9090 port connect refuse after app boot

app startup command
``` shell
export PROMETHEUS_MULTIPROC_DIR="/tmp/prometheus_multiproc"
export PROMETHEUS_METRICS_PORT="9090"
export PROMETHEUS_BIND_ADDRESS="0.0.0.0"
export GUNICORN_WORKERS="4"
BIND_ADDR=unix:/run/engine_receipt.sock;
PID_FILE=/run/engine_receipt.pid
rm -f /run/engine_receipt.pid
/tf212py311/bin/gunicorn \
-w 1 \
-c gunicorn.conf.py \
--threads 3 \
-t 1600 \
-b $BIND_ADDR \
-p $PID_FILE \
server:app
```
`gunicorn.conf.py` content
```python
# Basic Gunicorn settings
worker_class = "gunicorn_prometheus_exporter.PrometheusWorker"

# Prometheus configuration
import os
os.environ.setdefault("PROMETHEUS_MULTIPROC_DIR", "/tmp/prometheus_multiproc")
os.environ.setdefault("PROMETHEUS_METRICS_PORT", "9090")
os.environ.setdefault("PROMETHEUS_BIND_ADDRESS", "0.0.0.0")
os.environ.setdefault("GUNICORN_WORKERS","4")

```
after app startup, it shows logs:
```
[2025-09-15 15:32:20.690 |2572|MainThread|INFO] - setup_app finish..., takes 256.9923074245453
[2025-09-15 15:32:20.691 |2572|MainThread|WARNING] - gunicorn_error_logger.level: 20
[2025-09-15 15:33:13.286 |2572|MainThread|INFO] - Updated metrics for worker worker_1_1757921264: memory=46047232, cpu=0.0, uptime=328.3646159172058
[2025-09-15 15:37:38.818 |2572|MainThread|INFO] - Updated metrics for worker worker_1_1757921264: memory=46047232, cpu=0.0, uptime=593.8966326713562
[2025-09-15 15:37:39.537 |2572|MainThread|INFO] - Updated metrics for worker worker_1_1757921264: memory=46047232, cpu=0.0, uptime=594.6158511638641

```
but the 9090 port connect refused, how to solve it?