
*node 조회
1 2 3 4 5 6 | [root@host01-4 hk]# k get no NAME STATUS ROLES AGE VERSION host01-2.cloud.com Ready <none> 4h v1.10.3 host01-3.cloud.com Ready <none> 4h v1.10.3 host01-4.cloud.com Ready master 4h v1.10.3 | cs |
*Labeling 후 pod로 ssd가 생성이 되었는지 확인(disk=ssd)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | [root@host01-4 hk]# k get no NAME STATUS ROLES AGE VERSION host01-2.cloud.com Ready <none> 4h v1.10.3 host01-3.cloud.com Ready <none> 4h v1.10.3 host01-4.cloud.com Ready master 4h v1.10.3 [root@host01-4 hk]# k label no host01-2.cloud.com disk=ssd node "host01-2.cloud.com" labeled [root@host01-4 hk]# get no --show-labels -bash: get: command not found [root@host01-4 hk]# k get no --show-labels NAME STATUS ROLES AGE VERSION LABELS host01-2.cloud.com Ready <none> 4h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,disk=ssd,gpu=true,kubernetes.io/hostname=host01-2.cloud.com host01-3.cloud.com Ready <none> 4h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/hostname=host01-3.cloud.com host01-4.cloud.com Ready master 4h v1.10.3 beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/hostname=host01-4.cloud.com,node-role.kubernetes.io/master= [root@host01-4 hk]# k get ds NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE simple-ds 1 1 1 1 1 disk=ssd 3m [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE simple-ds-vv5lp 1/1 Running 0 1m 10.44.0.3 host01-2.cloud.com simple-rc-26m58 1/1 Running 0 22m 10.44.0.1 host01-2.cloud.com simple-rc-7htdc 1/1 Running 0 22m 10.36.0.1 host01-3.cloud.com simple-rc-bts7r 1/1 Running 0 17m 10.44.0.2 host01-2.cloud.com [root@host01-4 hk]# | cs |
*Labeling을 disk=hhd로 overwrite 할 시 다시 pod 리스트에서 빠져버린다.
1 2 3 4 | [root@host01-4 hk]# k get ds NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE simple-ds 0 0 0 0 0 disk=ssd 8m | cs |
1 2 3 4 5 6 7 8 9 | [root@host01-4 hk]# k label no host01-2.cloud.com disk=hdd --overwrite node "host01-2.cloud.com" labeled [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE simple-rc-26m58 1/1 Running 0 23m 10.44.0.1 host01-2.cloud.com simple-rc-7htdc 1/1 Running 0 23m 10.36.0.1 host01-3.cloud.com simple-rc-bts7r 1/1 Running 0 18m 10.44.0.2 host01-2.cloud.com | cs |
글
*내부 연결에 있어서는 Cluster IP가 Core Service!!!
(External LB > Node Port(Kubernetes) > Cluster IP(Kubernetes))
네임으로 날리면 DNS에서 제어해서
알아서 백단의 VIP로 redirection 시켜준다
1) External LB => Node Port=> DB(NS) => VIP => DB POD
2) External LB => Node Port=> LB(NS) => VIP => WEB POD1, WEB POD2
*kubectl pod Watch 명령어 :
1 2 3 4 5 | [root@host01-4 hk]# k get po -o wide --watch NAME READY STATUS RESTARTS AGE IP NODE simple-rc-bdvs6 0/1 Pending 0 38s <none> <none> simple-rc-d5h74 0/1 Pending 0 38s <none> <none> simple-rc-fq6hl 0/1 Pending 0 38s <none> <none> | cs |
*Service 만들기(Cluster IP = 내부 LB역할을 한다) - Pod IP는 가변적으로 기억할 필요 없음 :
-service 조회
1 2 3 4 5 | [root@host01-4 hk]# k get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 12m test-svc ClusterIP 10.96.25.31 <none> 80/TCP 8m | cs |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 | vi service yalm apiVersion: v1 kind: Service metadata: name: test-svc spec: selector: type: test ports: - port: 80 targetPort: 8080 [root@host01-4 hk]# k create -f service.yaml service "test-svc" created [root@host01-4 hk]# k get service NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 3m test-svc ClusterIP 10.96.25.31 <none> 80/TCP 6s [root@host01-4 hk]# k get po -o wide NAME READY STATUS RESTARTS AGE IP NODE simple-rc-c8bcs 1/1 Running 0 4m 10.36.0.1 host01-3.cloud.com simple-rc-nfssk 1/1 Running 0 4m 10.44.0.1 host01-2.cloud.com simple-rc-zgcjv 1/1 Running 0 4m 10.36.0.2 host01-3.cloud.com [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-zgcjv [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-c8bcs [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-zgcjv | cs |
