#!/bin/bash # TLinux 4 (TencentOS Server 4.0) A4X Worker — k8s 1.34 + IMEX channel # Phase 1: IMEX initramfs setup + reboot # Phase 2: k8s 1.34 worker setup + join (requires CP_IP and join token via metadata) set -euo pipefail exec > >(tee /var/log/kubeadm-setup.log) 2>&1 K8S_VERSION="1.34" K8S_PATCH="1.34.1" IMEX_STAMP=/etc/.imex-configured echo "=== [0/0] TLinux 4 base setup ===" dnf install -y sudo cloud-utils-growpart xfsprogs 2>/dev/null || true echo '%google-sudoers ALL=(ALL:ALL) NOPASSWD:ALL' > /etc/sudoers.d/google_sudoers chmod 440 /etc/sudoers.d/google_sudoers systemctl enable --now sshd sed -i 's/^#*PermitRootLogin.*/PermitRootLogin yes/' /etc/ssh/sshd_config systemctl restart sshd echo '* soft memlock unlimited' > /etc/security/limits.d/rdma.conf echo '* hard memlock unlimited' >> /etc/security/limits.d/rdma.conf BOOT_DEV=$(findmnt -n -o SOURCE / | sed 's/p[0-9]*$//') PART_NUM=$(findmnt -n -o SOURCE / | grep -o '[0-9]*$') growpart ${BOOT_DEV} ${PART_NUM} 2>/dev/null || true xfs_growfs / 2>/dev/null || resize2fs $(findmnt -n -o SOURCE /) 2>/dev/null || true nvidia-smi -pm ENABLED 2>/dev/null || true # Phase 1: IMEX initramfs setup (runs once, then reboots) if [ ! -f "$IMEX_STAMP" ]; then echo "=== Phase 1: Configuring IMEX channels ===" echo "options nvidia NVreg_CreateImexChannel0=1" > /etc/modprobe.d/nvidia.conf dracut --force --add-drivers "gve" touch "$IMEX_STAMP" echo "IMEX initramfs configured, rebooting..." reboot exit 0 fi echo "=== Phase 2: kubeadm worker setup (k8s $K8S_VERSION) ===" echo "=== [1/8] Disable swap and SELinux ===" swapoff -a sed -i '/swap/d' /etc/fstab setenforce 0 || true sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config 2>/dev/null || true echo "=== [2/8] Kernel modules ===" cat > /etc/modules-load.d/k8s.conf < /etc/sysctl.d/k8s.conf < /etc/yum.repos.d/docker-ce.repo < /etc/containerd/config.toml sed -i 's/SystemdCgroup = false/SystemdCgroup = true/' /etc/containerd/config.toml systemctl enable --now containerd echo "=== [4/8] Install NVIDIA container toolkit ===" curl -s -L https://nvidia.github.io/libnvidia-container/stable/rpm/nvidia-container-toolkit.repo | \ tee /etc/yum.repos.d/nvidia-container-toolkit.repo dnf install -y nvidia-container-toolkit nvidia-ctk runtime configure --runtime=containerd --set-as-default systemctl restart containerd echo "=== [5/8] Install kubeadm/kubelet/kubectl (k8s $K8S_VERSION) ===" cat > /etc/yum.repos.d/kubernetes.repo <&1 || echo "WARNING: IMEX channels not found" echo "=== [7/8] System info ===" cat /etc/os-release | head -3 uname -r nvidia-smi --query-gpu=driver_version --format=csv,noheader | head -1 kubeadm version echo "=== [8/8] kubeadm join ===" CP_IP=$(curl -sf -H "Metadata-Flavor: Google" \ http://metadata.google.internal/computeMetadata/v1/instance/attributes/cp-ip 2>/dev/null || true) JOIN_TOKEN=$(curl -sf -H "Metadata-Flavor: Google" \ http://metadata.google.internal/computeMetadata/v1/instance/attributes/join-token 2>/dev/null || true) JOIN_HASH=$(curl -sf -H "Metadata-Flavor: Google" \ http://metadata.google.internal/computeMetadata/v1/instance/attributes/join-hash 2>/dev/null || true) NODE_NAME=$(cat /proc/sys/kernel/hostname 2>/dev/null || curl -sf -H "Metadata-Flavor: Google" http://metadata.google.internal/computeMetadata/v1/instance/name || echo "unknown") if [[ -n "$CP_IP" && -n "$JOIN_TOKEN" && -n "$JOIN_HASH" ]]; then kubeadm join ${CP_IP}:6443 \ --token "${JOIN_TOKEN}" \ --discovery-token-ca-cert-hash "sha256:${JOIN_HASH}" \ --node-name "${NODE_NAME}" \ --ignore-preflight-errors=Hostname 2>&1 || echo "kubeadm join failed, manual join required" else echo "Missing CP metadata (cp-ip, join-token, join-hash). Manual join required." echo "Run on CP: kubeadm token create --print-join-command" fi echo "=== Done ===" echo "TLinux 4 k8s 1.34 worker setup complete at $(date)"