Merge pull request 'test(uptime-monitor): remove the audiobooks EndpointMonitor — REN-7 criterion 3 removal test' (#1018) from atlas/ren-7-delete-test-audiobooks into master

kubernetes/apps/uptime-monitor/pilot-endpointmonitors.yaml

@@ -1,4 +1,34 @@
 ---
+# ============================================================================
+# REN-7 CRITERION 3 — THE REMOVAL TEST. THIS COMMIT DELETES THE `audiobooks`
+# ENDPOINTMONITOR ON PURPOSE. IT IS NOT A CLEANUP AND NOT AN ABANDONMENT.
+#
+# The card requires proof in both directions: an EndpointMonitor creates a
+# monitor, and DELETING the EndpointMonitor removes it again. "A monitor that
+# cannot be removed is a leak." Create is proven — `audiobooks.renner.uno-
+# uptime-monitor`, UptimeRobot id 804092841, created 2026-09-26 03:26:27,
+# controller log `Monitor Added`. Removing the object here is the other half.
+#
+# What must be observed after this merges: the public feed
+# `https://uptime.renner.uno/api/getMonitorList/UgMXmP1iRV` goes from
+# `psp.totalMonitors: 10` to `9`, and id 804092841 is absent. The controller
+# log shows `Deleting Monitor: audiobooks.renner.uno-uptime-monitor`.
+#
+# PRECONDITION, CHECKED BEFORE THIS COMMIT WAS WRITTEN: v2.2.13 has no
+# finalizer, so a delete while the controller is down orphans the monitor
+# forever (see the caveat at the end of this header). Pod
+# `ingressmonitorcontroller-5d577cc58-pgqwg` was `Ready=True`, `restarts=0`,
+# up since 2026-09-26T00:43:45Z when this was authored. Anyone merging this
+# must re-check that the pod is Ready first; if it is not, hold the merge.
+#
+# RE-CREATING IT IS THE REVERT: revert this commit and Flux re-applies the
+# object, and the controller creates the monitor again under the same name.
+# That round trip is also the `idempotent reconcile` check, so the revert is
+# worth doing deliberately rather than only if something goes wrong. Expect a
+# NEW UptimeRobot id — the friendly name is what binds object to monitor, not
+# the id.
+# ============================================================================
+#
 # REN-7 — the pilot batch. Two of the 14 monitors from the survey's §A0 page
 # list, chosen because they are the two lowest-risk rows on it:
 #
@@ -135,6 +165,31 @@
 # unavailable. Check `kube_deployment_status_replicas_available{namespace=
 # "uptime-monitor"}` first, and if a monitor is orphaned anyway, delete it by id
 # through the API rather than hoping a later reconcile finds it.
+#
+# ---
+# STATE OF THE OBJECT BELOW: ITS CREATE IS BEING REJECTED, AND IT RETRIES
+# FOREVER. TRACKED ON REN-155 — LEFT IN PLACE DELIBERATELY.
+#
+# `www.renner.uno-uptime-monitor` does not exist in UptimeRobot. Controller log,
+# every 300s since 2026-09-26T03:26:27Z:
+#
+#   Creating Monitor: www.renner.uno-uptime-monitor
+#   AddMonitor Request failed. Status Code: 403
+#
+# That is the `else` branch of `Add()` (uptime-monitor.go:187) — a NON-200, and
+# therefore NOT the provider's own validation path, which returns HTTP 200 with
+# `{"stat":"fail", ...}` and is logged at :174 with the message text. So the
+# common guesses are excluded: a duplicate URL, a monitor-count cap and a bad
+# parameter would all arrive as 200 + `stat: fail`, not as 403. The 403 body is
+# discarded by that branch, so the provider's reason is not recoverable from the
+# log; settling it needs the API key or the account dashboard.
+#
+# It is left in git rather than removed because it is harmless and it is the
+# reproducer: `Add()` is fire-and-forget (`handleCreate` returns nil regardless,
+# endpointmonitor_created.go:12-31), so there is no error, no status and no
+# alert — only that log line, at INFO. Nothing pages, nothing degrades, and the
+# object is the only remaining evidence. Anyone tempted to "clean it up" should
+# close REN-155 first.
 apiVersion: endpointmonitor.stakater.com/v1alpha1
 kind: EndpointMonitor
 metadata:
@@ -157,18 +212,15 @@ spec:
     # matches the interval of the nine monitors already in the account. Written
     # out rather than left implicit so the value is reviewable here.
     interval: 300
----
-apiVersion: endpointmonitor.stakater.com/v1alpha1
-kind: EndpointMonitor
-metadata:
-  name: audiobooks.renner.uno
-  namespace: uptime-monitor
-spec:
-  # The path is in the URL rather than in `spec.healthEndpoint`, because
-  # `healthEndpoint` is only consulted on the ref-discovery path and is
-  # explicitly ignored when `spec.url` is set (pkg/kube/util/url.go:31-33).
-  # Endpoint taken from the existing Gatus check (apps/gatus/configmap.yaml:134).
-  url: https://audiobooks.renner.uno/healthcheck
-  uptimeRobotConfig:
-    monitorType: http
-    interval: 300
+# The `audiobooks.renner.uno` EndpointMonitor was here. It is removed by this
+# commit as REN-7's criterion-3 removal test — see the header block at the top
+# of this file for what that proves, what to observe, and how to put it back.
+# Its spec, for the revert:
+#
+#   url: https://audiobooks.renner.uno/healthcheck   # Gatus check,
+#                                                    # apps/gatus/configmap.yaml:134
+#   uptimeRobotConfig: { monitorType: http, interval: 300 }
+#
+# The path belongs in `spec.url` and not in `spec.healthEndpoint`, because
+# `healthEndpoint` is only consulted on the ref-discovery path and is explicitly
+# ignored when `spec.url` is set (pkg/kube/util/url.go:31-33).