fix: shutdown ends the live-event streams so a redeploy drains instead of waiting out its deadline (#5293)

http.Server.Shutdown waits for in-flight requests but never cancels them,
and an SSE stream only ends when its client leaves, so any open tab or
signed-in phone held every redeploy for the full 10s and the old task
exited 1. With the swarm's stop-first update that was ~10s of downtime per
deploy. The bus now closes on shutdown, ending each stream; clients
reconnect to the next instance.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

cmd/minstrel/main.go

@@ -481,6 +481,11 @@ func run() error {
 		// server's limitRequestBody middleware instead.
 		IdleTimeout: 120 * time.Second,
 	}
+	// Shutdown waits for in-flight requests but never cancels them, and an
+	// SSE stream only ends when its client leaves. Closing the bus ends
+	// every stream as shutdown begins, so the drain waits only on real work
+	// such as an audio stream; clients reconnect to the next instance (#5293).
+	httpServer.RegisterOnShutdown(bus.Close)
 
 	errCh := make(chan error, 1)
 	go func() {