Shoaib A.

Don't Miss to save the post - kubernetes new features launches in 2025 and learn the complete k8 here for free https://lnkd.in/drShEQAa Kubernetes introduced several significant new features and enhancements in 2025, especially with the upcoming v1.34 release scheduled for August 27, 2025. Here are the key highlights: Dynamic Resource Allocation (DRA) Graduation:  The DRA framework, which allows more flexible allocation of devices like GPUs or specialized hardware through structured resource claims, is moving to stable. This enables better management and sharing of advanced hardware in clusters. ServiceAccount Tokens for Image Pull Authentication:  Kubernetes is enhancing security by letting kubelets use short-lived, OIDC-compliant ServiceAccount tokens for authenticating image pulls—removing the need for long-lived secrets and aligning with best practices for workload identity. Pod Replacement Policy for Deployments:  A new .spec.podReplacementPolicy (alpha) gives you control over how pods are replaced during Deployment updates or scaling—choose faster rollouts or stricter resource control, addressing needs for clusters with tight resources or slow pod terminations. Production-Ready Tracing in Kubelet and API Server:  Deep tracing powered by OpenTelemetry is reaching stability for both kubelet and API servers. This enables operators to track events from control plane to node, crucial for debugging and performance tuning across distributed environments. Fine-Grained Autoscaling Control:  Horizontal Pod Autoscaler (HPA) now allows per-HPA configurable tolerance, letting you override the default autoscaling tolerance (10%) for individual workloads. This supports more responsive or conservative scaling, tailored to specific needs. KYAML Support (Kubernetes YAML Dialect):  KYAML, a stricter and less ambiguous subset of YAML optimized for Kubernetes, is being introduced. It offers safer syntax, always double-quotes string values, supports comments, allows trailing commas, and aims to reduce YAML’s classic error risks. kubectl will support KYAML output. Traffic Distribution Improvements for Services:  The Service resource adds new traffic distribution options, letting you prefer endpoints in the same zone or node, not just “close” ones. This enhances performance and availability in multi-zone or multi-node setups. Other broader 2025 trends include: AI-driven automation and better multi-cluster management Edge/IoT workflow optimizations Advanced security defaults and developer experience improvements 📍Register for free devops course - https://lnkd.in/drShEQAa [Classes starting from next week] 📍Download the syllabus - https://lnkd.in/daARkKq7 📍Join our DevOps Whatsapp group - https://lnkd.in/gJ4cvV77 Do make sure to share and like the post and follow Praveen Singampalli for more such content

٢٤٤

٢١ تعليق

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

٢٩٣

٣١ تعليق

In Kubernetes, HPA uses a fixed 10% tolerance for both scale-up and scale-down. Meaning, HPA waits for at least 10% change in CPU/memory usage before adding or removing pods For example, If you set 70% as the usage target, Kubernetes would only scale up if it went above 77%, and scale down if it went below 63%. Here is the problem with with fixed 10%? Sometimes you want the system to react quickly when traffic spikes. Other times, you want it to be more cautious to avoid constant scaling up and down. Now in Kubernetes v1.33 (alpha feature - HPAConfigurableTolerance), You can set separate tolerance levels for scaling up and scaling down. I have published a short guide with hands on example in my newsletter. 𝗥𝗲𝗮𝗱 𝗶𝘁 𝗵𝗲𝗿𝗲: https://lnkd.in/gyAJTMXS 10,000+ professionals read my DevOps newsletter I share deep dives and insights on Kubernetes, DevOps, Cloud, MLOps and GitOps ♻️ If you find this useful, Repost with your network! #kubernetes

