From: Remi Gacogne <remi.gacogne@powerdns.com>
Date: Mon, 3 Feb 2020 10:51:35 +0000 (+0100)
Subject: rec: Add steal and IO wait metrics
X-Git-Tag: auth-4.3.0-beta2~28^2~2
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=2e27f874790a07bff20b0b93c4a62920ce001b27;p=thirdparty%2Fpdns.git

rec: Add steal and IO wait metrics
---

diff --git a/pdns/rec_channel_rec.cc b/pdns/rec_channel_rec.cc
index 5acc418519..2e0e96adb8 100644
--- a/pdns/rec_channel_rec.cc
+++ b/pdns/rec_channel_rec.cc
@@ -1162,6 +1162,11 @@ void registerAllStats()
   addGetStat("user-msec", getUserTimeMsec);
   addGetStat("sys-msec", getSysTimeMsec);
 
+#ifdef __linux__
+  addGetStat("cpu-iowait", boost::bind(getCPUIOWait, string()));
+  addGetStat("cpu-steal", boost::bind(getCPUSteal, string()));
+#endif
+
   for(unsigned int n=0; n < g_numThreads; ++n)
     addGetStat("cpu-msec-thread-"+std::to_string(n), boost::bind(&doGetThreadCPUMsec, n));
 
diff --git a/pdns/recursordist/docs/metrics.rst b/pdns/recursordist/docs/metrics.rst
index de0e6eddd1..52e7f9cc8f 100644
--- a/pdns/recursordist/docs/metrics.rst
+++ b/pdns/recursordist/docs/metrics.rst
@@ -189,6 +189,18 @@ cpu-msec-thread-n
 ^^^^^^^^^^^^^^^^^
 shows the number of milliseconds spent in thread n. Available since 4.1.12.
 
+cpu-iowait
+^^^^^^^^^^
+.. versionadded:: 4.4
+
+Time spent waiting for I/O to complete by the whole system.
+
+cpu-steal
+^^^^^^^^^
+.. versionadded:: 4.4
+
+Stolen time, which is the time spent by the whole system in other operating systems when running in a virtualized environment.
+
 dlg-only-drops
 ^^^^^^^^^^^^^^
 number of records dropped because of :ref:`setting-delegation-only` setting
diff --git a/pdns/recursordist/rec_metrics.hh b/pdns/recursordist/rec_metrics.hh
index 5ad436aa31..3c6d7d7839 100644
--- a/pdns/recursordist/rec_metrics.hh
+++ b/pdns/recursordist/rec_metrics.hh
@@ -441,6 +441,13 @@ private:
     {"udp-sndbuf-errors",
       MetricDefinition(PrometheusMetricType::counter,
         "From /proc/net/snmp SndbufErrors")},
+
+    {"cpu-iowait",
+      MetricDefinition(PrometheusMetricType::counter,
+        "Time spent waiting for I/O to complete by the whole system")},
+    {"cpu-steal",
+      MetricDefinition(PrometheusMetricType::counter,
+        "Stolen time, which is the time spent by the whole system in other operating systems when running in a virtualized environment")},
   };
 };