fix: remove legacy RebalancingService and migrate WebSocket broadcast to AutoRebalancerService

[Jaydbrown]

Closes #5

Problem

RebalancingService (monitoring/rebalancer.ts) was left instantiated and started in index.ts even after it was superseded by AutoRebalancerService. Two problems remained:

1. Dead code at startup — new RebalancingService(wss) + rebalancingService.start() ran on every server boot, did nothing (the cron had already been removed), and added noise.
2. Lost WebSocket broadcast — notifyClients() and broadcastToAllClients() lived only inside RebalancingService and were never called, silently dropping real-time portfolio events.

Changes

backend/src/index.ts — 1 line removed, 9 lines removed, 2 added

-import { RebalancingService } from './monitoring/rebalancer.js'

-// Start existing rebalancing service (now queue-backed, no cron)
-try {
-    const rebalancingService = new RebalancingService(wss)
-    rebalancingService.start()
-    console.log('[REBALANCING-SERVICE] Monitoring service started (queue-backed)')
-} catch (error) {
-    console.error('Failed to start rebalancing service:', error)
-}
+// Wire wss into autoRebalancer so it can push real-time portfolio events to clients
+autoRebalancer.setWss(wss)

backend/src/services/autoRebalancer.ts — 54 lines added

• Added import { WebSocketServer } from 'ws'
• Added private wss: WebSocketServer | null = null field
• Added setWss(wss) — called from index.ts after the WebSocket server is created
• Added hasWss() — useful for health checks and tests
• Migrated notifyClients(portfolioId, event, data) from RebalancingService
• Migrated broadcastToAllClients(event, data) from RebalancingService

backend/src/monitoring/rebalancer.ts — deleted

All live functionality has been consolidated into AutoRebalancerService.

Why setWss() instead of constructor injection

autoRebalancer is instantiated at line 86 of index.ts but wss = new WebSocketServer(...) is created at line 214. setWss() avoids reordering the startup sequence and keeps the change minimal.

Failure scenarios resolved

Scenario	Before	After
Double rebalance (cron + queue)	Possible if cron accidentally restored	Impossible — one scheduler
Duplicate recordRebalanceEvent() writes	Possible	Impossible — one code path
WebSocket broadcast silently dropped	Always	Fixed — notifyClients() now reachable
Dead startup try/catch	Present	Removed