Keet's Pear Runtime for AI/Crypto Development

Deno’s not hype — it’s the fix for Node’s mess. Read more: https://lnkd.in/giGsQZWh TypeScript support, security by default, and serverless-ready features make it hard to ignore in 2025. #Deno #ModernBackend #TypeScript

Hello Debuggers👋 Over the past few weeks, I’ve been building and publishing open-source npm packages like Amazora Nexus and ChatStorm. One common piece of feedback I received was: “The projects are solid — we need great documentation to adopt them faster.” So today, I’m excited to share something equally important 🚀 📚 Introducing the ChatStorm Documentation Platform A single, unified documentation hub for all my open-source packages — designed to be simple, structured, and developer-friendly. 🔹 What you’ll find: • ✅ Complete setup & usage guides • ✅ API references with examples • ✅ TypeScript-first documentation • ✅ Clear architecture & flow explanations • ✅ One platform for multiple npm packages 📦 Currently Documented Projects: • Amazora Nexus – Amazon SP-API Node.js Client • ChatStorm – Socket.IO Chat Server • ChatStorm Client – React Hook for real-time chat 🌐 Documentation Platform: 👉 https://lnkd.in/g6-EJYYA This platform will continuously grow as I release new packages and improvements. The goal is simple: make integration faster and learning smoother for developers. If you’re into Node.js, React, TypeScript, real-time systems, or API integrations, I’d love your: ⭐ Feedback ⭐ GitHub stars ⭐ Contributions Thanks to everyone who supported the journey so far 🙌 More open-source coming soon. #OpenSource #Documentation #NodeJS #TypeScript #ReactJS #SocketIO #MongoDB #NPM #DeveloperTools #MERN

I accidentally built a self-DDoS machine. Then I fixed it. 📉📈 Integrating high-frequency trading protocols (Rithmic) into a modern Node.js & Kubernetes stack is not for the faint of heart. I encountered a fascinating bug this week: **The Listener Leak.** **The Setup**: I have a resilient `RithmicConnector` service running in our K8s cluster. It's designed to auto-reconnect if the websocket drops. **The Bug**: Every time the connection dropped and reconnected, our code registered a *new* event listener. 1st connect: 1 Listener. 2nd connect: 2 Listeners. 10th connect: 10 Listeners firing simultaneously for every market tick. 😱 I didn't just have a memory leak; I was spamming the broker with duplicate commands, risking an IP ban. **The Fix**: I refactored the architecture to separate the **Connection Lifecycle** from the **Event Registration**. Moved all listeners to `onModuleInit` (Singleton pattern) so they are registered exactly once, regardless of how many times the socket flickers. **The Result**: A stable, protocol-compliant trading engine running smoothly in a Linux container. Sometimes the hardest part of FinTech isn't the algo—it's the plumbing. 🔧 #FinTech #Kubernetes #NodeJS #Rithmic #TradingTech #SoftwareArchitecture #DevOps

Throwing hardware at a software problem is a band-aid, not a strategy. Earlier this week, I shared how a DDoS attack forced us to scale our infrastructure to 100 pods just to keep the lights on. It was expensive (€1,700/mo) and inefficient. We decided to stop brute-forcing it and start engineering it. In Part 2 of my case study, I explain how we re-architected for resilience: - From Local to Shared State: Why Redis was cheaper than CPU. - The "Undici Trap": Discovering Node.js default timeouts were killing our pods. - Vertical Scaling: Why 2 large pods beat 100 small ones. The Result? 📉 Costs dropped from €1,700 to €700 per month. 🛡️ We survived a 6,000 RPS attack without manual intervention. Reliability isn't an accident. It's an architectural feature. Read the full solution here: https://lnkd.in/eaUDU8AQ #SoftwareArchitecture #CostOptimization #DevOps #NodeJS #Efficiency #astro #web #react

Leaders: encourage engineers to spend a little time each week solving or answering fresh technical questions. The topics popping up now include transaction isolation in databases, OpenGL window rendering, cross-platform embedded installs, and SQL injection challenges in tooling. https://bit.ly/49rq6Wu https://owy.mn/4je4oZz 12262025000544Small contributions (a debugging hint, a link to docs, a tested snippet) compound into big team knowledge gains. #ContinuousLearning #TechLeadership

𝗠𝗼𝘃𝗲 𝗳𝗮𝘀𝘁. 𝗦𝘁𝗮𝘆 𝘀𝗲𝗰𝘂𝗿𝗲. 🚀 Backend systems shouldn’t slow innovation. With the right architecture and tools, development can remain lean while client data stays protected. Firebase enables rapid development without compromising on security, scalability, and reliability — when implemented correctly. 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗮𝗻𝗱 𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗮𝗿𝗲𝗻’𝘁 𝘁𝗿𝗮𝗱𝗲-𝗼𝗳𝗳𝘀. 𝗧𝗵𝗲𝘆’𝗿𝗲 𝘁𝗵𝗲 𝘀𝘁𝗮𝗻𝗱𝗮𝗿𝗱. #DeveloperLife #Firebase #BackendDevelopment #WebDevelopment #Programming #SecurityFirst #TechProfessionals

