Signals Outage Runbook

Detection

• Signal delivery latency > 5 minutes
• Discord bot offline or not posting
• Subscriber complaints
• Monitoring alert: signal_delivery_failures > 3

Response

1. Check signal service health: curl https://api.qgtm.ai/signals/health
2. Check Discord bot status: is it online in the server?
3. Check Redis: are signals being published to the stream?
4. Check Stripe webhooks: are subscription events processing?

Common Causes

Symptom	Likely Cause	Fix
Bot offline	Token expired or rate limited	Restart bot, check token
Signals not generating	Strategy service crashed	Check logs, restart
Signals generating but not delivering	Redis stream consumer lag	Check consumer group, restart
Delayed signals	API latency	Check Fly.io metrics, scale up

Communication

Post in Discord #announcements:

Signal delivery is currently delayed. We are investigating and will update within 30 minutes. Trading signals are still being generated — delivery will catch up once resolved.

Recovery

1. Fix root cause
2. Replay any missed signals from Redis stream
3. Post recovery notice in Discord
4. Review SLA impact for institutional tier subscribers