tencent cloud

PrevNext

masukan

cari berdasarkan kata kunci

Recent Pages

Dokumentasi
Tencent Kubernetes Engine
    Dokumentasi
    Download PDF

    Node Pool FAQs

    Terakhir diperbarui:2023-09-26 17:40:16
    Unduh PDF
      This document ‍lists the common questions and issues during the usage of general node pools.

      What is the relationship between a node pool and a scaling group?

      A node pool is a collection of nodes with similar specifications, configurations, and attributes. You can manage the nodes in batch through the TKE console, such as configuring node specifications, labels, taints, scripts and other parameters. You can create node pools with different billing modes (such as pay-as-you-go, spot) in a cluster. The underlying implementation of a node pool relies on Auto Scaling. There are two main concepts:
      A scaling group contains a collection of CVM instances that follow the same policies and are used for the same snario. Attributes of a scaling group include the maximum and minimum numbers of CVM instances and more.
      Launch configuration is the template for automatic creation of CVMs. It specifies the CVM instance type, system disk/data disk types and capacities, key pair, security group, etc.
      Each node pool corresponds to a scaling group, and each scaling group corresponds to a launch configuration. You can check the bound launch configuration and scaling group in the node pool details page in the TKE console.

      Which node pool parameters can I modified?

      Reminder:
      To ensure the proper scaling of node pools, DO NOT modify parameters in the AS console unless they are stated in this document.
      Parameter
      How to modify
      Node pool name
      TKE console > Node pool information
      
      Auto Scaling
      Node range
      Cloud tag, Deletion protection
      Labels, Taints
      TKE console > Node pool information: You can specify whether to apply the modifications to existing nodes.
      Operating system
      TKE console > Node configuration details: The modifications only apply to nodes newly added to the node pool.
      Runtime components
      Secondary models
      TKE console > Node configuration details
      A node pool can contain up to 10 models (including the primary model).
      The list shows only instance types available in the AZ of the node pool subnet.
      If a GPU-model taken as the primary model, , the driver is specified upon node pool creation. The secondary model must also be a GPU model.
      Custom data
      TKE console > Node configuration details: Modifications only apply to nodes newly added to the node pool.
      Note:
      DO NOT directly modify the custom data in the launch configuration of AS, which may cause the failure of adding nodes to the cluster.
      
      Security group
      AS console > Launch configuration details:. The modifications only apply to nodes newly added to the node pool.
      Instance name (node name)
      AS console > Launch configuration > Advanced: The modifications only apply to nodes newly added to the node pool.
      Subnet
      AS console > Scaling group details: Modifications only apply to nodes newly added to the node pool.
      Instance creation policy/Retry policy
      AS console > Scaling groups > Policy information: Modifications take effect from the next auto scaling task.

      Which node pool parameters should not be modified

      Parameter
      Description
      Billing mode
      You cannot modify the billing mode in the TKE console. In addition, we suggest NOT** to modify the instance billing mode of a launch configuration in the AS console.
      Data disk
      The data disk cannot be modified. In addition, it is recommended not to modify the capacity of data disks, and not to add or delete a data disk in the launch configuration of AS. Otherwise, the pending Pods may fail to be scheduled to the newly added nodes.
      Supported network (VPC)
      The VPC cannot be modified. In addition, it is recommended not to modify the supported network on the scaling group details page of AS, otherwise it may cause the failure of node expansion.
      Note:
      Each node pool corresponds to a unique scaling group and launch configuration. The launch configuration cannot be bound to other scaling groups, otherwise the node pool may fail to be deleted.
      It is recommended not to modify other parameters of the scaling group in the AS console
      
      Hubungi Kami

      Hubungi tim penjualan atau penasihat bisnis kami untuk membantu bisnis Anda.

      Hubungi Kami
      Dukungan Teknis

      Buka tiket jika Anda mencari bantuan lebih lanjut. Tiket kami tersedia 7x24.

      Kirim Tiket
      Dukungan Telepon 7x24
      Copyright © 2013-2024 Tencent Cloud. All Rights Reserved.
      Kebijakan PrivasiLegalKebijakan Cookie