k8s
nano /etc/systemd/system/k3s.service ExecStart=/usr/local/bin/k3s server --cluster-init --etcd-expose-metrics sudo systemctl daemon-reload sudo systemctl restart k3s etcdctl --endpoints=https://127.0.0.1:2379 endpoint status --write-out=table 如何安裝 etcdctl curl -L https://github.com/etcd-io/etcd/releases/download/v3.5.7/etcd-v3.5.7-linux-amd64.
k8s
Discourse 是一個現代化的開源論壇軟體，Bitnami 提供的 Docker 鏡像使其在 Kubernetes 或 Docker 環境中的部署更加方便。然而，過程中有一些細節需要特別留意，以下是我的實際使用經驗與解決方法。 1. Persisting 儲存的路徑 Discourse 的 Bitnami 鏡像需要持久化數據來確保論壇的穩定運行，尤其是 PostgreSQL 資料庫的數據存儲。 * Persisting 儲存路徑： Bitnami/Discourse 預設將 PostgreSQL 的數據儲存在 /bitnami/postgresql，這是需要持久化的關鍵目錄。 * 解決方案： 配合 NFS 或其他外部存儲解決方案，可以確保數據的持續性和高可用性。 2. 搭配 NFS Provider 使用 NFS Subdir External Provisioner 是一種高效的方法，能確保 Discourse
k8s
Title: Setting Up ArgoCD with GitHub Integration Introduction ArgoCD is a powerful GitOps tool for managing Kubernetes resources directly from your version control system. Integrating ArgoCD with GitHub enables automated synchronization of your cluster’s state with your Git repositories. In this guide, I’ll walk you through setting up
k8s
ArgoCD is a powerful GitOps tool for managing Kubernetes applications, but accessing its web interface often requires additional configuration. By default, ArgoCD uses a service accessible only within the cluster. To make it accessible externally, you can configure an Ingress controller. In this article, I’ll walk you through how
k8s
Introduction Installing ArgoCD on Kubernetes can sometimes come with unexpected challenges, especially in environments with existing configurations or resource issues. Recently, I faced several issues while deploying ArgoCD, including CrashLoopBackOff errors, namespaces stuck in Terminating status, and connectivity issues. In this article, I’ll walk you through the problems I
k8s
ArgoCD is a popular GitOps continuous delivery tool that simplifies Kubernetes application deployments. In this article, I’ll guide you through the basic steps to install ArgoCD using Helm, configure access for both external and internal environments, and ensure your KUBECONFIG is set correctly. Step 1: Ensure KUBECONFIG is Properly
k8s
Introduction When using Ghost, a modern and powerful content management system, I encountered an issue while trying to implement the ads.txt file. Ghost doesn’t natively support serving static files like ads.txt at the root level without additional configuration, and I needed a solution to comply with Google
zh
建立多個 interfaces ，Google 本身的設定會只能夠從第一張網卡進出。即使有第二張以上的網卡和 external IP，依然是不可通的情況，封包進得去出不來。 這時，可參考文件 [1] 敘述在 VM 輸入以下指令。 1sudo ifconfig NIC IP_ADDRESS netmask NETMASK broadcast IP_ADDRESS mtu 1430 2echo "1 rt1" | sudo tee -a /etc/iproute2/rt_tables 3sudo ip route add GATEWAY src IP_ADDRESS dev NIC table
gcp
GKE ingress of backend service connected with GCS of bucket. Please follow the References [1] to install the Ingress with NGINX controller on Google Kubernetes Engine Noted: Open the 8443 port in firewall Install ingress kubectl apply -f google-storage-buckets.yaml kubectl apply -f ingress-bucket.yaml kubectl apply -f ingress-pod.yaml
gcp
Background: 使用兩個 LB 對應到相同 VM 但是不同的 Port。因需要對應到相同的 VM，所以需要使用 unmanaged instance group 如果使用下列架構 ex-lb-1 -> unmanaged-instance-group-1(vm-1, vm-2) ex-lb-2 -> unmanaged-instance-group-2(vm-1, vm-2) 在建立第二個 LB 時會出現此錯誤訊息：Operation type [patch] failed with message "Validation failed for instance 'projects/PROJECT_ID/zones/ZONE/instances/INSTANCE_NAME'
gcp
有些 VM 會無法解析 private zone 裡的 CNAME ，而有些則可以解析的情況。 Short description 查看文件 [1] 所述，當 VM 使用 metadata server 169.254.169.254 為 NS 並且沒有其他替代 NS 的時候， DNS 會 follow this table. 您可查看第一項 Cloud DNS private zone 的 behavior Cloud DNS systems perform queries for the CNAME answer in Cloud
gcp
您可以查看文件 [1]，有提供兩個文件可供參考。第一個文件 [2] 是 Google 自己使用的網段內，第二個文件 [3] 是全球或地域性的 IP 範圍 for Google Cloud 的資源。另外，我們也發現 Github 上有個 repository [9] 幫忙列出所有 IP range from Google (Cloud & GoogleBot)。 URL 的部分，這邊為您收集以下常見之服務的 Service endpoint。 GCE API： https://compute.googleapis.com [4] GKE API： https://container.googleapis.com [5]