Slow / Long Loading
1. The Very First Message Is Slow
After the workspace is idle, Brain cold-starts; the first message may take 30–60 seconds to respond. This is expected. Subsequent messages are noticeably faster.
2. Panel as a Whole Is Slow
- Chat too long: thousands of messages slow rendering — start a new chat to shard.
- Too many browser tabs — close some.
- Clear browser cache and Service Worker (DevTools → Application → Clear storage).
3. Page Switching Lags
- Task, Memory and Database pages can load a lot — with thousands of entries it's naturally slow.
- Search or filter to narrow the view.
4. AI Replies Feel Slow
- During peak hours (8–11 PM Asia, UTC+8) models are under heavy load; each reply may be 2–3 seconds slower.
- Running several complex tasks concurrently also queues.
- For short messages, add "please keep it brief" to cut generation time.
5. Slow Uploads
- Depends on your network path to AWS; corporate networks or VPNs are often the bottleneck.
- Large files (>10MB) taking 30+ seconds is normal — compress images first.
6. Channel Round-Trip Delay
- Telegram, LINE and Email each add their own transport delay (Email in particular can exceed 10 seconds).
- To see how fast DYNARA itself is, compare timestamps directly in the Panel chat.
