1. 创建 Dataset 1 2 3 4 5 6 7 8 9 10 kubectl apply -f - <<EOF apiVersion: v1 kind: Secret metadata: name: my-s3 type: Opaque stringData: aws.accessKeyId: xxx aws.secretKey: xxx EOF 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 kubectl apply -f - <<EOF apiVersion: data.fluid.io/v1alpha1 kind: Dataset metadata: name: my-s3 spec: mounts: - mountPoint: s3://BUCKET/ name: s3 options: alluxio.underfs.s3.endpoint: ks3-cn-beijing-internal.ksyun.com alluxio.underfs.s3.disable.dns.buckets: "false" encryptOptions: - name: aws.accessKeyId valueFrom: secretKeyRef: name: my-s3 key: aws.accessKeyId - name: aws.secretKey valueFrom: secretKeyRef: name: my-s3 key: aws.secretKey accessModes: - ReadWriteMany EOF 2. 创建 Runtime 1 2 3 4 5 6 7 8 9

1. 分析 Fluid 挂载 NFS 存储 查看 Fuse 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 38 39 40 41 42 kubectl get pod nfs-demo-fuse-f9wg8 -oyaml apiVersion: v1 kind: Pod metadata: generateName: nfs-demo-fuse- spec: containers: - command: - /usr/local/bin/entrypoint.sh env: - name: FLUID_RUNTIME_TYPE value: thin - name: FLUID_RUNTIME_NS value: default - name: FLUID_RUNTIME_NAME value: nfs-demo - name: MOUNT_POINT value: /runtime-mnt/thin/default/nfs-demo/thin-fuse - name: MOUNT_OPTIONS value: ro image: fluidcloudnative/nfs:v0.1 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - sh - -c - umount /runtime-mnt/thin/default/nfs-demo/thin-fuse name: thin-fuse securityContext: privileged: true volumeMounts:

1. 创建 Dataset 1 2 3 4 5 6 7 8 9 10 kubectl apply -f - <<EOF apiVersion: data.fluid.io/v1alpha1 kind: Dataset metadata: name: nfs-demo spec: mounts: - mountPoint: x.x.x.x:/x-x/ name: nfs-demo EOF 2. 创建 Runtime 1 2 3 4 5 6 7 8 9 10 11 12 13 14 kubectl apply -f - <<EOF apiVersion: data.fluid.io/v1alpha1 kind: ThinRuntimeProfile metadata: name: nfs spec: fileSystemType: nfs fuse: image: fluidcloudnative/nfs imageTag: v0.1 imagePullPolicy: IfNotPresent command: - "/usr/local/bin/entrypoint.sh" EOF 1 2 3 4 5 6 7 8 kubectl apply -f - <<EOF apiVersion: data.fluid.io/v1alpha1 kind: ThinRuntime metadata: name: nfs-demo spec: profileName: nfs EOF 3. 创建测试 Pod 1 2 3 4 5 6 7 8 9 10 11 12

1. 让 Ops Copilot 成为 Ops Coilot 在 2023 年 09 月，我写过一版 Ops Copilot，也有文章发出 我在给 Ops 工具写 Copilot 。 实现的效果是这样的： 1 2 3 4 5 6 7 8 9 10 Opscli> 打开浏览器 Open a browser and navigate to 'https://www.google.com'. ↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓ import webbrowser webbrowser.open('https://www.google.com') ↑↑↑↑↑↑↑↑↑↑↑

1. 什么是 MLOps MLOps 是 Machine Learning Operations 的缩写，描述的是围绕模型研发整个生命周期过程的标准化和工程化。 MLOps 包括以下几个关键步骤: 数据管理，数据的存储、访问、清洗、转换 模型开发，算法开发、模型构建 模型训练与调优，使用数据训练模型，调整超参数优化模型，微调模型 模型评