chore: average alerts across namespace for 1 hour (#10827)

We now require the *average* of the increase in proven chain across a namespace to be 0 for an *hour* to trigger a slack alert.
AztecProtocol · Dec 17, 2024 · 962a7a2 · 962a7a2
1 parent 4ac13e6
commit 962a7a2
Showing 1 changed file with 6 additions and 6 deletions.
diff --git a/spartan/metrics/terraform/grafana.tf b/spartan/metrics/terraform/grafana.tf
@@ -59,11 +59,11 @@ resource "grafana_mute_timing" "mute_timing_always" {
   }
 }
 
-resource "grafana_rule_group" "rule_group_minutely" {
+resource "grafana_rule_group" "rule_group_hourly" {
   org_id           = 1
-  name             = "minutely-evaluation-group"
+  name             = "hourly-evaluation-group"
   folder_uid       = grafana_folder.rule_folder.uid
-  interval_seconds = 60
+  interval_seconds = 3600
 
   rule {
     name      = "Proven Chain is Live"
@@ -81,7 +81,7 @@ resource "grafana_rule_group" "rule_group_minutely" {
       model = jsonencode({
         disableTextWrap     = false,
         editorMode          = "code",
-        expr                = "increase(aztec_archiver_block_height{aztec_status=\"proven\"}[30m])",
+        expr                = "avg by(k8s_namespace_name) (increase(aztec_archiver_block_height{aztec_status=\"proven\"}[60m]))",
         fullMetaSearch      = false,
         includeNullMetadata = true,
         instant             = true,
@@ -118,15 +118,15 @@ resource "grafana_rule_group" "rule_group_minutely" {
           expression    = "A",
           intervalMs    = 1000,
           maxDataPoints = 43200,
-          refId         = "C",
+          refId         = "B",
           type          = "threshold"
         }
       )
     }
 
     no_data_state  = "NoData"
     exec_err_state = "Error"
-    for            = "1m"
+    for            = "1h"
     annotations    = {}
     labels         = {}
     is_paused      = false