fix(sentinel): reduce resource churn from health flaps

Ignore health status changes in Sentinel push deduplication when the container lifecycle state is unchanged.

Scope stale resource checks to Sentinel servers whose heartbeat is stale, and avoid refreshing resource last_online_at on unchanged statuses.

config/constants.php

@@ -93,9 +93,15 @@ return [
 
     'sentinel' => [
         // How often (seconds) PushServerUpdateJob is force-dispatched even when
-        // the container state hash is unchanged. Keeps last_online_at,
-        // exited-detection and storage checks from going stale.
+        // the container state hash is unchanged. Keeps exited-detection and
+        // storage checks from going stale without writing every resource row on
+        // every push.
         'push_force_interval_seconds' => env('SENTINEL_PUSH_FORCE_INTERVAL_SECONDS', 300),
+
+        // How long a Sentinel-enabled server may go without a heartbeat before
+        // ResourcesCheck considers its resources stale. Per-resource
+        // last_online_at is only updated on real status changes, not every push.
+        'resource_stale_seconds' => env('SENTINEL_RESOURCE_STALE_SECONDS', 300),
     ],
 
     'proxy' => [