Moved a Core Web Vitals collection setup designed for shared observability infrastructure to my personal blog. Fixed a dashboard left empty by sparse traffic and changed when metrics are collected for ordinary browsers. Along the way I found why some query results went missing, and worked out how much five days of data can support.
Sent the Core Web Vitals measured in an app webview to an observability platform shared by many teams. The webview gave no signal on leaving, and tail sampling, delta-to-cumulative conversion and Tempo queries each dropped values in their own way without raising an error. For every setting that could not be changed, re-measured what was lost in a reproduction environment.
A feature where the server opens a URL the user handed it is, by itself, the textbook setup for SSRF. I ran six whitelist bypasses and the five defensive principles that block them on Node. Strip IPv4-mapped by hand and hex notation gets through, and undici lookup hook is never called when the host is an IP literal.
The "10% error rate" of a link preview server mixes five kinds of failure with different causes. This post works out the four points where runtime choice actually matters, lowers the error rate by fixing User-Agent and encoding handling, and verifies the "P95 under one second" target by working backwards from the cache hit rate with a two-million-run simulation. Along the way it shows that Node built-in TextDecoder turns CP949 extension characters into different characters without raising an error.
Raise traffic 12x and it takes 31.5 seconds for a new pod to receive its first request, 22.5 of them spent before the HPA notices the load. The post also covers the conditions under which the autoscaler stops responding in the five minutes right after a deploy, why scale-down comes down in steps, why memory-based HPA does not fit Node apps, and scaling ahead of time with KEDA.
Build-time transformations, cache key serialization, ResumeDataCache, cacheHandler, and Cache Components - everything created by a single 'use cache' line
The uncomfortable truth benchmarks reveal
What question does vinext really ask?
Hey Edge Middleware, how have you been?
What happens behind a single line of "use server"?
A preview of section 5.2 (Pitfalls of the vm Module) from the upcoming Node.js Deep Dive book.
Please show lots of interest and support!
How to survive when JSON.parse() becomes overwhelming
CVE-2025-55182, CVE-2025-55184, CVE-2025-55183, and the React hidden inside Next.js
A look at V8's generational garbage collection, adjusting heap memory limits, and practical ways to diagnose memory leaks.
Why GitHub Actions scheduled workflows get delayed by dozens of minutes: the structural causes, and free alternatives that actually run on time.
