Web Performance | WebPixie BlogPosts tagged with Web Performance on the WebPixie Blog.https://webpixie.io/blog/en-gbWhy TTFB alone won't tell you what's actually slowhttps://webpixie.io/blog/post/ttfb-alone-wont-tell-you-whats-slowhttps://webpixie.io/blog/post/ttfb-alone-wont-tell-you-whats-slowA lower TTFB does not guarantee a faster page, and a higher one does not guarantee a slower one. Cloudflare has the data to prove it. Here is what to measure instead.Sat, 19 Sep 2026 17:30:00 GMTWeb PerformanceMonitoringMatt ScottHSTS preload is a one-way street: what removal actually takeshttps://webpixie.io/blog/post/hsts-preload-removal-timelinehttps://webpixie.io/blog/post/hsts-preload-removal-timelineMost guides say removal "takes months." The real number, from the official removal page, is 6 to 12 weeks to reach most Chrome users, and Firefox and Safari run their own separate clocks entirely.Mon, 07 Sep 2026 22:01:00 GMTWeb PerformanceMonitoringMatt Scott504 vs 503: what actually triggers each in nginx, Cloudflare, and ALBhttps://webpixie.io/blog/post/504-vs-503-nginx-cloudflare-albhttps://webpixie.io/blog/post/504-vs-503-nginx-cloudflare-alb503 means a server is deliberately not serving right now. 504 means something upstream never answered in time. The trigger conditions for each are different in nginx, AWS ALB, and Cloudflare, and none of them share the same defaults.Sun, 06 Sep 2026 19:49:00 GMTWeb PerformanceMonitoringMatt ScottThe Cache-Control directives map: no-store, no-cache, private, must-revalidatehttps://webpixie.io/blog/post/cache-control-directives-maphttps://webpixie.io/blog/post/cache-control-directives-mapFive directives, one table, and four ways getting one of them wrong actually breaks a site: a personal page cached on a shared CDN, a logged-out user seeing a previous session on the back button, and two more.Tue, 01 Sep 2026 12:12:00 GMTWeb PerformanceMonitoringMatt Scott