feat: per-workload resources, storage GiB metrics, optional service disks

- Expose DB/Redis/RabbitMQ usage plus sidecars in getResourceUsage
- Storage API: GiB fields, Redis/Rabbit PVC usage, fix du/exec container names
- PATCH /resources accepts workload; persist entity fields only for app
- App detail: workload cards, disk bars, DB expand, scale target select

Co-authored-by: Cursor <cursoragent@cursor.com>

frontend/src/types/index.ts

@@ -185,9 +185,33 @@ export interface PodMetric {
   name: string;
   cpu: string;
   memory: string;
+  /** When metrics-server returns multiple containers (e.g. Fluent Bit sidecar) */
+  containers?: { name: string; cpu: string; memory: string }[];
+}
+
+export interface WorkloadResourceUsage {
+  key: 'app' | 'database' | 'redis' | 'rabbitmq';
+  title: string;
+  deploymentName: string;
+  configured: {
+    cpuRequest: string;
+    cpuLimit: string;
+    memoryRequest: string;
+    memoryLimit: string;
+    replicas: number;
+    readyReplicas: number;
+    availableReplicas: number;
+  };
+  pods: PodInfo[];
+  metrics: PodMetric[];
+  sidecars?: { name: string; cpuRequest?: string; cpuLimit?: string; memoryRequest?: string; memoryLimit?: string }[];
 }
 
 export interface ResourceUsage {
+  /** Per-deployment breakdown (app, DB, optional Redis/RabbitMQ) */
+  workloads?: WorkloadResourceUsage[];
+  /** Note when Elasticsearch logging sidecar is enabled */
+  loggingNote?: string;
   configured: {
     cpuRequest: string;
     cpuLimit: string;