Real-Time Analytics

Bklit includes real-time features that update your dashboard instantly as visitors interact with your site.

What You Get

Live Visitor Notifications

Get notified the moment someone visits your site. Toast notifications appear within 1 second showing:

• Visitor's location (country and city)
• Device type (mobile or desktop)
• Entry page

Live Map Updates

The globe on your Live analytics page updates instantly when visitors arrive. Watch pins appear in real-time showing exactly where your visitors are from.

Accurate Live Counts

The live visitor counter in your dashboard header updates immediately. No more stale numbers - you always know exactly how many people are on your site right now.

How It Works

Real-time analytics uses a queue-based architecture with Redis pub/sub and WebSockets:

1. Event queued - Ingestion service pushes to Redis queue (1-3ms)
2. Background processing - Worker pops events in batches
3. ClickHouse insert - Batch of 100 events saved (8-50ms)
4. Publish to Redis - Worker publishes to live-events channel
5. WebSocket broadcast - Server pushes to connected dashboards
6. Dashboard updates - You see the update within 1 second

This architecture ensures:

• Fast API responses (no waiting for database)
• Zero data loss (Redis queue persists)
• Efficient database writes (batching)
• Real-time updates (pub/sub)

Automatic Fallback

If real-time isn't available (Redis not configured or WebSocket down), everything automatically falls back to polling mode. Your analytics continue working normally - you just won't get instant updates.

Self-Hosted Setup

Real-time features are completely optional. If you don't configure Redis:

• Analytics work exactly as before
• Polling mode (30-60 second updates)
• No errors or degraded functionality

To enable real-time on your self-hosted instance, see our deployment guide.

What's Different

Before (Polling):

• Updates every 15-30 seconds
• Multiple ClickHouse queries per user
• Slight delays in live counts

Now (Real-Time):

• Updates within 1 second
• Fewer database queries
• Instant, accurate live data

The best part? You don't need to do anything - it works automatically in production!