global/overlay/etc/puppet/cosmos-rules.yaml

@@ -21,7 +21,6 @@
 
 '^internal-sto4-prod-k8sc-[0-9].rut.sunet.se$':
   rut::infra_ca_rp:
-  rut::controller_nrpe:
   sunet::microk8s::node: 
     channel: 1.31/stable
     drain_reboot_cron: true
@@ -47,7 +46,7 @@
 
 '^internal-sto4-prod-monitor-[0-9].rut.sunet.se$':
   sunet::dockerhost2:
-  rut::rut_mon:
+  rut::controller_checks:
   sunet::naemon_monitor:
     domain: monitor-prod.rut.sunet.se
     naemon_tag: latest

global/overlay/etc/puppet/modules/rut/manifests/controller_checks.pp

@@ -0,0 +1,7 @@
+# Everything is awesome!
+class rut::controller_checks {
+    nagioscfg::service {'check_rut_pods':
+    hostgroup_name => ['internal-sto4-prod-k8sc-0.rut.sunet.se'],
+    check_command  => 'check_nrpe!check_rut_pods',
+    description    => 'Microk8s cluster health',
+}

global/overlay/etc/puppet/modules/rut/manifests/controller_nrpe.pp

@@ -1,17 +0,0 @@
-class rut::controller_nrpe {
-    sunet::nagios::nrpe_command {'check_rut_pods':
-      command_line => '/usr/lib/nagios/plugins/check_rut_pods.sh'
-    }
-
-    file { "/usr/lib/nagios/plugins/check_rut_pods.sh":
-       ensure  => "file",
-       content => template("rut/check_rut_pods.sh.erb"),
-       mode    => '0755',
-    }
-
-    user { 'nagios':
-      ensure     => present,
-      groups     => ['microk8s'],
-      membership => minimum,
-    }
-}

global/overlay/etc/puppet/modules/rut/manifests/rut_mon.pp

@@ -1,7 +0,0 @@
-class rut::rut_mon {
-    nagioscfg::service {'check_rut_pods':
-    host_name      => ['internal-sto4-prod-k8sc-0.rut.sunet.se', 'internal-sto4-prod-k8sc-1.rut.sunet.se', 'internal-sto4-prod-k8sc-2.rut.sunet.se'],
-    check_command  => 'check_nrpe!check_rut_pods',
-    description    => 'Microk8s cluster health',
-    }
-}

global/overlay/etc/puppet/modules/rut/templates/check_rut_pods.sh.erb

@@ -1,69 +0,0 @@
-#!/bin/bash
-# This file is managed by puppet.
-
-STATUS=$(/snap/bin/kubectl get events --all-namespaces -o json)
-
-# number warnings required to make critical status (any warning makes warning and any critical makes critical)
-critical_warning_num_threshold=3
-
-num_warnings=$(echo "$STATUS" | jq '[.items[] | select(.type == "Warning")] | length')
-num_normal=$(echo "$STATUS" | jq '[.items[] | select(.type == "Normal")] | length')
-num_critical=$(echo "$STATUS" | jq '[.items[] | select(.type == "Critical")] | length')
-
-function print_info {
-#	echo "$msg: Criticals: $num_critical", "Warnings: $num_warnings"
-	output="$msg - "
-
-	if [[ $num_critical -gt 0 ]]; then
-	    output+="Criticals: $num_critical "
-	fi
-
-	if [[ $num_warnings -gt 0 ]]; then
-	    [[ $num_critical -gt 0 ]] && output+=", "  # Add a comma if both exist
-	    output+="Warnings: $num_warnings "
-	fi
-
-	echo "$output"
-	if [[ "$num_critical" -gt 0 ]]; then
-	    echo "----------------------------------------"
-	    echo "$STATUS" | jq -r '
-	        .items[] | select(.type == "Critical") |
-	        "Host: " + .source.host +
-	        "\nType: " + .type +
-		"\nPod: " + .involvedObject.name +
-	        "\nMessage: " + .message +
-	        "\n----------------------------------------"
-	    '
-	fi
-
-
-	if [[ "$num_warnings" -gt 0 ]]; then
-	    echo "----------------------------------------"
-	    echo "$STATUS" | jq -r '
-	        .items[] | select(.type == "Warning") |
-	        "Host: " + .source.host +
-	        "\nType: " + .type +
-		"\nPod: " + .involvedObject.name +
-	        "\nMessage: " + .message +
-	        "\n----------------------------------------"
-	    '
-	fi
-
-	echo "run \"kubectl get events --all-namespaces\" on $HOSTNAME to get more info"
-	}
-
-if [[ "$num_critical" -gt 0 || "$num_warnings" -ge "$critical_warning_num_threshold" ]]; then
-	msg="CRITICAL"
-	print_info
-	exit 2
-	fi
-
-if [[ "$num_warnings" -gt 0 ]]; then
-	msg="WARNING"
-	print_info
-	exit 1
-	fi
-
-msg="OK"
-print_info
-exit 0