Lite Cluster Resource Management

On the ModelArts console, you can manage created resources. You can click a resource pool name to go to the resource pool details page and perform the following operations:

Managing Lite Cluster Nodes: A node is a fundamental component of a container cluster. You can manage an individual node within a resource pool, including tasks such as replacing, deleting, and resetting the node.
Managing Lite Cluster Node Pools: To help you better manage nodes in a Kubernetes cluster, ModelArts allows you to manage nodes using node pools. A node pool is a group of nodes with the same configuration in a cluster. A node pool contains one or more nodes. You can create, update, and delete node pools.
Managing Lite Cluster Resource Pool Tags: ModelArts enables you to add tags to resource pools. Tags serve as identifiers for cloud resources, allowing you to quickly locate resource pools.
Resizing a Lite Cluster Resource Pool: After a cluster resource pool has been created and utilized for a period of time, the resource requirements may evolve due to changes in AI development services. In such cases, ModelArts offers a scaling function that allows you to dynamically adjust the resources as needed.
Upgrading the Lite Cluster Resource Pool Driver: If nodes in a resource pool include GPU/Ascend resources, you may need to customize the GPU/Ascend driver according to your service requirements. ModelArts enables you to independently upgrade the GPU/Ascend driver for the dedicated resource pool.
Monitoring Lite Cluster Resources: ModelArts leverages AOM and Prometheus to monitor resources, providing insights into current resource usage.
Releasing Lite Cluster Resources: You can release Lite Cluster resources that are no longer used.