Frank Perrakis

You can code and run your MCP server remotely on Kubernetes. Here's how: 1. Write code which conforms to the MCP standards. What does this mean? it means, it should work with JSON RPC 2.0 and can process standard MCP methods such as tools/list should send the name and description of the tools, tools/call should be able to invoke the tools etc. These standards are defined by Anthropic 2. You can either code the methods for MCP yourself, which can be tedious. Or use MCP implementation with library like FastMCP 3. Run Dockerfile to create a container and save it in ECR 4. Deploy to the Kubernetes cluster such as in Amazon EKS, and expose it using a ALB via service or ingress. ALB and container support streamable HTTP out of the box, and works nicely 5. Invoke your MCP Server using the ALB Url. For Streamable HTTP, you need to initialize a session, and then use the sessionID for subsequent calls to do tool discovery, and tool calls Check out detailed step by step demo, along with related codes, and pros-cons of local vs remote MCP server: https://lnkd.in/eXKENy6R --- Get byte sized system design, behavioral, Gen AI, and other interview and career switch tips in weekly newsletter : https://lnkd.in/eG7XdHmN #genai #ai #systemdesign

293

31 Comments

Deleting a pod ≠ stopping it instantly Here is the hidden shutdown flow Kubernetes follows👇 - The API server sends a deletion event. - Kubernetes executes the PreStop hook if any exists. - kubelet sends a SIGTERM signal to the container's main process (PID 1). - Applications have 30 seconds by default to handle the signal and shut down gracefully. - If the application process does not exit within the grace period, Kubernetes sends a SIGKILL signal to terminate the pod. 𝗡𝗼𝘁𝗲: Kubernetes does not give extra time for the PreStop hook. If your pod has a 30-second grace period, the PreStop hook and your app’s shutdown process must both finish within those same 30 seconds. We covered the following in detail in our newsletter edition. - What are SIGTERM and SIGKILL - How Kubernetes handles SIGTERM and SIGKILL - How applications can handle SIGTERM for graceful shutdown - The role of PreStop hooks and why it matters - The link between SIGTERM and container PID 1 → 𝗥𝗲𝗮𝗱 𝗶𝘁 𝗛𝗲𝗿𝗲: https://lnkd.in/gC-tsrgu Over to you… Have you written any custom logic to handle SIGTERM? #devops #kubernetes

330

3 Comments

Before You Ship to Production… Read This. Too many teams treat Kubernetes like it’s still running in staging. Then comes the late-night pager. Downtime. Blame games. After mentoring dozens of engineers and seeing real outages firsthand, I have documented the 10 Non-Negotiables for Running Kubernetes in Production, no fluff, just hard-earned lessons. From probes to autoscaling, from secrets to security, this is the checklist you never skip. Follow Bavithran M for more DevOps, Kubernetes, and cloud-native insights. 🔁Don't keep this to yourself, share it with your network and help someone improve their Kubernetes resilience! #kubernetes #devops #cloudnative #productionready #platformengineering #careerbytecode CareerByteCode #bavicnative #DevSecOps #Azure #AWS #GCP #Job #Interview #Career #Linkedin #Troubleshooting

284

47 Comments

Learn Kubernetes Weekly 151 is hot off the press, and those are my top picks: ⚙️ Kubernetes Node Stability and Performance: Tuning Kubelet for Better Resource Management Jean-François PUCHEU — This article explains how to configure kubelet resource reservations, eviction thresholds, and graceful shutdown settings to improve Kubernetes node stability and prevent crashes. 📦 Dynamic PVC provisioning on AWS EKS Tomas Sapak — This tutorial demonstrates how to set up dynamic PersistentVolumeClaim (PVC) provisioning on AWS EKS using the AWS EBS CSI driver and Terraform. 🎨 GPU sharing with MPS and OKE Fernando Harris — This tutorial explains how to configure GPU sharing via NVIDIA MPS on Oracle Kubernetes Engine (OKE). It walks through GPU node pool setup, cloud-init customization, device-plugin Helm deploy, and testing shared access with 10 replicas. 🧹 K8s cleaner: K8s cleaner is a controller that identifies, removes, or updates stale/orphaned or unhealthy resources to maintain a clean and efficient Kubernetes cluster. If you prefer to read online (or receive it next week in your inbox), you can do so here: https://lnkd.in/gHm3cDQr

37

𝟯 𝗦𝘁𝗲𝗽𝘀 𝗧𝗼 𝗦𝗲𝘁𝘁𝗶𝗻𝗴 𝗨𝗽 𝗠𝘂𝗹𝘁𝗶-𝗖𝗹𝘂𝗱 𝗧𝗲𝗿𝗿𝗮𝗳𝗼𝗿𝗺: Most Terraform repos 𝘤𝘰𝘭𝘭𝘢𝘱𝘴𝘦 the second you add more than one Cloud. Here’s the 3-part structure that keeps 𝐦𝐮𝐥𝐭𝐢-𝐜𝐥𝐨𝐮𝐝 Terraform clean and scalable 👇🏽 1️⃣ 𝐂𝐥𝐨𝐮𝐝-𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐜𝐨𝐝𝐞 ▶ multi-cloud-terraform/ holds separate folders for 𝐀𝐖𝐒 and 𝐀𝐳𝐮𝐫𝐞. ▶ Each has its own 𝐦𝐚𝐢𝐧.𝐭𝐟, 𝐯𝐚𝐫𝐢𝐚𝐛𝐥𝐞𝐬.𝐭𝐟, and 𝐨𝐮𝐭𝐩𝐮𝐭𝐬.𝐭𝐟. ▶ This keeps the logic for each provider clear and isolated. 2️⃣ 𝐑𝐞𝐮𝐬𝐚𝐛𝐥𝐞 𝐦𝐨𝐝𝐮𝐥𝐞𝐬 ▶ Inside modules/, you define building blocks like: AWS storage Azure compute Shared networking ▶ Write once, use many times. No duplication. 3️⃣ 𝐄𝐧𝐯𝐢𝐫𝐨𝐧𝐦𝐞𝐧𝐭 𝐜𝐨𝐧𝐟𝐢𝐠𝐬 ▶ envs/ contains dev, test, and prod. ▶ Each environment has its own backend, variables, and provider files. ▶ Same codebase, different settings, isolated deployments. 𝐖𝐡𝐲 𝐢𝐭 𝐦𝐚𝐭𝐭𝐞𝐫𝐬: ✅ No duplication across clouds ✅ Teams scale without stepping on each other ✅ Environments stay clean and consistent 𝗧𝗲𝗿𝗿𝗮𝗳𝗼𝗿𝗺 𝗶𝘀 𝗽𝗼𝘄𝗲𝗿𝗳𝘂𝗹, 𝗯𝘂𝘁 𝗼𝗻𝗹𝘆 𝗶𝗳 𝘆𝗼𝘂 𝗱𝗲𝘀𝗶𝗴𝗻 𝗳𝗼𝗿 𝗴𝗿𝗼𝘄𝘁𝗵. 𝗛𝗼𝘄 𝗵𝗮𝘃𝗲 𝘆𝗼𝘂 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲𝗱 𝗺𝘂𝗹𝘁𝗶-𝗰𝗹𝗼𝘂𝗱 𝗧𝗲𝗿𝗿𝗮𝗳𝗼𝗿𝗺 𝗶𝗻 𝘆𝗼𝘂𝗿 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀?

231

9 Comments