If a Container Service for Kubernetes (ACK) cluster contains vGPU-accelerated nodes, you must purchase NVIDIA GRID licenses and set up GRID license servers to enable the nodes to work as normal. This topic describes how to renew the NVIDIA driver license of a vGPU-accelerated node in an ACK cluster.
- An NVIDIA driver license is purchased. For more information, see Purchase an NVIDIA driver license.
- This topic is applicable to dedicated, managed, and professional Kubernetes clusters.
An ACK cluster contains vGPU-accelerated nodes
- Log on to a vGPU-accelerated node and modify the following settings in the /etc/nvidia/gridd.conf file:
Note For more information about how to log on to a vGPU-accelerated node, see Connect to a Linux instance by using password authentication and Connect to a Windows instance by using password authentication.
ServerAddress=<your License Server address> ServerPort=<License Server port>
- Run the following commands to restart the NVIDIA driver:
systemctl daemon-reload systemctl restart nvidia-gridd
- Run the following command and check the output:
The output indicates that the
nvidia-smicommand is successfully run. This means that the NVIDIA driver license of the vGPU-accelerated instance is renewed.