fixes and alert improvements

[distro-setup] / filesystem / etc / prometheus / rules / iank.yml

diff --git a/filesystem/etc/prometheus/rules/iank.yml b/filesystem/etc/prometheus/rules/iank.yml
index 971e392fd63b55387089c47529060ebd91543d03..4f3bb04ee5e8e0ee5129638dcbc43c6c6458ea9b 100644 (file)
--- a/filesystem/etc/prometheus/rules/iank.yml
+++ b/filesystem/etc/prometheus/rules/iank.yml
@@ -115,6 +115,14 @@ groups:
     annotations:
       summary: '12 minutes down'
 
+  - alert: mailtest_check_unexpected_spamd_vps
+    expr: |-
+      mailtest_check_unexpected_spamd_results >= 1
+    labels:
+      severity: day
+    annotations:
+      summary: 'jr -u mailtest-check -e'
+
   - alert: mailtest_check_mailhost
     expr: |-
       time() - max by (folder,from) (mailtest_check_last_usec{job="node"}) >= 60 * 12
@@ -123,14 +131,14 @@ groups:
     annotations:
       summary: '12 minutes down'
 
-  # 42 mins: enough for a 30 min queue run plus 12
+  # 20 minutes. just allow for more due to prod alert.
   - alert: mailtest_check_gnu_mailhost
     expr: |-
-      time() - max by (folder,from) (mailtest_check_last_usec{folder="/m/md/l/testignore", from="iank@gnu.org"}) >= 60 * 42
+      time() - max by (folder,from) (mailtest_check_last_usec{folder="/m/md/l/testignore", from="iank@gnu.org"}) >= 60 * 20
     labels:
       severity: prod
     annotations:
-      summary: '42 minutes down'
+      summary: '20 minutes down'
 
 
   - alert: 1pmtest