--- title: Why Is Kubernetes Scaling Up When the Real Problem Is... description: A discussion post raises the issue of traffic skew in Kubernetes clusters, where persistent TCP, HTTP/2, or gRPC connections can pin traffic to a single pod,... canonical: https://daily.dev/posts/why-is-kubernetes-scaling-up-when-the-real-problem-is-traffic-distribution--pxdvb0s4d twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why Is Kubernetes Scaling Up When the Real Problem Is Traffic Distribution? | daily.dev og:description: A discussion post raises the issue of traffic skew in Kubernetes clusters, where persistent TCP, HTTP/2, or gRPC connections can pin traffic to a single pod,... og:url: https://daily.dev/posts/why-is-kubernetes-scaling-up-when-the-real-problem-is-traffic-distribution--pxdvb0s4d og:image: https://api.daily.dev/og/posts/pXDVb0S4d.png og:image:alt: Why Is Kubernetes Scaling Up When the Real Problem Is Traffic Distribution? og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Why Is Kubernetes Scaling Up When the Real Problem Is Traffic Distribution? **[DevOps & Platform Engineering Hub](https://daily.dev/sources/sanjay)** · [@divyansh30](https://daily.dev/divyansh30) · 1 min read · 6 upvotes · 0 comments ## Summary A discussion post raises the issue of traffic skew in Kubernetes clusters, where persistent TCP, HTTP/2, or gRPC connections can pin traffic to a single pod, causing it to run hot while other replicas sit idle. Horizontal Pod Autoscaler may respond by adding more pods, increasing compute cost without resolving the underlying imbalance. The author links to a write-up that reproduced the behavior on EKS and explored fixes ranging from connection recycling to cluster-level and Layer 7 routing, then asks readers how they detect and solve traffic skew in production. ## Content Ever seen a Kubernetes workload where one pod is running hot while several other replicas are barely doing anything? The obvious reaction is usually to add more replicas or let HPA scale the deployment. But what if the problem isn't capacity? With persistent TCP connections, traffic can stay pinned to the same backend pod. That means HPA can keep adding pods while the original hot pod is still handling most of the traffic. Now you're paying for more compute without actually fixing the bottleneck. I'm curious how other DevOps and platform teams handle this: - Have you seen HPA scale out while traffic remained concentrated on one pod? - How do you identify traffic skew in production? - Are you dealing with long-lived TCP, HTTP/2, or gRPC connections? - Do you solve this at the application, cluster, or platform layer? I came across a detailed write-up that reproduced this behavior on EKS and looked at different ways to address it, from connection recycling to cluster-level and Layer 7 routing. [https://www.kubeblogs.com/kubernetes-traffic-skew-why-one-pod-gets-hot-while-others-sit-idle/](https://www.kubeblogs.com/kubernetes-traffic-skew-why-one-pod-gets-hot-while-others-sit-idle/) Have you ever had a “hot pod” problem that turned out to be a traffic-distribution issue rather than a capacity problem? #DevOps #PlatformEngineering #Kubernetes #SRE #CloudInfrastructure --- Tags: [#aws](https://daily.dev/tags/aws), [#kubernetes](https://daily.dev/tags/kubernetes) [View this post on daily.dev](https://daily.dev/posts/why-is-kubernetes-scaling-up-when-the-real-problem-is-traffic-distribution--pxdvb0s4d) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"DiscussionForumPosting","mainEntityOfPage":"https://daily.dev/posts/why-is-kubernetes-scaling-up-when-the-real-problem-is-traffic-distribution--pxdvb0s4d","headline":"Why Is Kubernetes Scaling Up When the Real Problem Is Traffic Distribution?","text":"A discussion post raises the issue of traffic skew in Kubernetes clusters, where persistent TCP, HTTP/2, or gRPC connections can pin traffic to a single pod, causing it to run hot while other replicas sit idle. Horizontal Pod Autoscaler may respond by adding more pods, increasing compute cost without resolving the underlying imbalance. The author links to a write-up that reproduced the behavior on EKS and explored fixes ranging from connection recycling to cluster-level and Layer 7 routing, then asks readers how they detect and solve traffic skew in production.","url":"https://daily.dev/posts/why-is-kubernetes-scaling-up-when-the-real-problem-is-traffic-distribution--pxdvb0s4d","datePublished":"2026-09-01T04:40:54.633Z","dateModified":"2026-09-01T04:41:15.656Z","author":{"@type":"Person","name":"Divyansh","url":"https://daily.dev/divyansh30","image":"https://media.daily.dev/image/upload/s--f9lbkpFv--/f_auto/v1744873189/avatars/avatar_JMQ4qarKTe7zPurSPRrzM","description":"Aspiring engineer with an interest in AI, tech and DevOps :)","interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"EndorseAction"},"userInteractionCount":800}},"image":"https://media.daily.dev/image/upload/s--kdv4Xjru--/f_auto/v1788237655/posts/pXDVb0S4d?_a=BAMAMicg0","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":6},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"isPartOf":{"@type":"WebPage","url":"https://daily.dev/squads/sanjay","name":"DevOps & Platform Engineering Hub"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"DevOps & Platform Engineering Hub","item":"https://daily.dev/squads/sanjay"},{"@type":"ListItem","position":3,"name":"Why Is Kubernetes Scaling Up When the Real Problem Is Traffic Distribution?"}]} ```