<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Chandan Kumar · Blog</title><link>https://chanupen.github.io/blog/</link><description>Notes from running production systems: Kubernetes, AWS, CI/CD, observability, incident response and the tooling around them.</description><language>en-IN</language><managingEditor>Chandan Kumar</managingEditor><lastBuildDate>Sat, 24 Oct 2026 00:35:00 +0530</lastBuildDate><atom:link href="https://chanupen.github.io/blog/index.xml" rel="self" type="application/rss+xml"/><item><title>Why your pods drop requests during a rolling deploy</title><link>https://chanupen.github.io/blog/zero-downtime-rolling-deploys/</link><pubDate>Sat, 24 Oct 2026 00:35:00 +0530</pubDate><guid>https://chanupen.github.io/blog/zero-downtime-rolling-deploys/</guid><description>Kubernetes sends SIGTERM and removes the pod from the load balancer at the same time, not one after the other. That race is behind most 502s during deploys. Here is how to close it.</description><category>kubernetes</category><category>aws</category><category>eks</category><category>reliability</category><category>deployments</category></item><item><title>How a pod gets AWS credentials without a single access key</title><link>https://chanupen.github.io/blog/pod-aws-credentials-without-keys/</link><pubDate>Sat, 24 Oct 2026 00:10:00 +0530</pubDate><guid>https://chanupen.github.io/blog/pod-aws-credentials-without-keys/</guid><description>IRSA from the token up: what actually happens when a pod calls S3 or Bedrock, how to debug it when it fails, and when EKS Pod Identity is the better choice.</description><category>aws</category><category>kubernetes</category><category>eks</category><category>iam</category><category>security</category></item><item><title>The first five minutes on a sick Linux box</title><link>https://chanupen.github.io/blog/first-five-minutes-on-a-linux-box/</link><pubDate>Fri, 23 Oct 2026 23:45:00 +0530</pubDate><guid>https://chanupen.github.io/blog/first-five-minutes-on-a-linux-box/</guid><description>The triage I have used for 13+ years, from RHEL servers in a telecom data centre to Kubernetes pods with no shell, and why a container can be OOMKilled while the node has free memory.</description><category>linux</category><category>troubleshooting</category><category>kubernetes</category><category>sre</category></item><item><title>Starting this blog</title><link>https://chanupen.github.io/blog/starting-this-blog/</link><pubDate>Fri, 23 Oct 2026 22:00:00 +0530</pubDate><guid>https://chanupen.github.io/blog/starting-this-blog/</guid><description>Why I am writing in public after 13 years in operations, and what to expect here.</description><category>meta</category></item></channel></rss>