From 580ee5ede725108f5f33c6a60ce605f8acdaeb33 Mon Sep 17 00:00:00 2001
From: Pratik Mankawde <3397372+pratikmankawde@users.noreply.github.com>
Date: Wed, 13 May 2026 14:40:58 +0100
Subject: [PATCH] fix(telemetry): StatsD gauge and io_latency first-sample emit

Two fixes so gauges register in Prometheus (via StatsD) even when their
initial/steady-state value is 0:

1. StatsDGaugeImpl m_dirty: default-init to true so the initial value
   (0) is emitted on the first flush. Previously, gauges whose value
   never changed from 0 were never flushed and never appeared
   downstream.

2. io_latency_sampler firstSample_: new atomic<bool>, init true.
   m_event.notify now fires when either firstSample_ is true (exchanged
   to false) or lastSample >= 10 ms. This guarantees the io_latency
   metric is registered on startup; subsequent sub-10 ms samples are
   still suppressed to avoid flooding.
---
 src/libxrpl/beast/insight/StatsDCollector.cpp | 5 ++++-
 src/xrpld/app/main/Application.cpp            | 8 ++++++--
 2 files changed, 10 insertions(+), 3 deletions(-)

diff --git a/src/libxrpl/beast/insight/StatsDCollector.cpp b/src/libxrpl/beast/insight/StatsDCollector.cpp
index d11e85830f..6116987447 100644
--- a/src/libxrpl/beast/insight/StatsDCollector.cpp
+++ b/src/libxrpl/beast/insight/StatsDCollector.cpp
@@ -166,7 +166,7 @@ private:
     std::string m_name;
     GaugeImpl::value_type m_last_value{0};
     GaugeImpl::value_type m_value{0};
-    bool m_dirty{false};
+    bool m_dirty{true};
 };
 
 //------------------------------------------------------------------------------
@@ -583,6 +583,9 @@ StatsDEventImpl::do_notify(EventImpl::value_type const& value)
 StatsDGaugeImpl::StatsDGaugeImpl(std::string name, std::shared_ptr<StatsDCollectorImp> const& impl)
     : m_impl(impl), m_name(std::move(name))
 {
+    // Start dirty so the initial value (0) is emitted on the first flush.
+    // Without this, gauges whose value never changes from 0 would never
+    // appear in downstream metric stores (e.g. Prometheus via StatsD).
     m_impl->add(*this);
 }
 
diff --git a/src/xrpld/app/main/Application.cpp b/src/xrpld/app/main/Application.cpp
index 29fb2dd0e9..bb43349eb9 100644
--- a/src/xrpld/app/main/Application.cpp
+++ b/src/xrpld/app/main/Application.cpp
@@ -149,6 +149,7 @@ private:
         beast::Journal m_journal;
         beast::io_latency_probe<std::chrono::steady_clock> m_probe;
         std::atomic<std::chrono::milliseconds> lastSample_;
+        std::atomic<bool> firstSample_;
 
     public:
         io_latency_sampler(
@@ -156,7 +157,7 @@ private:
             beast::Journal journal,
             std::chrono::milliseconds interval,
             boost::asio::io_context& ios)
-            : m_event(std::move(ev)), m_journal(journal), m_probe(interval, ios)
+            : m_event(std::move(ev)), m_journal(journal), m_probe(interval, ios), firstSample_(true)
         {
         }
 
@@ -175,7 +176,10 @@ private:
 
             lastSample_ = lastSample;
 
-            if (lastSample >= 10ms)
+            // Always emit the first sample so the metric is registered in
+            // downstream stores (Prometheus via StatsD).  After that, only
+            // report latency >= 10 ms to avoid flooding with sub-ms values.
+            if (firstSample_.exchange(false) || lastSample >= 10ms)
                 m_event.notify(lastSample);
             if (lastSample >= 500ms)
             {