For simplicity, this guide uses a flat cluster configuration where each node acts as both a master (control-plane) and a worker. ProxMox Hardware Configuration When setting up ProxMox for Talos Kubernetes nodes, follow these instructions to configure the virtual machine: Memory: Allocate at least 3 GB of RAM per node to ensure smooth operation of the Kubernetes components. Processors: Set the CPU configuration to 2 cores (1 socket, 2 cores). ...
Adding more nodes to an existing Talos Kubernetes cluster is straightforward. This guide includes optional automation steps that streamline the process, such as editing node configuration files, but these can also be performed manually. Assumptions Before proceeding, ensure the following conditions are met: You have deployed a single-node Talos-based Kubernetes cluster following the instructions in this guide. The additional node will also have the control-plane role. (If this is not the case, additional node configuration editing will be required.) Talos is already installed on the new node, and it is currently in “Maintenance” mode with an IP address assigned. (Optional) You are using GitOps for storing Kubernetes configuration. Steps to Add a Node 1. (Optional) Install yq yq is a YAML processor required for modifying the node configuration. Install the correct version using the commands below, do not use apt install yq: ...
Notes To use image with QEMU support use tag ce4c980550dd2ab1b17bbf2b08801c7eb59418eafe8f279833297925d67c7515 For example: factory.talos.dev/installer/ce4c980550dd2ab1b17bbf2b08801c7eb59418eafe8f279833297925d67c7515:v1.9.3 See Talos Upgrading Guide for more details If you are using Talos image for booting, don’t forget to update it as well Upgrade Process Use talosctl upgrade command to upgrade the nodes: talosctl upgrade --nodes $N2_IP \ --image factory.talos.dev/installer/ce4c980550dd2ab1b17bbf2b08801c7eb59418eafe8f279833297925d67c7515:v1.9.3 Expected Output You’ll see live remote diagnostics on your screen during the upgrade process: # watching nodes: [xxx.xxx.xxx.182] # * xxx.xxx.xxx.182: waiting for actor ID # watching nodes: [xxx.xxx.xxx.182] # * xxx.xxx.xxx.182: task: stopAllPods action: START # watching nodes: [xxx.xxx.xxx.182] # * xxx.xxx.xxx.182: unavailable, retrying... # watching nodes: [xxx.xxx.xxx.182] # * xxx.xxx.xxx.182: stage: RUNNING ready: false unmetCond: [name:"nodeReady" reason:"node \"n1\" status is not available yet"] # watching nodes: [xxx.xxx.xxx.182] # * xxx.xxx.xxx.182: post check passed Post-upgrade Update talosctl and kubectl ...