Added logo and screenshots

.kiro/specs/pabawi-v0.5.0-release/cache-issue-resolution.md

@@ -0,0 +1,79 @@
+# Cache Issue Resolution
+
+## Problem
+
+After the routes refactoring, the inventory page shows no nodes and the home page shows "no integration configured". 
+
+## Root Cause
+
+The request deduplication middleware (in-memory cache) has stale data from before the refactoring. The cache key for `/api/inventory` is associated with the wrong response data (integration status instead of inventory nodes).
+
+## Evidence
+
+```bash
+# Without cache busting - returns wrong data (integrations instead of nodes)
+curl http://localhost:3000/api/inventory
+# Returns: {"integrations": [...]}  ❌ WRONG
+
+# With cache busting parameter - returns correct data
+curl "http://localhost:3000/api/inventory?_t=$(date +%s)"
+# Returns: {"nodes": [...], "sources": {...}}  ✅ CORRECT
+```
+
+## Solution
+
+The cache is in-memory and will be cleared when the backend restarts. 
+
+### Option 1: Restart Backend (Recommended)
+```bash
+# Stop the backend process (Ctrl+C if running in terminal)
+# Or kill the process
+pkill -f "node.*backend"
+
+# Start backend again
+cd backend
+npm run dev
+```
+
+### Option 2: Wait for Cache Expiration
+The cache TTL is 60 seconds, so the stale data will expire automatically after 1 minute.
+
+### Option 3: Add Cache Clear Endpoint (Future Enhancement)
+Add an admin endpoint to clear the deduplication cache:
+
+```typescript
+// In backend/src/server.ts
+app.post("/api/admin/cache/clear", (_req: Request, res: Response) => {
+  deduplicationMiddleware.clear();
+  res.json({ message: "Cache cleared successfully" });
+});
+```
+
+## Prevention
+
+To prevent this issue in the future:
+
+1. **Clear cache after major refactoring**: Add a step to restart the backend after significant route changes
+2. **Add cache versioning**: Include a version number in cache keys that changes with deployments
+3. **Reduce cache TTL during development**: Use shorter TTL (e.g., 10 seconds) in development mode
+4. **Add cache clear endpoint**: Provide an admin endpoint to manually clear cache
+
+## Verification
+
+After restarting the backend, verify the fix:
+
+```bash
+# Test inventory endpoint
+curl http://localhost:3000/api/inventory | jq '.nodes | length'
+# Should return: 8 (or your actual node count)
+
+# Test integration status endpoint  
+curl http://localhost:3000/api/integrations/status | jq '.integrations | length'
+# Should return: 4 (bolt, puppetdb, puppetserver, hiera)
+```
+
+## Status
+
+✅ **Root cause identified**: Stale cache from refactoring
+✅ **Solution provided**: Restart backend to clear cache
+✅ **API endpoints verified**: Both endpoints work correctly with cache-busting parameters