Sai

@livingdevops Consistency leads to success, congratulations!🎉

Word of mouth is the best marketing there is, and that is how my DevOps Bootcamps make me good money (without any Insta/Google Ads)

What is the role of annotations in Kubernetes YAML file?

Created a website for everything resource related to AI and ML. It's an open-source project, so anyone can contribute it. I think this will be immensely useful for the people who are actively learning and practicing ML/AI concepts. GitHub Repo: github.com/viveknaskar/ev… Live site: viveknaskar.github.io/everything-ai-…

@techxutkarsh Please share

A senior Google engineer just dropped a 421-page doc called Agentic Design Patterns. Every chapter is code-backed and covers the frontier of AI systems: → Prompt chaining, routing, memory → MCP & multi-agent coordination → Guardrails, reasoning, planning This isn’t a blog post. It’s a curriculum. And it’s free.

@uday_devops Congratulations buddy 💫

We just hit 3,500 followers! 2.5 months of grinding from 0 to 3,500. Thank you everyone for your support ❤️🥳

Best 53 seconds on the internet.

I recently asked: “What would you do if your VPC is running out of IPs?” A lot of answers were: “switch to IPv6” Lets be clear: IPv6 is NOT a quick fix. Switching to IPv6 is a full transformation 👇 1️⃣Upgrade your network Routers, firewalls, load balancers, VPNs must support IPv6 2️⃣Redesign IP addressing IPv6 is huge, but you still need structure. Plan CIDR blocks (/56, /64) 3️⃣Validate OS & systems Your servers, containers, and nodes must support IPv6 4️⃣Enable IPv6 in cloud VPC, subnets, ALB/NLB must support IPv6 5️⃣Update DNS Add AAAA records. No AAAA = no IPv6 traffic 6️⃣Rethink security No NAT in IPv6 - Everything becomes publicly reachable - Rewrite firewall & security rules 7️⃣Fix your applications Update configs, APIs, DB connections 8️⃣Choose a transition strategy Dual-stack (most common) NAT64 / DNS64 IPv6-only (rare) 9️⃣Upgrade observability Logs, metrics, tracing must support IPv6. Many tools still assume IPv4 1️⃣0️⃣Test everything Connectivity, latency, failover, DNS. Expect surprises

@devops_nk Have fun buddy…

Taking a short break 🌴 For the past few months, I’ve been pushing myself hard constantly building, learning, and growing. And honestly I’ve achieved a lot too. Now it’s time to pause for a bit. Going on a 2–3 day vacation with a friend ✌️ Don’t worry content is already scheduled for your learning & entertainment. Will be back soon. Thanks for all the insane love & support ❤️

@devops_nk Jfrog

Which artifact repository are you using to store the artifacts in your project ?

Built a private EKS cluster setup on AWS with a clean network design — and learned a lot comparing it with Azure 👇 🔹 Architecture highlights: 🧱 Separate Access VPC (Hub-like) with Bastion host ☸️ EKS VPC (Spoke) with private subnets only 🔐 Private API endpoint (no public exposure) 🔗 VPC Peering for secure communication 🌐 Controlled access via security groups + routing 🚀 SSH → Bastion → EKS nodes + kubectl access to API server 💡 Key learnings (AWS vs Azure mindset shift): 1. Networking is more explicit in AWS AWS: You must wire everything manually (routes, peering, SGs) Azure: More “batteries included” (VNet peering + NSGs feel simpler) 👉 AWS gives flexibility, but also more room for mistakes (like my intermittent API timeouts 😅) 2. EKS control plane access is SG-driven AWS: API access controlled via cluster security group Azure (AKS): More abstracted, easier private cluster setup 👉 Missing SG rules = random connectivity issues (learned the hard way) 3. Private endpoint behavior differs AWS EKS: API resolves to multiple private IPs All paths must be reachable → otherwise intermittent failures Azure: Private endpoints feel more predictable out of the box 4. Peering vs Hub-Spoke maturity AWS: VPC Peering = simple but not scalable Transit Gateway = real hub-spoke Azure: Hub-Spoke is more native and common pattern 5 SSH vs SSM vs Bastion AWS: multiple options (Bastion, SSM Session Manager) Azure: Bastion service is more integrated 🔥 Biggest takeaway: AWS gives you low-level control, Azure gives you higher-level abstractions Both are powerful — but AWS forces you to truly understand networking. Would I change anything? 👉 Next step: move to Transit Gateway-based hub-spoke + SSM (no SSH) Github repo can be found here github.com/DashrathMundka… #AWS #EKS #Terraform #CloudArchitecture #DevOps #Kubernetes #Azure #CloudLearning

High-Level Monitoring & Alerting in an EKS Cluster. Save it for quick interview revision and follow for more. How production observability usually works in Kubernetes: Logs pipeline: 🟢 EKS Application Pods → Fluent Bit (log collector) → Elasticsearch/Kibana (cluster logs) → Splunk (application logs) Metrics pipeline: 🔵 EKS Application Pods → Prometheus (metrics scraping) → Grafana (dashboards & visualization) Alerting: ⚫️ Prometheus → Alertmanager → Slack / Email alerts Simple way to remember: Fluent Bit → Logs Prometheus → Metrics Grafana → Visualization Alertmanager → Notifications This is one of the most common monitoring setups used in production EKS/Kubernetes clusters.

