From 31b8e82e475cb2ee70e166ae5ea1f8cf2188bcbc Mon Sep 17 00:00:00 2001 From: marcusDenslow Date: Thu, 30 Apr 2026 08:44:56 +0200 Subject: [PATCH] fix: widen histogram buckets to resolve sub-5ms latency MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Default prometheus.DefBuckets bottoms out at 5ms, which made every route's p95 collapse to 4.75ms (histogram_quantile interpolation within the smallest bucket). Custom buckets from 100µs to 5s now distinguish microsecond-fast routes (health) from slower ones (stripe-backed cart endpoints). Co-Authored-By: Claude Opus 4.7 --- api/middleware/metrics.go | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/api/middleware/metrics.go b/api/middleware/metrics.go index ab4ba7b..e22077d 100644 --- a/api/middleware/metrics.go +++ b/api/middleware/metrics.go @@ -24,7 +24,7 @@ var ( prometheus.HistogramOpts{ Name: "http_request_duration_seconds", Help: "HTTP request latency in seconds, labeled by method and route", - Buckets: prometheus.DefBuckets, + Buckets: []float64{.0001, .0005, .001, .0025, .005, .01, .025, .05, .1, .25, .5, 1, 2.5, 5}, }, []string{"method", "route"}, )