Make FS space alerts thresholds configurable (prometheus#1624)

* Make FS space alerts thresholds configurable (#1)

This makes it possible to tweak the thresholds for
the NodeFilesystemSpaceFillingUp alerts. Which
might be necessary in systems like Kubernetes,
where the image garbage collector runs at 85%,
so it's not a problem that the disk reaches that usage %.

Signed-off-by: iuri aranda <[email protected]>

Author: iuriaranda

docs/node-mixin/alerts/alerts.libsonnet

@@ -8,7 +8,7 @@
             alert: 'NodeFilesystemSpaceFillingUp',
             expr: |||
               (
-                node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s} / node_filesystem_size_bytes{%(nodeExporterSelector)s,%(fsSelector)s} * 100 < 40
+                node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s} / node_filesystem_size_bytes{%(nodeExporterSelector)s,%(fsSelector)s} * 100 < %(fsSpaceFillingUpWarningThreshold)d
               and
                 predict_linear(node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s}[6h], 24*60*60) < 0
               and
@@ -28,7 +28,7 @@
             alert: 'NodeFilesystemSpaceFillingUp',
             expr: |||
               (
-                node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s} / node_filesystem_size_bytes{%(nodeExporterSelector)s,%(fsSelector)s} * 100 < 20
+                node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s} / node_filesystem_size_bytes{%(nodeExporterSelector)s,%(fsSelector)s} * 100 < %(fsSpaceFillingUpCriticalThreshold)d
               and
                 predict_linear(node_filesystem_avail_bytes{%(nodeExporterSelector)s,%(fsSelector)s}[6h], 4*60*60) < 0
               and

docs/node-mixin/config.libsonnet

@@ -35,6 +35,18 @@
     // just a warning for K8s nodes.
     nodeCriticalSeverity: 'critical',
 
+    // Available disk space (%) thresholds on which to trigger the
+    // 'NodeFilesystemSpaceFillingUp' alerts. These alerts fire if the disk
+    // usage grows in a way that it is predicted to run out in 4h or 1d
+    // and if the provided thresholds have been reached right now.
+    // In some cases you'll want to adjust these, e.g. by default Kubernetes
+    // runs the image garbage collection when the disk usage reaches 85%
+    // of its available space. In that case, you'll want to reduce the
+    // critical threshold below to something like 14 or 15, otherwise
+    // the alert could fire under normal node usage.
+    fsSpaceFillingUpWarningThreshold: 40,
+    fsSpaceFillingUpCriticalThreshold: 20,
+
     grafana_prefix: '',
   },
 }