Arena is a lightweight client that is used to manage Kubernetes-based machine learning tasks. Arena allows you to streamline data preparation, model development, model training, and model prediction throughout a complete lifecycle of machine learning. This improves the work efficiency of data scientists. Arena is also deeply integrated with the basic services of Alibaba Cloud. It supports GPU sharing and Cloud Paralleled File System (CPFS). Arena can run in deep learning frameworks optimized by Alibaba Cloud. This maximizes the performance and utilization of heterogeneous computing resources provided by Alibaba Cloud.


Step 1: Install ack-arena

Note If you have installed ack-arena on the Cloud-native AI Component Set page, skip this step and go to Step 2. For more information, see Deploy the cloud-native AI component set.
  1. Log on to the ACK console.
  2. In the left-side navigation pane of the ACK console, click Clusters.
  3. On the Clusters page, find the cluster that you want to manage and choose More > Manage System Components in the Actions column.
  4. Find ack-arena and click Install.

Step 2: Configure the Arena client

If you use a dedicated Kubernetes cluster, log on to a master node in the cluster by using SSH and run the arena command. For more information about how to log on to a node by using SSH, see Connect to the master nodes of a dedicated Kubernetes cluster by using SSH.

If you use a managed Kubernetes cluster, you must install the Arena client on your on-premises machine, such as a PC that runs macOS. This is because a managed Kubernetes cluster does not contain master nodes. Before you install the Arena client, make sure that the kubeconfig file is in the $HOME/.kube/config directory. For more information, see Connect to ACK clusters by using kubectl. Then, perform the following steps to install and configure the Arena client:

Note You can run the kubectl get nodes command to check whether the configurations in the kubeconfig file are correct.
  1. Download the Arena client.
  2. Decompress the package.
    • To install the Arena client on Linux, run the following command to decompress the package:
      tar -xvf arena-installer-0.8.6-a2bec8c-linux-amd64.tar.gz
    • To install the Arena client on macOS, run the following command to decompress the package:
      tar -xvf arena-installer-0.8.6-a2bec8c-darwin-amd64.tar.gz
  3. Run the following command to install the Arena client:
    cd arena-installer
    bash  --only-binary
  4. Optional:Install bash-completion.
    The auto completion feature of bash-completion can automatically fill in partially typed commands.
    • Run the following command to install bash-completion on CentOS or Alibaba Cloud Linux 2:
      yum install bash-completion -y
    • Run the following command to install bash-completion on Debian or Ubuntu:
      apt-get install bash-completion
    • Run the following command to install bash-completion on macOS:
      brew install bash-completion@2
  5. Run the following command to add the auto completion feature to the profile file.
    • Linux
      echo "source <(arena completion bash)" >> ~/.bashrc
      chmod u+x ~/.bashrc
    • macOS
      echo "source $(brew --prefix)/etc/profile.d/" >> ~/.bashrc
    Then, you can press Tab in a CLI to automatically complete a partially typed command.

Step 3: Test whether Arena works as expected

You can perform the following steps to check whether Arena works as expected:

  1. Run the following command to query the available GPU resources in the cluster:
    arena top node
    The output shows information about the nodes and GPUs. This indicates that Arena works as expected.
    NAME                        IPADDRESS      ROLE    STATUS  GPU(Total)  GPU(Allocated)
    cn-huhehaote.192.1xx.x.xx7  192.1xx.x.xx7  <none>  ready   8           0
    cn-huhehaote.192.1xx.x.xx8  <none>  ready   8           0
    cn-huhehaote.192.1xx.x.xx9  <none>  ready   8           0
    cn-huhehaote.192.1xx.x.xx0  <none>  ready   8           0
    Allocated/Total GPUs In Cluster:
    0/32 (0%)
  2. Use Arena to submit a training job. The output shows that the job is submitted.
    arena submit tf \
          --name=firstjob \
          --gpus=1 \
          "python /app/"

    Expected output:

    configmap/firstjob-tfjob created
    configmap/firstjob-tfjob labeled created
    INFO[0001] The Job firstjob has been submitted successfully
    INFO[0001] You can run `arena get firstjob --type tfjob` to check the job status
  3. Run the arena list command to query all jobs.
    Expected output:
    firstjob  RUNNING  TFJOB    5s
  4. Run the following command to query the state of the submitted job:
    arena get firstjob

    Expected output:

    NAMESPACE: default
    firstjob  SUCCEEDED  TFJOB    14m  firstjob-chief-0
  5. Run the following command to query the log of the job:
    arena logs --tail=10 firstjob

    Expected output:

    Accuracy at step 910: 0.9694
    Accuracy at step 920: 0.9687
    Accuracy at step 930: 0.9676
    Accuracy at step 940: 0.9678
    Accuracy at step 950: 0.9704
    Accuracy at step 960: 0.9692
    Accuracy at step 970: 0.9721
    Accuracy at step 980: 0.9696
    Accuracy at step 990: 0.9675
    Adding run metadata for 999