@devops_nk @kbkthebolt What do you do in Crypto?

@kbkthebolt There are different ways bro Crypto Freelancing X monitization

Once you start earning online, there is no going back.

Amazon is introducing account regional namespaces for Amazon S3 general-purpose buckets. For 18 years, Amazon S3 had one frustrating rule: Your bucket name had to be globally unique across the entire internet. Which meant: Create a bucket… ❌ Name already taken. Try another one… ❌ Still taken. Try again… ❌ Also taken. Today, that finally changes. AWS has introduced account regional namespaces for Amazon S3 general purpose buckets. Now bucket names can exist within your account and region namespace, instead of competing globally. Example: logs-123456789012-us-east-1-an What this means for cloud engineers: 🚀 No more fighting for globally unique bucket names 🚀 Easier naming conventions across environments 🚀 Better automation for Terraform / IaC pipelines 🚀 Simpler governance using IAM and SCP policies If you want to read the full AWS announcement: aws.amazon.com/blogs/aws/intr… It’s a small change on the surface… but a big improvement for teams managing multi-account AWS environments.

This week I’m diving deep into MCP (Model Context Protocol). You’ll hear MCP a lot in the AI world this year. It’s the protocol that lets AI models interact with: • APIs • Databases • Developer tools • Infrastructure • Internal systems In simple terms: MCP = a standard way for AI to use tools. Over the next 7 days, I’ll go from zero → building real MCP systems, and I’ll share everything I learn. Plan: Day 1 — MCP fundamentals Day 2 — Build an MCP server Day 3 — Connect MCP to APIs Day 4 — MCP resources & context Day 5 — Designing tools for LLMs Day 6 — Build an AI DevOps assistant Day 7 — MCP in production I’ll post 2 updates daily so we can learn together. If you're a: • Backend engineer • DevOps engineer • AI builder You’ll want to understand MCP.

Launching The AI SRE Watchlist, built on top of awesome-ai-sre open source repo. Check it out here aisrewatchlist.vercel.app And leave a start on the repo github.com/pavangudiwada/…

I stepped away from AI SRE for a few months...coming back, I realized the tech has moved miles ahead but most companies are still debating RAG and MCP. Best teams are already crossing The Trust Barrier!! Where is your team?

Most people use containers. Very few actually understand what happens from build → runtime 👇 Containerization explained: Step 1: Build Phase - It starts with a Dockerfile. When you run: docker build It creates a Docker Image that contains: • Your application code • Dependencies • Required libraries • Runtime environment That image is portable. Same image → Runs on your laptop, CI server, staging, production, cloud. This is where “Build once, run anywhere” becomes real. Step 2: Runtime Phase When you run the image: docker run It becomes a Container. A container is: • An isolated process • With its own filesystem • Own network stack • Own process space But here’s the key 👇 All containers share the Host OS kernel. That’s why they are: - Lightweight like processes - Isolated like VMs Who Manages All This? The Container Engine: • Docker • containerd • CRI-O • Podman It handles: • Container lifecycle • Networking • Isolation • Resource allocation That’s the real flow: Dockerfile → Image → Container → Managed by Engine → Runs on Host Kernel - Simple concept. - Powerful impact. When deploying apps, do you prefer Docker, containerd, or Podman and why? #DevOps #Docker

If Kubernetes networking confused you in the beginning. you are not alone. Imagine your Kubernetes cluster like a restaurant 🍽️ 👥 External Traffic → Customers walking in 🚪 Ingress → The security guard at the door 🧑‍💼 Service → The receptionist directing customers 👨‍🍳 Pod → The chef actually cooking the food 📦 Container → The recipe the chef follows Request Flow: External Traffic → Ingress → Service → Pod → Container • Ingress decides who can enter the cluster • Service decides which Pod should handle the request • Pod/Container actually runs the application Simple concept but debugging Kubernetes networking still feels like this: “Why is my service not reachable?” “Why is the pod healthy but traffic not reaching it?” “Is it DNS… again?” Curious: What confused you the most when you first learned Kubernetes networking? • Ingress • Services • ClusterIP / NodePort / LoadBalancer • DNS inside the cluster Let me know so I can make the next post on that Happy Learning !

Essential Terraform commands every DevOps and Cloud engineer should master 🚀 From `terraform init` to `terraform apply`, these commands form the backbone of Infrastructure as Code. If you work in DevOps & Cloud, this is your daily toolkit. Save this post for work 👇