concurrently load companies and users

README.md

@@ -108,6 +108,29 @@ These settings allow you to optimize performance for your specific infrastructur
 - **Memory Constrained**: Reduce channel sizes (50-100 each)
 - **High CPU**: Increase worker count up to 2x CPU cores
 
+### Async Initial Loading Configuration (New!)
+For faster connection setup and improved resilience with large datasets (>1000 companies/users):
+
+- `USE_ASYNC_LOADING`: Enable asynchronous initial data loading (`true`/`false`, default: `false`)
+- `ASYNC_LOADER_PAGE_SIZE`: Page size for initial data loading (default: `100`)
+- `ASYNC_LOADER_CIRCUIT_BREAKER_THRESHOLD`: API call failures before circuit breaker opens (default: `5`)
+- `ASYNC_LOADER_CIRCUIT_BREAKER_TIMEOUT`: Circuit breaker recovery timeout (default: `30s`)
+
+**When to use Async Loading**:
+- ✅ **Large datasets** (2000+ companies/users): Reduces connection time from ~60s to ~2s
+- ✅ **Production environments**: WebSocket connects quickly, data loads in background
+- ✅ **High availability**: Circuit breaker protects against API overload
+- ❌ **Small datasets** (<500 companies/users): Synchronous loading is sufficient
+- ❌ **Development**: Synchronous loading provides simpler debugging
+
+Example for production with large datasets:
+```bash
+export USE_ASYNC_LOADING="true"
+export ASYNC_LOADER_PAGE_SIZE="100"
+export ASYNC_LOADER_CIRCUIT_BREAKER_THRESHOLD="5"
+export ASYNC_LOADER_CIRCUIT_BREAKER_TIMEOUT="30s"
+```
+
 ### Redis Configuration (Required)
 
 **Note**: Redis is mandatory for the datastream replicator. The application will exit if it cannot connect to Redis.