TechBadger

DevOps changed my life man 😔, i was in a rut, stuck for good in a ‘decent’ role .. I just took a leap faith and everything turned out as I wanted .

Top resources for DevOps learning: - - KodeKloud - A Cloud Guru - Linux Foundation - TechWorld with Nana - Bret Fisher (Docker/K8s) - DevOps Toolkit - The DevOps Handbook - Kubernetes Docs - Helm Docs - Terraform Docs - AWS Skill Builder - Google Cloud Skills Boost - Microsoft Learn - CNCF Landscape - Katacoda (archived but gold) - Play with Kubernetes - Killercoda - DevOps Exercises (GitHub) - 90DaysOfDevOps - DevOps Cube - Spacelift Blog - Codefresh Blog - Learnk8s - The New Stack - SRE Weekly - Last Week in AWS - DevOps'ish Newsletter - CNCF YouTube - DevOps Directive - That DevOps Guy - ByteByteGo - System Design Primer - GitHub Skills - HashiCorp Learn - Pulumi Docs - ArgoCD Docs - FluxCD Docs - 👀

@twtayaan 🙌🏽😂

Parents be like: We'll just make another kid.

FluxCD vs ArgoCD ⚡️ Here’s the truth: tools are just a means to an end. As a DevOps engineer, what matters isn’t chasing every shiny new too - it’s whether it helps you deliver reliably, scale effectively, and solve real problems. Pick what works. Master it. Ship value. #DevOps #GitOps #PlatformEngineering

@ghumare64 claude

Reply with "claude" for free claude pro invites.

learn kubernetes

All the fundamentals you need to know before jumping into GitHubActions:

Agentic AI Concepts

Question: We have a full GitFlow setup. We fixed a critical bug directly in `master`, but now `develop` and feature branches are missing the fix — causing regressions and merge conflicts. How can we safely propagate production fixes back into ongoing development? Answer: Use Git Backmerge. Git Backmerge automatically merges changes from `master` back into `develop` (or other long-lived branches), ensuring hotfixes made in production are not lost. ✅ Keeps branches aligned ✅ Prevents re-introducing fixed bugs ✅ Reduces painful future merges GitHub setup: Create a workflow that triggers on push to `master` → automatically opens or performs a merge into `develop` using GitHub Actions. Production fixes flow *back* into development --exactly what GitFlow needs.

@livingdevops @elonmusk 🙌🏽

Never knew this day will come when Elon Musk @elonmusk himself will repost my tweet.

You need to learn to Claude Code

@__karnati Alloy - just a wrapper over OpenTelemetry

Monitoring tells you something is wrong. Observability tells you WHY. Most teams have one. 1️⃣Monitoring = tracking known failure modes. You define the metrics. You define the thresholds. You get alerts. It's reactive. You can only catch problems you predicted. 2️⃣Observability = ability to understand any system state from its outputs. Built on 3 pillars: Logs, Metrics, Traces. Logs → what happened (event records) Metrics → how much / how fast Traces → why it happened All three together = a full picture of whats happening in the system ✅Logs: almost everyone has them. ✅Metrics: most teams have them. ❌The gap most teams have: traces. Distributed tracing: rare. And it's the one that shows you WHERE in a multi-service chain, a latency spike is happening. Start with OpenTelemetry. @opentelemetry is the open standard for instrumenting apps for all 3 pillars. Vendor-neutral. Works with Jaeger, Grafana Tempo, Honeycomb, Datadog. Deploy once. Send anywhere. . What's your observability stack? Drop below 👇👇

DevOps Project Idea 💡 Deploy a TO-DO app with Helm and Kustomize and integrate CI/CD checks for security and compliance:

reminder: the most important software in history started as someone's side project.

@twtayaan @devops_nk 😂

How can you forget > Linux is free > Docker is free > Kubernetes is free > Git and Github are free > GitHub Actions is free > Python is free > AWS, GCP, Azure are free (limited use) > Terraform is free > ArgoCD and Flux are free > Prometheus and Grafana are free your laptop and internet connection, that's all you need to start ?

I'm on my way to dislike shit posters whose hooks start with - be honest - mac vs windows - what you will do in this situation - I love my pdf post 100th time - brave browser post 100th time - whoever has built this bla bla.. - and many more...

𝗿𝗺 -𝗿𝗳 /

@__karnati wire a Slack notification for any new “drift” … as soon as a PR is merged and it causes drift that notification fires

In my opinion, the most important file in infrastructure setup is 👇 Your Terraform state file. Terraform state file (.tfstate) is Terraform's map of reality. It tracks which resources exist, what they are, and their IDs. When you run "terraform apply", Terraform compares the desired state in this state file and creates the resources. 🤯If the State file is wrong, your Terraform brain is wrong. A wrong state file can CREATE a resource that already exists, or it might DELETE a resource that is being used. Wrong state = wrong plan = production incidents Common ways a state gets corrupted: → Two engineers run applications simultaneously → Manual changes to cloud resources outside Terraform → Partially apply that failed midway → Someone deleted the state file to start fresh.' The fix: remote state with locking. This should be non-negotiable for any team environment. Drift detection tip: Run "terraform plan" or "terraform refresh" regularly in CI, even when you're not deploying. If the plan shows changes, but you didn't change any .tf files → someone drifted prod manually. Catch it before it causes problems. .......................................... Remote state + locking = the minimum for any team. How are you securing your Terraform state? Drop your setup 👇

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 !

@DThompsonDev hi Danny , is this open to remote speakers?

Since I opened the Call For Speakers on Sunday we have over 50 speaker submissions 🥹 Thank y'all for supporting the idea of CYC26! Just the fact that y'all want to be there with the community means the absolute world to me! Want to speak at CYC26? sessionize.com/the-commit-you…

@iximiuz aha 😂

@techbadger_ Ah, right. That's for building though :) I was ranting about distribution.

How Multi-Platform Container Images Work 🧐 When you "docker pull nginx:1.29" on an AMD64 server and an ARM64 laptop, Docker fetches totally different image builds. But how can it happen when you used exactly the same image name in both cases? To represent multiple builds with a single image reference (e.g., "nginx:1.29"), container registries use a special file - Image Index that lists manifests for individual platform builds. The pull process then gets an extra step: - First, fetch the index from https://registry.example[.]com/v2/REPO/manifests/TAG - Then find a manifest describing the image for the requested platform in the index and fetch it by digest https://registry.example[.]com/v2/REPO/blobs/DIGEST - Finally, fetch the actual image configuration and filesystem layer blobs using the digests from the manifest Whereas for a single-platform image, the https://registry.example[.]com/v2/REPO/manifests/TAG link would point directly to its manifest. Hence, one less layer of indirection. Read more about the internals of container images at labs.iximiuz.com/tutorials/cont…