*Service 이름 만으로 Curl 가능
1 2 3 4 5 6 7 8 9 10 11 12 13 14 | [root@host01-4 hk]# k exec -it simple-rc-c8bcs bash root@simple-rc-c8bcs:/# curl http://test-svc You've hit simple-rc-c8bcs root@simple-rc-c8bcs:/# root@simple-rc-c8bcs:/# cat /etc/resolv.conf nameserver 10.96.0.10 search default.svc.cluster.local svc.cluster.local cluster.local cloud.com options ndots:5 root@simple-rc-c8bcs:/# curl http://test-svc.default You've hit simple-rc-zgcjv root@simple-rc-c8bcs:/# curl http://test-svc.default.svc You've hit simple-rc-zgcjv | cs |
*한번 연결되면 동일 Client는 동일 IP로 가도록 설정(similar to Sticky Session) :
1 2 | sessionAffinity: ClientIP # 부분을 ClientIP로 바꿔주면 Sticky Session으로 변경된다 | cs |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 | [root@host01-4 hk]# k edit svc test-svc service "test-svc" edited # Please edit the object below. Lines beginning with a '#' will be ignored, # and an empty file will abort the edit. If an error occurs while saving this file will be # reopened with the relevant failures. # apiVersion: v1 kind: Service metadata: creationTimestamp: 2018-05-25T06:55:52Z name: test-svc namespace: default resourceVersion: "25644" selfLink: /api/v1/namespaces/default/services/test-svc uid: aa440afa-5fe8-11e8-8e09-005056b28b62 spec: clusterIP: 10.96.25.31 ports: - port: 80 protocol: TCP targetPort: 8080 selector: type: test sessionAffinity: ClientIP # 부분을 ClientIP로 바꿔주면 Sticky Session으로 변경된다 type: ClusterIP status: loadBalancer: {} [root@host01-4 hk]# k get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 15m test-svc ClusterIP 10.96.25.31 <none> 80/TCP 11m [root@host01-4 hk]# k get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE simple-rc-c8bcs 1/1 Running 0 15m 10.36.0.1 host01-3.cloud.com simple-rc-nfssk 1/1 Running 0 15m 10.44.0.1 host01-2.cloud.com simple-rc-zgcjv 1/1 Running 0 15m 10.36.0.2 host01-3.cloud.com [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-nfssk [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-nfssk [root@host01-4 hk]# curl 10.96.25.31 You've hit simple-rc-nfssk [root@host01-4 hk]# | cs |
*Service Endpoint 확인(매핑 정보를 알 수 있다) :
1 2 3 4 5 | [root@host01-4 hk]# k get ep NAME ENDPOINTS AGE kubernetes 10.10.12.14:6443 17m test-svc 10.36.0.1:8080,10.36.0.2:8080,10.44.0.1:8080 13m | cs |
*Endpoint 만들기
endpoint는 service와 이름(label name)이 같아야 한다:
> k get svc
입력 시 뜨는 name과 동일...그 이후에
> k get ep
를 입력 시 endpoint로 ip port가 잘 매핑되어 출력되는 것을 확인 할 수 있다
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 | [root@host01-4 hk]# k get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE headless-svc ClusterIP 10.96.102.169 <none> 80/TCP 2m kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 23m test-svc ClusterIP 10.96.25.31 <none> 80/TCP 19m vi endpoint.yaml apiVersion: v1 kind: Endpoints metadata: name: headless-svc subsets: - addresses: - ip: 10.10.10.10 - ip: 11.11.11.11 ports: - port: 80 #위에서 metadata.name = service name(headless-svc )과 같아야 한다 [root@host01-4 hk]# k create -f endpoint.yaml endpoints "headless-svc" created [root@host01-4 hk]# k get service NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE headless-svc ClusterIP 10.96.102.169 <none> 80/TCP 6m kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 27m test-svc ClusterIP 10.96.25.31 <none> 80/TCP 23m [root@host01-4 hk]# k get ep NAME ENDPOINTS AGE headless-svc 10.10.10.10:80,11.11.11.11:80 19s kubernetes 10.10.12.14:6443 27m test-svc 10.36.0.1:8080,10.36.0.2:8080,10.44.0.1:8080 24m | cs |
'Docker(도커) Kubernetes' 카테고리의 다른 글
| Kubernetes Namespace 생성 및 리소스 사용 (0) | 2018.05.25 |
|---|---|
| Docker Kubernetes, Readiness Probe 생성 방법 (0) | 2018.05.25 |
| Kubernetes 설치 및 사용법, Yaml으로 pod 및 rc 생성 (0) | 2018.05.25 |
| 도커 관련 설정 파일 모음 (2) | 2018.05.24 |
| docker swarm 사용법 (0) | 2018.05.24 |