fix(prometheus): give cadvisor room for a full swarm node

cadvisor's footprint grows with the number of containers it watches. On a
swarm lab node carrying the whole stack it reached 436MB and the cgroup
OOM-killer reaped it, taking the deploy down with rc 137 minutes later.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

roles/web-app-prometheus/meta/services.yml

@@ -100,8 +100,8 @@ cadvisor:
   version: v0.52.1
   name: prometheus-cadvisor
   cpus: "0.25"
-  mem_reservation: 64m
-  mem_limit: 512m
+  mem_reservation: 128m
+  mem_limit: 1g
   pids_limit: 64
   min_storage: 1GB
   ports: