From f9b03ddd9d92143645a248ae34956efe0db38f31 Mon Sep 17 00:00:00 2001
From: Andrey Klimentyev <andrey.klimentyev@flant.com>
Date: Tue, 10 Apr 2018 14:54:48 +0300
Subject: [PATCH] kube-prometheus: fixed CPU accounting

Currently, node recording rules feature
an incorrect idle CPU accounting. This
change aims to fix that.
---
 assets/prometheus/rules/node.rules.yaml        | 6 +++---
 manifests/prometheus/prometheus-k8s-rules.yaml | 6 +++---
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/assets/prometheus/rules/node.rules.yaml b/assets/prometheus/rules/node.rules.yaml
index 9a9d599c..e678ca84 100644
--- a/assets/prometheus/rules/node.rules.yaml
+++ b/assets/prometheus/rules/node.rules.yaml
@@ -2,7 +2,7 @@ groups:
 - name: node.rules
   rules:
   - record: instance:node_cpu:rate:sum
-    expr: sum(rate(node_cpu{mode!="idle",mode!="iowait",mode!~"^(?:guest.*)$"}[3m]))
+    expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[3m]))
       BY (instance)
   - record: instance:node_filesystem_usage:sum
     expr: sum((node_filesystem_size{mountpoint="/"} - node_filesystem_free{mountpoint="/"}))
@@ -12,10 +12,10 @@ groups:
   - record: instance:node_network_transmit_bytes:rate:sum
     expr: sum(rate(node_network_transmit_bytes[3m])) BY (instance)
   - record: instance:node_cpu:ratio
-    expr: sum(rate(node_cpu{mode!="idle"}[5m])) WITHOUT (cpu, mode) / ON(instance)
+    expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[5m])) WITHOUT (cpu, mode) / ON(instance)
       GROUP_LEFT() count(sum(node_cpu) BY (instance, cpu)) BY (instance)
   - record: cluster:node_cpu:sum_rate5m
-    expr: sum(rate(node_cpu{mode!="idle"}[5m]))
+    expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[5m]))
   - record: cluster:node_cpu:ratio
     expr: cluster:node_cpu:rate5m / count(sum(node_cpu) BY (instance, cpu))
   - alert: NodeExporterDown
diff --git a/manifests/prometheus/prometheus-k8s-rules.yaml b/manifests/prometheus/prometheus-k8s-rules.yaml
index 403e4383..552aad5e 100644
--- a/manifests/prometheus/prometheus-k8s-rules.yaml
+++ b/manifests/prometheus/prometheus-k8s-rules.yaml
@@ -498,7 +498,7 @@ data:
     - name: node.rules
       rules:
       - record: instance:node_cpu:rate:sum
-        expr: sum(rate(node_cpu{mode!="idle",mode!="iowait",mode!~"^(?:guest.*)$"}[3m]))
+        expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[3m]))
           BY (instance)
       - record: instance:node_filesystem_usage:sum
         expr: sum((node_filesystem_size{mountpoint="/"} - node_filesystem_free{mountpoint="/"}))
@@ -508,10 +508,10 @@ data:
       - record: instance:node_network_transmit_bytes:rate:sum
         expr: sum(rate(node_network_transmit_bytes[3m])) BY (instance)
       - record: instance:node_cpu:ratio
-        expr: sum(rate(node_cpu{mode!="idle"}[5m])) WITHOUT (cpu, mode) / ON(instance)
+        expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[5m])) WITHOUT (cpu, mode) / ON(instance)
           GROUP_LEFT() count(sum(node_cpu) BY (instance, cpu)) BY (instance)
       - record: cluster:node_cpu:sum_rate5m
-        expr: sum(rate(node_cpu{mode!="idle"}[5m]))
+        expr: sum(rate(node_cpu{mode!="idle",mode!="iowait"}[5m]))
       - record: cluster:node_cpu:ratio
         expr: cluster:node_cpu:rate5m / count(sum(node_cpu) BY (instance, cpu))
       - alert: NodeExporterDown