pod
apiVersion: v1 #指定api版本,此值必须在kubectl apiversion中 kind: Pod #指定创建资源的角色/类型 metadata: #资源的元数据/属性 name: web04-pod #资源的名字,在同一个namespace中必须唯一 labels: #设定资源的标签,详情请见http://blog.csdn.net/liyingke112/article/details/77482384 k8s-app: apache version: v1 kubernetes.io/cluster-service: "true" annotations: #自定义注解列表 - name: String #自定义注解名字 spec: #specification of the resource content 指定该资源的内容 restartPolicy: Always #表明该容器一直运行,默认k8s的策略,在此容器退出后,会立即创建一个相同的容器 nodeSelector: #节点选择,先给主机打标签kubectl label nodes kube-node1 zone=node1 zone: node1 containers: - name: web04-pod #容器的名字 image: web:apache #容器使用的镜像地址 imagePullPolicy: Never #三个选择Always、Never、IfNotPresent,每次启动时检查和更新(从registery)images的策略, # Always,每次都检查 # Never,每次都不检查(不管本地是否有) # IfNotPresent,如果本地有就不检查,如果没有就拉取 command: ['sh'] #启动容器的运行命令,将覆盖容器中的Entrypoint,对应Dockefile中的ENTRYPOINT 例如 ["python","main.py","${type}"] args: ["$(str)"] #启动容器的命令参数,对应Dockerfile中CMD参数 env: #指定容器中的环境变量 - name: str #变量的名字 value: "/etc/run.sh" #变量的值 resources: #资源管理,请求请见http://blog.csdn.net/liyingke112/article/details/77452630 #如果用不到可以删掉 requests: #容器运行时,最低资源需求,也就是说最少需要多少资源容器才能正常运行 cpu: 0.1 #CPU资源(核数),两种方式,浮点数或者是整数+m,0.1=100m,最少值为0.001核(1m) memory: 32Mi #内存使用量 limits: #资源限制 cpu: 0.5 memory: 32Mi ports: - containerPort: 80 #容器开发对外的端口 name: httpd #名称 protocol: TCP livenessProbe: #pod内容器健康检查的设置,详情请见http://blog.csdn.net/liyingke112/article/details/77531584 httpGet: #通过httpget检查健康,返回200-399之间,则认为容器正常 path: / #URI地址 port: 80 #host: 127.0.0.1 #主机地址 scheme: HTTP initialDelaySeconds: 180 #表明第一次检测在容器启动后多长时间后开始 timeoutSeconds: 5 #检测的超时时间 periodSeconds: 15 #检查间隔时间 #也可以用这种方法 #exec: 执行命令的方法进行监测,如果其退出码不为0,则认为容器正常 # command: # - cat # - /tmp/health #也可以用这种方法 #tcpSocket: //通过tcpSocket检查健康 # port: number lifecycle: #生命周期管理 postStart: #容器运行之前运行的任务 exec: command: - 'sh' - 'yum upgrade -y' preStop:#容器关闭之前运行的任务 exec: command: ['service httpd stop'] volumeMounts: #详情请见http://blog.csdn.net/liyingke112/article/details/76577520 - name: volume #挂载设备的名字,与volumes[*].name 需要对应 mountPath: /data #挂载到容器的某个路径下 readOnly: True volumes: #定义一组挂载设备 声明这是一个卷 - name: volume #定义一个挂载设备的名字 #meptyDir: {} hostPath: path: /opt #挂载设备类型为hostPath,路径为宿主机下的/opt,这里设备类型支持很多种
job
{ "apiVersion": "batch/v1", "kind": "Job", "metadata": { "name": "${jobName}", "namespace": "dataproc", "labels": { "app": "seldon", "ksonnet.io/component": "antispam" }, "annotations": { "batch_version": "v1", "project_name": "antispam" } }, "spec": { "template": { "spec": { "containers": [ { "name": "antispam", "image": "${image}", "command": ["python", "main.py", "${thres}", "${projectTag}"], "env": [ { "name": "PROCESS_NUM", "value": "1" }, { "name": "DP_DISABLE_HEALTHCHECKS", "value": "xids" } ], "resources": { "limits": { "nvidia.com/gpu": "1" } }, "imagePullPolicy": "Always", "volumeMounts": [ { "mountPath": "/usr/local/nvidia/bin", "name": "bin" }, { "mountPath": "/usr/lib/nvidia", "name": "lib" }, { "mountPath": "/usr/lib/x86_64-linux-gnu/libcuda.so", "name": "libcuda-so" }, { "mountPath": "/usr/lib/x86_64-linux-gnu/libcuda.so.1", "name": "libcuda-so-1" }, { "mountPath": "/images", "name": "source-data-storage" }, { "mountPath": "/data", "name": "result-data-storage" } ] } ], "imagePullSecrets": [ { "name": "regcred-secret" } ], "volumes": [ { "name": "bin", "hostPath": { "path": "/usr/lib/nvidia-384/bin" } }, { "name": "lib", "hostPath": { "path": "/usr/lib/nvidia-384" } }, { "name": "libcuda-so", "hostPath": { "path": "/usr/lib/x86_64-linux-gnu/libcuda.so" } }, { "name": "libcuda-so-1", "hostPath": { "path": "/usr/lib/x86_64-linux-gnu/libcuda.so.1" } }, { "name": "source-data-storage", "flexVolume": { "driver": "fstab/cifs", "fsType": "cifs", "secretRef": { "name": "${cifsSecretRef}" }, "options": { "networkPath": "${inputPath}", "mountOptions": "" } } }, { "name": "result-data-storage", "flexVolume": { "driver": "fstab/cifs", "fsType": "cifs", "secretRef": { "name": "${cifsSecretRef}" }, "options": { "networkPath": "${resultPath}", "mountOptions": "" } } } ], "restartPolicy": "Never" } } } }
secret
{ "apiVersion": "v1", "kind": "Secret", "metadata": { "name": "${cifsSecretRef}", "namespace": "dataproc" }, "type": "fstab/cifs", "data": { "username": "${username}", "password": "${password}" } }