Install the ack-cgpu component

Updated at:

Container Service for Kubernetes provides the GPU sharing feature that allows multiple prediction models to share one GPU and supports GPU memory isolation based on the NVIDIA kernel mode driver. This topic describes how to install the ack-cgpu component, which can be used to share GPUs, isolate GPU memory, and query GPU allocation information.

Prerequisites

Limits

Do not set the CPU policy to static for nodes that enable GPU sharing.

The following table describes other limits.

Item

Requirement

Kubernetes

Kubernetes 1.12.6 or later

Operating system

Operating systems supported by ACK except Windows

GPU model

For more information about the GPU model supported by ACK, see GPU-accelerated ECS instance types supported by ACK

Step 1: Add labels to GPU-accelerated nodes

  1. Log on to the ACK console. In the left navigation pane, click Clusters.

  2. On the Clusters page, click the name of your cluster. In the left navigation pane, click Nodes > Nodes.

  3. On the Nodes page, click Manage Labels and Taints in the upper-right corner.

  4. On the Manage Labels and Taints page, select the desired nodes and click Add Label.

  5. In the Add dialog box, enter a Name and a Value for the label, and then click OK.

    Set Name to cgpu and Value to true to enable the GPU sharing feature for the node.

Important

If you need to disable the GPU sharing feature for a node, you can set the Name to cgpu and the Value to false. Deleting the label does not disable the GPU sharing feature.

Step 2: Install the ack-cgpu component on the labeled nodes

  1. On the Clusters page, click the name of your cluster. In the left navigation pane, click Applications > Helm.

  2. In the upper-left corner of the page, click Create. On the Chart tab, select ack-cgpu and follow the on-screen instructions to install the ack-cgpu component.

    On the Helm page, verify that the Status of ack-cgpu is Deployed.