Serverless助力AI计算：阿里云ACK Serverless/ECI发布GPU容器实例

我会带着你远行 2022-01-18 20:19 300阅读 0赞

ACK Serverless（Serverless Kubernetes）近期基于ECI（弹性容器实例）正式推出GPU容器实例支持，让用户以serverless的方式快速运行AI计算任务，极大降低AI平台运维的负担，显著提升整体计算效率。

AI计算离不开GPU已经是行业共识，然而从零开始搭建GPU集群环境是件相对复杂的任务，包括GPU规格购买、机器准备、驱动安装、容器环境安装等。GPU资源的serverless交付方式，充分的展现了serverless的核心优势，其向用户提供标准化而且“开箱即用”的资源供给能力，用户无需购买机器也无需登录到节点安装GPU驱动，极大降低了AI平台的部署复杂度，让客户关注在AI模型和应用本身而非基础设施的搭建和维护，让使用GPU/CPU资源就如同打开水龙头一样简单方便，同时按需计费的方式让客户按照计算任务进行消费， 避免包年包月带来的高成本和资源浪费。

在ACK Serverless中创建挂载GPU的pod也非常简单，通过annotation指定所需GPU的类型，同时在resource.limits中指定GPU的个数即可（也可指定instance-type）。每个pod独占GPU，暂不支持vGPU，GPU实例的收费与ECS GPU类型收费一致，不产生额外费用，目前阿里云ECI提供如下几种规格的GPU类型：（详情请参考[help.aliyun.com/document\_de…][help.aliyun.com_document_de]）

<table> 
 <thead> 
  <tr> 
   <th>vCPU</th> 
   <th>内存(GiB)</th> 
   <th>GPU类型</th> 
   <th>GPU count</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>2</td> 
   <td>8.0</td> 
   <td>P4</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>4</td> 
   <td>16.0</td> 
   <td>P4</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>8</td> 
   <td>32.0</td> 
   <td>P4</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>16</td> 
   <td>64.0</td> 
   <td>P4</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>32</td> 
   <td>128.0</td> 
   <td>P4</td> 
   <td>2</td> 
  </tr> 
  <tr> 
   <td>56</td> 
   <td>224.0</td> 
   <td>P4</td> 
   <td>4</td> 
  </tr> 
  <tr> 
   <td>8</td> 
   <td>32.0</td> 
   <td>V100</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>32</td> 
   <td>128.0</td> 
   <td>V100</td> 
   <td>4</td> 
  </tr> 
  <tr> 
   <td>64</td> 
   <td>256.0</td> 
   <td>V100</td> 
   <td>8</td> 
  </tr> 
 </tbody> 
</table>

下面让我们通过一个简单的图片识别示例，展示如何在ACK Serverless中快速进行深度学习任务的计算。

### 创建Serverless Kubernetes集群 ###

### 使用tensorflow进行图片识别 ###

对于我们人类此图片的识别是极其简单不过的，然而对于机器而言则不是一件轻松的事情，其中依赖大量数据的输入和模型算法的训练，下面我们将基于已有的tensorflow模型对上个图片进行识别。

> 在这里我们选用了[tensorflow的入门示例][tensorflow]  
> 镜像registry-vpc.cn-hangzhou.aliyuncs.com/ack-serverless/tensorflow是基于tensorflow官方镜像tensorflow/tensorflow:1.13.1-gpu-py3构建，在里面已经下载了示例所需models仓库：[github.com/tensorflow/…][github.com_tensorflow]

在serverless集群控制台基于模版创建或者使用kubectl部署如下yaml文件，pod中指定GPU类型为P4，GPU个数为1。

apiVersion: v1
    kind: Pod
    metadata:
      name: tensorflow
      annotations:
        k8s.aliyun.com/eci-gpu-type : "P4"
    spec:
      containers:
      - image: registry-vpc.cn-hangzhou.aliyuncs.com/ack-serverless/tensorflow
        name: tensorflow
        command:
        - "sh"
        - "-c"
        - "python models/tutorials/image/imagenet/classify_image.py"
        resources:
          limits:
            nvidia.com/gpu: "1"
      restartPolicy: OnFailure复制代码

