DevOps DevSecOps & SRE

🛠️ DevOps Tool of the Week: AirLLM Running large LLMs locally usually requires massive GPU memory. AirLLM takes a different approach by loading model layers into VRAM only when needed during inference. This enables models as large as 70B parameters to run on GPUs with as little as 4GB VRAM. #AI #LLMOps #DevOps Content Credit: DevOpsCube link to the post- https://www.linkedin.com/posts/llm-mlops-devopstools-share-7471516866088452096-qt_9/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY GitHub Repository: https://github.com/lyogavin/airllm

🚀 _helpers.tpl helps keep Helm charts clean and maintainable. By defining reusable named templates and including them across Deployments, Services, and other resources, teams can centralize labels, annotations, and common logic. Similar to shared libraries in CI/CD, but scoped to a Helm chart. #Helm #Kubernetes #PlatformEngineering #DevOps Content Credit: Sulaiman S (DevOps & SRE Discussions) link: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7468157671863545856/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY Source Article: https://devopscube.com/_helpers-tpl-file-in-helm-charts/

🔐 SSL/TLS termination is more than decrypting HTTPS traffic. Where you terminate TLS—at CloudFront, ALB, NGINX Ingress, or inside a service mesh—directly affects security, observability, routing, compliance, and performance. Choosing between TLS termination, passthrough, or end-to-end mTLS is a key cloud architecture decision. #AWS #Kubernetes #CloudSecurity #mTLS #DevOps Content Credit: Kalijavedu BhanuPrakash Reddy Link : https://www.linkedin.com/posts/kalijavedubhanu_aws-kubernetes-eks-share-7465947569047736320-v_mr/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY

AI agents are moving into production, but they still need observability, governance, and operational controls. kagent brings agent workloads directly to Kubernetes, allowing teams to deploy, monitor, and manage AI agents using existing cloud-native tooling and workflows. An interesting project from the creators of Istio. #AIAgents #Kubernetes #PlatformEngineering #DevOps #AIOps source: https://github.com/kagent-dev/kagent

Kubernetes governance isn't about blocking developers—it's about building trust at scale. Gatekeeper enables Policy as Code, helping teams enforce security, compliance, resource limits, and deployment standards before workloads reach production. Well-designed guardrails reduce risk while allowing developers to move faster with confidence. #Kubernetes #Gatekeeper #OPA #PlatformEngineering #DevSecOps Content Credit: Sagar Ikhankar (DevOps & SRE Discussions) link: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7467129825359708160/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY

🚀 Apache Airflow on Kubernetes Airflow is no longer just a data engineering tool—it's becoming a key component of modern MLOps and AI infrastructure. Understanding DAGs, task orchestration, executors, GitSync, and KubernetesExecutor can help DevOps engineers support scalable ML and GenAI workflows. Airflow 3 is also expanding support for AI, ML, and near real-time data processing. #MLOps #Airflow #Kubernetes #DevOps #AIInfrastructure Content Credit: Aswin Vijayan (DevOps & SRE Discussions) LinkedIn Post: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7464923227123855361/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY Source Article (DevOpsCube): https://devopscube.com/apache-airflow-on-kubernetes/

🚀 AIOps vs MLOps vs DevOps vs LLMOps They share the same foundation: pipelines. The difference is what flows through them. 🔹 DevOps → Code → Build → Test → Deploy 🔹 MLOps → Data → Train → Deploy → Retrain 🔹 AIOps → Telemetry → Detect → Correlate → Auto-remediate 🔹 LLMOps → Prompts → Evaluate → Ground → Monitor The evolution of Ops is clear: from software pipelines to AI-driven systems that can learn, reason, and respond. #DevOps #MLOps #AIOps #LLMOps #AI #CloudComputing Content Credit: Vishakha Sadhwani Post link: https://www.linkedin.com/posts/vsadhwani_aiops-vs-mlops-vs-devops-vs-llmops-same-share-7465441593009790976-h9Ys/?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY

🔐 Managing SSH, Kubernetes, MySQL, and PostgreSQL access across environments can get messy fast. Warpgate is an open-source bastion host that provides transparent access, session recording, audit trails, SSO, and 2FA without requiring custom clients. A useful project for teams looking to improve infrastructure access control and visibility. #DevOps #Kubernetes #Security #OpenSource source: https://github.com/warp-tech/warpgate

🚨 What really happens when a Kubernetes pod crashes? Kubernetes doesn't simply restart the pod and move on. Kubelet detects failures, restart policies take effect, CrashLoopBackOff may be triggered, and controllers work to maintain desired replicas. Understanding the internal flow helps identify the real root cause faster. 🔍 #Kubernetes #DevOps #SRE #CrashLoopBackOff #K8s #Troubleshooting Content Credit: Sai Praneeth (DevOps & SRE Discussions) Original Post: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7465995620583981056?utm_source=social_share_send&utm_medium=android_app&rcm=ACoAAFCO0B4BD6DEmkdKDOMFYP9rusautbTY9lY&utm_campaign=copy_link

🚨 Kubernetes pod eviction doesn't always mean application failure. When nodes face memory, disk, or resource pressure, Kubernetes evicts pods to protect cluster stability. Eviction follows QoS priorities—BestEffort first, Burstable next, Guaranteed last. Understanding node pressure helps teams troubleshoot root causes before outages occur. 🚀 #Kubernetes #DevOps #SRE Content Credit: Sai Praneeth (DevOps & SRE Discussions) Source: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7466731645564514304/

🚨 Kubernetes creating hundreds of pods without any traffic increase? The issue is often not traffic—it’s autoscaling, metrics, readiness failures, deployment changes, or controller misconfigurations. Before deleting pods, identify who created them, what triggered scaling, and what changed recently. Debug first. Scale smarter. 🚀 #Kubernetes #DevOps #CloudComputing Original Content: Singam Swetha (DevOps & SRE Discussions) Source: https://www.linkedin.com/feed/update/urn:li:groupPost:6585254-7465657994886791168/

Posting on behalf of @Cartee1 Hi everyone, thought this might be relevant for a lot of you. I kept running into the same problem with coding agents: Git tracks code, but no one tracks the agent. The agent changes a bunch of files, the session moves on, and later you’re stuck trying to understand what happened, which prompt caused what, and how to roll it back. So we built re_gent: version control for AI agent activity. It’s Apache 2.0, free, open source, and still alpha. We posted it on Show HN yesterday and it started gaining traction pretty fast, so it seems like we’re not the only ones feeling this pain. Would love feedback, issues, and contributors. Repo: https://github.com/regent-vcs/re_gent

https://github.com/KusionStack/karpor Karpor is an open-source Kubernetes visualization and intelligence tool that adds powerful search, insights, and AI assistance for managing resources across multiple clusters.

https://github.com/freelensapp/freelens Freelens is a free, open-source Kubernetes IDE. Basically a community-driven Lens/OpenLens-style desktop app for managing clusters across Linux, macOS, and Windows.