From 5b4a500d0cbddc1d6710264a9774491faedcdde9 Mon Sep 17 00:00:00 2001
From: Robert Gildein <gildeinrobert@gmail.com>
Date: Thu, 22 Aug 2024 21:13:00 +0200
Subject: [PATCH] Add alert rules to istio-pilot based on the KF093 spec (#515)

---
 .../KubeflowIstioPilotServices.rules          | 24 +++++++++++++++++++
 1 file changed, 24 insertions(+)
 create mode 100644 charms/istio-pilot/src/prometheus_alert_rules/KubeflowIstioPilotServices.rules

diff --git a/charms/istio-pilot/src/prometheus_alert_rules/KubeflowIstioPilotServices.rules b/charms/istio-pilot/src/prometheus_alert_rules/KubeflowIstioPilotServices.rules
new file mode 100644
index 00000000..b4671042
--- /dev/null
+++ b/charms/istio-pilot/src/prometheus_alert_rules/KubeflowIstioPilotServices.rules
@@ -0,0 +1,24 @@
+groups:
+- name: KubeflowIstioPilotServices
+  rules:
+  - alert: KubeflowServiceDown
+    expr: up{} < 1
+    for: 5m
+    labels:
+      severity: critical
+    annotations:
+      summary: "{{ $labels.juju_charm }} service is Down ({{ $labels.juju_model }}/{{ $labels.juju_unit }})"
+      description: |
+       One or more targets of {{ $labels.juju_charm }} charm are down on unit {{ $labels.juju_model }}/{{ $labels.juju_unit }}.
+       LABELS = {{ $labels }}
+
+  - alert: KubeflowServiceIsNotStable
+    expr: avg_over_time(up{}[10m]) < 0.5
+    for: 0m
+    labels:
+      severity: warning
+    annotations:
+      summary: "{{ $labels.juju_charm }} service is not stable ({{ $labels.juju_model }}/{{ $labels.juju_unit }})"
+      description: |
+        {{ $labels.juju_charm }} unit {{ $labels.juju_model }}/{{ $labels.juju_unit }} has been unreachable at least 50% of the time over the last 10 minutes.
+        LABELS = {{ $labels }}