Containers vs Virtual Machines ⚡ 🔹 Virtualization VM 🖥️: Full OS, kernel, apps, like an apartment Container 📦: App + dependencies, shares OS, like a room in a house 🔹 How they run VM: Hypervisor boots a full OS, like powering on a PC Container: Runs as a process using OS features, like opening a browser tab 🔹 Speed & efficiency VM 🐢: Heavy, slow, more resources Container 🚀: Lightweight, instant start, high density 🎯 Takeaway VMs = strong isolation, different OS Containers = fast, scalable, cloud-native #containers #virtualmachines #docker #kubernetes #cloudcomputing #devops #softwarearchitecture

How Real-Time Systems Work (Socket.io) 🚀 Ever wondered how apps update instantly without a page refresh? Traditional HTTP follows a simple flow: request → response → connection closed This approach works well for standard APIs, but it falls short for real-time use cases. Using WebSockets with Socket.io, I’ve worked on implementing persistent, bi-directional communication between client and server to enable real-time functionality. This approach makes features like the following possible in production environments: • Instant chat and notifications • Live order and status updates • Real-time data and price changes Working with Socket.io helped me understand how event-driven architecture, connection lifecycle management, and scalability play a key role in building responsive applications. Real-time isn’t magic — it’s smart engineering. Have you implemented real-time features in your applications? #FullStackDevelopment #SocketIO #WebSockets #NodeJS #SoftwareEngineering #MERNStack

Built a real-time presence system the way production systems actually work “Online / Offline” looks trivial on the surface. Designing it correctly is not. I recently built Lannister, a real-time user presence system, to deeply understand how large-scale products track user activity reliably. Instead of focusing only on implementation, I focused on design decisions: • Why tracking presence per device is more accurate than per user • How to detect silent crashes without relying on disconnect events • How TTLs in Redis can replace fragile cleanup jobs • How to design presence so it works across multiple servers • How to prevent “ghost online” users in distributed systems The system uses: WebSockets for real-time connectivity Redis TTL + expiration events for crash recovery Distributed state instead of server memory Event-driven cleanup rather than polling-heavy logic What I enjoyed most was thinking through failure scenarios: Server crashes Client crashes Network drops Multi-device users Redis unavailability I documented the full reasoning, trade-offs, and architecture in a blog here 👇 🔗 https://lnkd.in/d2hPCZwf This project helped me sharpen how I think about backend systems, scalability, and correctness under failure — the things that matter in real production environments. Slack Discord Instagram Would love to hear: 👉 How would you design presence at scale? 👉 What edge cases have you seen in real-world systems? #BackendEngineering #SystemDesign #DistributedSystems #Redis #WebSockets #ScalableSystems #SoftwareEngineer #Hiring

Buzzword galore incoming!!!! 😄 Let’s start in layman’s terms. Imagine you’re scrolling Instagram and then Ronaldo posts a pic. Suddenly millions (more likely billions) of people want to see the same thing at the same time. If you only had a fixed number of servers, two bad things happen: 1. the app gets very slow or... 2. or it just dies So instead of guessing traffic, modern systems scale automatically. That’s where Kubernetes + autoscaling come in. Now here are some TLDRs for all of the matter 👇 Scaling on CPU only is kinda… dumb in real systems. Users don’t care about CPU. They care about: *how slow the app feels (latency) *whether requests are stuck in queues *whether the app recovers fast after spikes That’s where SLOs come in. An SLO (Service Level Objective) is basically: “I want p95 latency under X ms” “I don’t want backlog to grow uncontrollably” So instead of reacting after things break, you scale to protect those goals. This is where KEDA shines. Keda lets Kubernetes scale workloads based on real signals, not just “CPU went brrr”. Sooooooooooooooooooooooooooooooooo... I put together a small event-driven demo app where: traffic ramps from 1 → 1000 RPS 1. jobs go through a real message queue (NATS JetStream) 2. workers are written in Go (real CPU work, not fake sleep()) 3. metrics are scraped every second (Cos I am NOT testing this for 30 mins lol) 4. KEDA decides how many worker pods should exist I’m posting a short ramp-up video(ramps up from 1 to 1000 RPS) below, and I’ll push the project to GitHub with a proper README + manifests so anyone can reproduce the demo. If nothing else, this was a fun way to really feel how autoscaling behaves under pressure 🙂 #kubernetes #keda #autoscaling #golang #observability #mandatoryTagsLol