Linkerd

The official Escape Room Party track (+ lyrics) by DJ Linky 🤩 (Register for the party if you haven't yet ☞ https://lnkd.in/dW2uMdjY) Something’s wrong inside the cluster. Latency rising, logs all flustered. Pods are failing, traffic spiking. Unknown Controller now is striking. Metrics screaming in the night, Tracing packets left and right. Linky says, “We’ve got a clue." Service mesh is watching you. Follow the traces, Read the flow. Something’s breaking, Deep below. We won’t panic, We won’t flip, Stabilize the service mesh. Save Phippy, save the day. Debug the cluster, find the way. Kubernetes in the zone, Mesh is routing, not alone. Save Phippy, lock the vault. Fix the outage, find the fault. Trace it, patch it, ship it fast, Service mesh will hold it fast. Rolling updates, zero trust, Memory safe, no buffer bust. Rust inside the data plane, Latency dropping once again. Multi-cluster, zone-aware, Traffic's shifting everywhere. From the ingress to the pod, Observability is God.   Check the metrics, Read the graph. Is it liveness? Is it a path? Scale it, route it, Fix the flow. Mesh is running, go go go! Save Phippy, save the day, Engineers will lead the way. Packets dancing through the mesh, Production-ready, fresh. Save Phippy, breach the vault, No more errors, no more faults. Kubernetes standing tall, Service mesh just saved us all.   Trace. Route. Scale. Deploy. Trace. Route. Scale. Deploy. Linky’s watching. Cluster steady. Mesh engaged. Production-ready. Save Phippy, save the night. Debug heroes in the light. From the logs to service mesh, We restore, and we refresh. Save Phippy, case complete. Unknown Controller tastes defeat. Vault unlocked, the crowd goes wild, Cluster stable. Chaos exiled. Listen to the track ☞ https://lnkd.in/eRv6MsqM

[PT-BR] 🚨 Requisições que nunca terminam podem derrubar todo um sistema distribuído. Esse tipo de problema acontece mais do que imaginamos em arquiteturas de microservices. Sem timeouts bem definidos, um serviço lento pode consumir conexões, bloquear threads e gerar falhas em cascata. Quando utilizamos um service mesh como o Linkerd, esses controles podem ser aplicados diretamente na camada de rede — sem alterar o código da aplicação. No meu novo artigo, mostro na prática como implementar isso. --- (EN) 🚨 Requests that never end can take down an entire distributed system. This is more common than we think in microservices architectures. Without well-defined timeouts, a slow service can consume connections, block threads, and trigger cascading failures. With a service mesh like Linkerd, these controls can be applied directly at the network level—without changing a single line of application code. In my latest article, I show how to implement this in practice. What’s inside / O que você vai encontrar: ✅ Request vs. Response vs. Idle timeouts. ✅ Configuration using Kubernetes annotations. ✅ Using Gateway API (HTTPRoute) for timeout policies. ✅ Real-time validation with linkerd viz tap. ✅ The "Golden Rule" between timeouts and retries. 👇 Check it out / Leia o artigo completo: [PT-BR]: https://lnkd.in/dAyBmJ5d --- [EN] https://lnkd.in/dAyBmJ5d

(PT-BR) Meu artigo de Deep Dive agora faz parte do site oficial da Linkerd! 🔥 É uma honra ter esse reconhecimento da equipe do projeto. Esse estudo nasceu a partir de desafios reais que enfrentamos no dia a dia como SREs na Jusbrasil, investigando como o componente linkerd-destination traduz o estado dinâmico do Kubernetes em decisões de tráfego para milhares de proxies no mesh. --- (EN) My Deep Dive technical guide is now part of the official Linkerd website! 🔥 It’s an honor to have this recognition from the project team. This study reflects real production challenges we face as SREs at Jusbrasil, exploring how the Linkerd Destination component translates Kubernetes cluster state into routing decisions for the entire service mesh. What’s inside/ O que você vai encontrar ✅ Service Discovery: How Kubernetes Services and EndpointSlices are translated into proxy endpoints. ✅ Policy Distribution: How security and routing policies are propagated to the data plane. ✅ gRPC Streaming Updates: How proxies receive real-time endpoint updates. ✅ Destination Architecture: Understanding the containers inside the Destination pod and their responsibilities. What’s inside / O que você vai encontrar: A special thanks to the team at Buoyant for the support and for fostering such an incredible open-source community! 🚀 👇 Check it out here: https://lnkd.in/eZNwsSve

EMERGENCY KUBE-NAPPING ALERT 🚨 The cloud native world is in crisis. Phippy, the beloved cloud native mascot, has vanished. A cryptic note, signed only by the "Unknown Controller," confirms a Kube-napping and points to a corrupted Kubernetes cluster on the canals of Amsterdam. Phippy's liveness probe is failing, and time is running out. YOUR ASSISTANCE IS REQUIRED Merge Forward is assembling an elite team of Cluster Detectives for an emergency, hands-on investigation: The ultimate Cloud Native Escape Room Party at KubeCon. YOUR MISSION Join Linky the Lobster on March 25th to follow the metrics, read the traces, and debug the deployments using your Kubernetes and Linkerd skills. This is a skills-gated rescue operation AND party! Phippy's only hope is you. THE LIVENESS PROBE LOUNGE Merge Forward is also hosting a dedicated space for inclusive networking, perfect for connecting with fellow cloud native enthusiasts outside the technical deep dive. Secure your spot today: https://lnkd.in/e_Pk5CCh #KubeNapping #SavePhippy #KubeCon #Linkerd #Kubernetes #CloudNative 

Great article by Bezaleel Silva detailing the runtime behavior of one of Linkerd’s core control-plane components.....linkerd-destination. Check it out. #kubernetes #CloudNative #ServiceMesh