Merge branch 'master' into dist/eole/2.6.2/master

diagnose/70-grafana

@@ -1,6 +1,7 @@
 #!/bin/bash
 
-if [ $(CreoleGet activer_grafana) = "oui" ];then
+if [[ $(CreoleGet activer_grafana) == "oui" ]]
+then
     . /usr/lib/eole/diagnose.sh
 
     EchoGras "*** Accès au serveur grafana"

diagnose/70-prometheus

@@ -1,6 +1,7 @@
 #!/bin/bash
 
-if [ $(CreoleGet activer_prometheus) = "oui" ];then
+if [[ $(CreoleGet activer_prometheus) == "oui" ]]
+then
     . /usr/lib/eole/diagnose.sh
 
     EchoGras "*** Accès au serveur Prometheus"

dicos/70_prometheus.xml

@@ -1,9 +1,10 @@
 <?xml version="1.0" encoding="utf-8"?>
 <creole>
     <files>
-        <file filelist='prometheus'   name='/etc/prometheus/prometheus.yml'   mkdir='True' rm='True'/>
+        <file filelist='prometheus'   name='/etc/prometheus/prometheus.yml'          mkdir='True' rm='True'/>
-        <file filelist='alertmanager' name='/etc/prometheus/alertmanager.yml' mkdir='True' rm='True'/>
+        <file filelist='alertmanager' name='/etc/prometheus/alertmanager.yml'        mkdir='True' rm='True'/>
-        <file filelist='grafana'      name='/etc/grafana/grafana.ini'         mkdir='True' rm='True'/>
+		<file filelist='alertmanager' name='/etc/prometheus/rules.d/alert-rules.yml' mkdir='True' rm='True'/>
+        <file filelist='grafana'      name='/etc/grafana/grafana.ini'                mkdir='True' rm='True'/>
 
         <service>prometheus</service>
         <service>alertmanager</service>
@@ -102,6 +103,9 @@
                 </family>
 
         <family name="alertes prometheus">
+			<variable name='alSMTPUseSys' type='oui/non' description="Utiliser la passerelle SMTP du système ?">
+				<value>non</value>
+			</variable>
             <variable name='alSMTPHost' type='string' description="Adresse du serveur SMTP pour l'envois des alertes"/>
             <variable name='alSMTPPort' type='string' description="Port d'écoute du serveur SMTP pour l'envois des alertes"/>
             <variable name='alFrom' type='string' description="Adresse d'origine des emails d'alerte"/>
@@ -210,6 +214,16 @@
             <param>['','service','severity']</param>
         </check>
 
+        <condition name='disabled_if_in' source='alSMTPUseSys'>
+            <param>oui</param>
+            <target type='variable'>alSMTPUser</target>
+            <target type='variable'>alSMTPPass</target>
+            <target type='variable'>alSMTPPort</target>
+            <target type='variable'>alSMTPTLS</target>
+			<target type='variable'>alSMTPHost</target>
+			<target type='variable'>alSMTPAuth</target>
+        </condition>
+
         <condition name='disabled_if_in' source='alSMTPAuth'>
             <param>non</param>
             <target type='variable'>alSMTPUser</target>

tmpl/alert-rules.yml

@@ -0,0 +1,41 @@
+#
+# Alert Rules
+#
+groups:
+- name: EoleRules
+  rules:
+  # Instance is Down
+  - alert: JobInstanceDown
+    expr: up == 0
+    for: 1m
+    annotations:
+      DESCRIPTION: Job {{ $labels.job }} instance {{ $labels.instance }} is down.
+      SUMMARY: Job instance is down
+
+  # Heavy CPU usage
+  - alert: cpu_threshold_exceeded
+    expr: (100 * (1 - avg by(instance) (irate(node_cpu{job="%%{job_name_node}",mode="idle"}[5m]))))
+      > 80
+    annotations:
+      description: This device's cpu usage has exceeded the threshold with a value
+        of {{ $value }}.
+      summary: Instance {{ $labels.instance }} CPU usage is dangerously high
+
+  # Heavy Memory usage
+  - alert: mem_threshold_exceeded
+    expr: (node_memory_MemFree{job="%%{job_name_node}"} + node_memory_Cached{job="%%{job_name_node}"} + node_memory_Buffers{job="%%{job_name_node}"})
+      / 1e+06 < 80
+    annotations:
+      description: This device's memory usage has exceeded the threshold with a value
+        of {{ $value }}.
+      summary: Instance {{ $labels.instance }} memory usage is dangerously high
+
+  # Heavy "/" use
+  - alert: filesystem_threshold_exceeded
+    expr: node_filesystem_avail{job="%%{job_name_node}",mountpoint="/"} / node_filesystem_size{job="%%{job_name_node}"}
+      * 100 < 90
+    annotations:
+      description: This device's filesystem usage has exceeded the threshold with
+        a value of {{ $value }}.
+      summary: Instance {{ $labels.instance }} filesystem usage is dangerously high
+

tmpl/alertmanager.yml

@@ -1,16 +1,28 @@
 global:
   # The smarthost and SMTP sender used for mail notifications.
+%if %%alSMTPUseSys ==  'oui'
+  %if %%tls_smtp == "non"
+  smtp_smarthost: '%%exim_relay_smtp:25'
+  %elif %%tls_smtp == "port 25"
+  smtp_smarthost: '%%exim_relay_smtp:25'
+  smtp_require_tls: true
+  %else
+  smtp_smarthost: '%%exim_relay_smtp:465'
+  smtp_require_tls: true
+  %end if
+%else
   smtp_smarthost: '%%alSMTPHost:%%alSMTPPort'
   smtp_from: '%%alFrom'
-%if %%getVar('alSMTPAuth','non') == 'oui'
+  %if %%getVar('alSMTPAuth','non') == 'oui'
   smtp_auth_username: '%%alSMTPUser'
   smtp_auth_password: 'alSMTPPass'
-%end if
+  %end if
 
-%if %%getVar('alSMTPTLS','non') == 'oui'
+  %if %%getVar('alSMTPTLS','non') == 'oui'
   smtp_require_tls: true
-%else
+  %else
   smtp_require_tls: false
+  %end if
 %end if
 
   # The auth token for Hipchat.

tmpl/prometheus.yml

@@ -5,7 +5,7 @@ global:
   scrape_timeout: %%prometheusScrapeTimeout
 
 rule_files:
-- "/etc/prometheus/rules.d/*.yml"
+  - "/etc/prometheus/rules.d/*.yml"
 
 scrape_configs:
   - job_name: %%prometheusJobName