创建pod等待执行完成，查看pod日志:

# kubectl get pod -a
    NAME         READY     STATUS      RESTARTS   AGE
    tensorflow   0/1       Completed   0          6m
    
    
    # kubectl logs tensorflow
    >> Downloading inception-2015-12-05.WARNING:tensorflow:From models/tutorials/image/imagenet/classify_image.py:141: __init__ (from tensorflow.python.platform.gfile) is deprecated and will be removed in a future version.
    Instructions for updating:
    Use tf.gfile.GFile.
    2019-05-05 09:43:30.591730: W tensorflow/core/framework/op_def_util.cc:355] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
    2019-05-05 09:43:30.806869: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
    2019-05-05 09:43:31.075142: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
    2019-05-05 09:43:31.075725: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x4525ce0 executing computations on platform CUDA. Devices:
    2019-05-05 09:43:31.075785: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla P4, Compute Capability 6.1
    2019-05-05 09:43:31.078667: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2494220000 Hz
    2019-05-05 09:43:31.078953: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x4ad0660 executing computations on platform Host. Devices:
    2019-05-05 09:43:31.078980: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>
    2019-05-05 09:43:31.079294: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties:
    name: Tesla P4 major: 6 minor: 1 memoryClockRate(GHz): 1.1135
    pciBusID: 0000:00:08.0
    totalMemory: 7.43GiB freeMemory: 7.31GiB
    2019-05-05 09:43:31.079327: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0
    2019-05-05 09:43:31.081074: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:
    2019-05-05 09:43:31.081104: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0
    2019-05-05 09:43:31.081116: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N
    2019-05-05 09:43:31.081379: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 7116 MB memory) -> physical GPU (device: 0, name: Tesla P4, pci bus id: 0000:00:08.0, compute capability: 6.1)
    2019-05-05 09:43:32.200163: I tensorflow/stream_executor/dso_loader.cc:152] successfully opened CUDA library libcublas.so.10.0 locally
    >> Downloading inception-2015-12-05.tgz 100.0%
    Successfully downloaded inception-2015-12-05.tgz 88931400 bytes.
    giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca (score = 0.89107)
    indri, indris, Indri indri, Indri brevicaudatus (score = 0.00779)
    lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens (score = 0.00296)
    custard apple (score = 0.00147)
    earthstar (score = 0.00117)复制代码

pod的日志显示模型已经成功检测到图片为panda。可以看到在整个机器学习计算过程中，我们只是运行了一个pod，当pod变成terminated状态后任务完成，没有ecs环境准备，没有购买GPU机器，没有安装Nivida GPU驱动，没有安装docker软件，计算力如同水电一样按需使用。

### 最后 ###

ACK中虚拟节点也同样基于ECI实现了GPU的支持，使用方式与ACK Serverless相同（但需要把pod指定调度到虚拟节点上，或者把pod创建在有virtual-node-affinity-injection=enabled label的namespace中），基于虚拟节点的方式可以更灵活的支持多种深度学习框架，如kubeflow、arena或其他自定义CRD。

示例如下：

本文作者：贤维

[原文链接 ][Link 1]

本文为云栖社区原创内容，未经允许不得转载。

转载于:https://juejin.im/post/5cee2b47f265da1bbf68ff4a

[help.aliyun.com_document_de]: https://link.juejin.im?target=https%3A%2F%2Fhelp.aliyun.com%2Fdocument_detail%2F114581.html
[tensorflow]: https://link.juejin.im?target=https%3A%2F%2Fwww.tensorflow.org%2Ftutorials%2Fimages%2Fimage_recognition
[github.com_tensorflow]: https://link.juejin.im?target=https%3A%2F%2Fgithub.com%2Ftensorflow%2Fmodels
[Link 1]: https://link.juejin.im?target=https%3A%2F%2Fyq.aliyun.com%2Farticles%2F701311%3Futm_content%3Dg_1000060141