Multi-deployment: one k8s Deployment per pod in stack.yml

Each pod entry in stack.yml now creates its own k8s Deployment with
independent lifecycle and update strategy. Pods with PVCs get Recreate,
pods without get RollingUpdate. This enables maintenance services that
survive main pod restarts.

- cluster_info: get_deployments() builds per-pod Deployments, Services
- cluster_info: Ingress routes to correct per-pod Service
- deploy_k8s: _create_deployment() iterates all Deployments/Services
- deployment: restart swaps Ingress to maintenance service during Recreate
- spec: add maintenance-service key

Single-pod stacks are backward compatible (same resource names).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

stack_orchestrator/deploy/spec.py

@@ -264,5 +264,14 @@ class Spec:
     def is_kind_deployment(self):
         return self.get_deployment_type() in [constants.k8s_kind_deploy_type]
 
+    def get_maintenance_service(self) -> typing.Optional[str]:
+        """Return maintenance-service value (e.g. 'dumpster-maintenance:8000') or None.
+
+        When set, the restart command swaps Ingress backends to this service
+        during the main pod Recreate, so users see a branded maintenance page
+        instead of a bare 502.
+        """
+        return self.obj.get("maintenance-service")
+
     def is_docker_deployment(self):
         return self.get_deployment_type() in [constants.compose_deploy_type]