٢٨٩

٧ تعليق

Senior DevOps Engineer Interview at OpenAI Round 1 – Infra & AI-Scale Foundations 1. How would you design a pipeline for deploying model-serving microservices across multiple regions? 2. Explain how you’d secure secret distribution for fine-tuning pipelines across AWS + Azure. 3. A single shard of Redis (serving embeddings cache) is throttling. How do you redesign under load? 4. How do you auto-scale inference clusters for unpredictable GPT traffic without burning GPU $$ on idle nodes? 5. How do you secure GitHub Actions or Jenkins pipelines that deploy AI workloads without hardcoding secrets? Round 2 – Chaos, RCA & Reliability 1. Latency spikes for a subset of GPT users in APAC. Infra looks healthy. Walk me through your triage. 2. Your inference-serving pods show “200 OK” but model outputs are corrupted. Where do you start? 3. An etcd cluster hit compaction issues at 200k writes/sec. How do you stabilize without downtime? 4. S3 replication silently failed for a model checkpoint. Debug path + prevention? 5. eBPF-based observability shows kernel-level throttling during inference. RCA? Round 3 – Leadership, Culture & Engineering Judgment 1. How do you balance reliability with research velocity in an AI-first org? 2. You’re asked to simulate a full GPU cluster outage. What’s your chaos drill plan? 3. How do you explain infra trade-offs (cost vs latency) to researchers who just want more GPUs? 4. What’s your checklist before calling an AI-serving infra “production-ready”? 💡 TL;DR If you haven’t: 1. Debugged a GPU node leak mid-inference 2. Simulated regional outages during live AI traffic 3. Balanced Redis/Vector DB bottlenecks under GPT-scale 4. Built chaos drills for model-serving clusters Then OpenAI interviews will expose your weak spots. But you can prepare. That’s exactly what we do at InfraThrone: war room drills, RCA simulations, and infra challenges. 🎯 Next Cohort: Nov 1, 2025 Early bird open for first 30 engineers → infrathrone.xyz 📌 Source: insights from engineers working at OpenAI (shared publicly). #DevOps #SRE #Kubernetes #OpenAI #InfraThrone #ChaosEngineering

٦٨

٧ تعليق

🚢 What is Kubernetes? aka K8s – the powerful container orchestration tool that’s revolutionizing how we deploy and manage apps! 💻🌍 🧠 In simple terms: Kubernetes = A traffic controller 🚦 for your containerized apps 🧳 It helps you run your apps anywhere, scale them automatically, and recover quickly from crashes 🔁⚙️ 📦 Key Features: ✅ Auto-scaling ⬆️⬇️ ✅ Self-healing 💉 ✅ Load balancing ⚖️ ✅ Rollouts & rollbacks 🔄 ✅ Platform-agnostic (cloud, on-premise, hybrid) ☁️🏠🌐 🧰 Works perfectly with: 🐳 Docker | ☁️ AWS/GCP/Azure | 📂 Helm | 🔧 CI/CD tools | 🔐 Secrets & ConfigMaps 🔥 If your app is in containers, Kubernetes is the tool that keeps it running — smartly, reliably, and automatically. 💬 “Build it, ship it, scale it — with Kubernetes 🚀” #Kubernetes #K8s #Containers #DevOps #CloudComputing #Docker #CloudNative #Microservices #Helm #KubeCTL #AWS #Azure #GCP #PlatformEngineering #SRE #SiteReliabilityEngineering #CloudEngineer #InfrastructureAsCode #Terraform #CI_CD #TechTrends #DevOpsEngineer #CloudArchitecture #DigitalTransformation #SoftwareEngineering #OpenSource #SelfHealingSystems #CloudNativeApps #HybridCloud #MultiCloud #KubernetesCluster #KubeCon #Automation #Orchestration #ITInfrastructure #DeveloperTools #SysAdmin #PlatformOps #EngineeringCulture #AgileDevelopment #BackendEngineering #DevSecOps #ScalingApps #KubernetesMaster #CloudDeployment #KubernetesPods #DockerKubernetes #Serverless #IaC #CloudSolutions #CloudStrategy #KubernetesDeployment #AppScaling #CloudMonitoring #Prometheus #Grafana #Observability #KubernetesCommunity #DevLife #TechTalk #TechLeadership #TechTools #ContainerizedApps #CloudSecurity #ModernApps #DigitalOps #CodeNewbie #EngineeringGoals #TechInnovation #ITTrends #KubernetesLove #SRELife #OpenSourceTools #MicroservicesArchitecture #ManagedKubernetes #CNCF #DevOpsCulture #ModernInfrastructure #OpsEngineer #TechCareer #TechLearning #ContainersEverywhere #TechWorld #KubernetesForAll #CloudDevelopers #DevOpsLife #CloudBased #LearningK8s #DevOpsCommunity #CloudEngineerJobs #SoftwareInfra #CI_CD_Pipeline #ResilienceEngineering #CodeToCloud #AgileOps #TechUpSkill #KubernetesTraining #ServerInfra #TechWizards #CloudWarrior #KubernetesForBeginners #CloudNativeJourney #BuildDeployScale

٣٣

#Kubernetes Journey – Day 8: Volumes & Persistent Volumes In Kubernetes, containers are ephemeral – once a Pod is deleted or restarted, its data is lost. That’s where Volumes & Persistent Volumes (PV) come into play. 🔹 Volume – Provides storage to a Pod, ensuring data is available during container restarts inside the same Pod. 🔹 Persistent Volume (PV) – A cluster-wide storage resource that exists independently of Pods. 🔹 Persistent Volume Claim (PVC) – A request made by a Pod to use storage from a PV. ✅ Key Takeaways: Volumes solve the ephemeral nature of containers. PVs abstract the underlying storage (NFS, AWS EBS, GCP PD, Azure Disk, etc.). PVCs let developers request storage without worrying about backend implementation. StorageClasses automate dynamic provisioning of volumes. 💡 Real-World Example: Imagine running a database Pod (MySQL/Postgres) – you cannot afford to lose data every time the Pod restarts. By using PVCs with PVs, you ensure your data persists reliably. #Interview Question for You: Q.What is the difference between Volume, PV, and PVC in Kubernetes? #Kubernetes #DevOps #CloudComputing #Containers #LearningJourney Pod → PVC → PV → Storage

٦٠

٦ تعليق

******Interview Questions****** 𝐊𝐮𝐛𝐞𝐫𝐧𝐞𝐭𝐞𝐬 Q. What is the difference between a Deployment and a StatefulSet in Kubernetes? Q. When should you use a StatefulSet instead of a Deployment? Q. Can you attach a volume to a Deployment? If yes, how is it different from a StatefulSet? Q. What could cause a StatefulSet pod to fail when rescheduled to a different availability zone? Q. How do PV/PVC behave across zones in EKS or Kubernetes in general? Q. What is a DaemonSet and when would you use it? Q. If you want two pods per node (instead of one), what alternatives to DaemonSet can you use? Q. What is a Pod Disruption Budget (PDB) and how is it useful? Q. How do you handle certificate rotation in on-prem Kubernetes clusters? Q. What are the challenges with scheduling pods in a multi-node, multi-AZ setup? Q. How does the Kubernetes scheduler decide where to place pods? Q. What happens when a StatefulSet pod cannot mount its volume after moving to another node? 🔹 𝐓𝐞𝐫𝐫𝐚𝐟𝐨𝐫𝐦 Q. What are common challenges faced while working with Terraform? Q. How do you handle state file management in Terraform? Q. How do you detect and resolve drift in Terraform-managed infrastructure? Q. How do you manage secrets securely in Terraform? Q. Why should you use a remote backend for Terraform? 🔹 𝐀𝐖𝐒 & 𝐍𝐞𝐭𝐰𝐨𝐫𝐤𝐢𝐧𝐠 Q. What are all the possible ways to deploy an Nginx server on AWS? Q. What are the pre-requisites for VPC peering between two VPCs? Q. What problems occur when two VPCs have overlapping CIDR blocks? Q. How can you enable communication between overlapping CIDR VPCs? Q. What is a Transit Gateway and how does it help in VPC communication? Q. How can a jump server be used in overlapping network scenarios? Q. Can you explain transitive routing between VPCs A, B, and C? 🔹 𝐂𝐈/𝐂𝐃 – 𝐆𝐢𝐭𝐇𝐮𝐛 𝐀𝐜𝐭𝐢𝐨𝐧𝐬 Q. What CI/CD tools have you used in your current role? Q. How are you integrating tools like SonarQube, Docker, and Trivy in your pipelines? Q. How do you trigger a GitHub Actions workflow in another repository? Q. What is the purpose of repository_dispatch in GitHub Actions? Q. How would you trigger a CI/CD pipeline in Repo A from changes in Repo B? #AWS #Linux #DevOps

١٦٠

٣ تعليق

🚨 Troubleshooting Unstable Kubernetes Pods: A Practical Guide Troubleshooting Kubernetes workloads can be challenging, especially when symptoms are unclear or logs are incomplete. To streamline the process, I’ve put together a clear, step-by-step activity diagram that covers the most effective diagnostics tactics used in real-world environments. 📌Here’s a practical, step-by-step troubleshooting activity flow I use when debugging pods in production environments. 🔍 10 Steps to Troubleshoot a Pod Efficiently: 1️⃣ Check the logs – kubectl logs <pod_name> 2️⃣ Analyze pod status – kubectl get pod <pod_name> 3️⃣ Describe the pod – kubectl describe pod <pod_name> 4️⃣ Verify the YAML configuration 5️⃣ Check Kubernetes events – kubectl get events 6️⃣ Validate container images & tags 7️⃣ Restart the pod – kubectl rollout restart deployment/<deployment_name> 8️⃣ Review service dependencies 9️⃣ Test network connectivity – kubectl exec, ping, curl 🔟 Inspect CPU/memory – kubectl top pod <pod_name> 💡 Whether you're in dev, staging, or prod — this flow helps streamline your debugging process. hashtag#Kubernetes hashtag#DevOps hashtag#CloudNative hashtag#Troubleshooting hashtag#SRE hashtag#PlatformEngineering hashtag#K8s hashtag#ContainerOps hashtag#InfrastructureAsCode

٩٦

٣ تعليق