[Bug] Sending heartbeat msg without a return value resulted in session expiration

### Search before asking

- [X] I had searched in the [issues](https://github.com/apache/eventmesh/issues?q=is%3Aissue) and found no similar issues.

### Environment

Windows

### EventMesh version

1.2.0

### What happened

After the client subscribes to the message [TCP], asynchronous [thread pool with 5 threads] reads the message and sleeps for 5 minutes. The server pushes a large number of messages;

Problem: Heartbeat heartbeat, unable to send to the server, resulting in session timeout being cleared;

### How to reproduce

1. Subscribe to a topic, and then sleep for 5 minutes after receiving the message;

2. Send 1000 messages to the topic

### Debug logs

```Java
2024-01-29 19:00:13.970 WARN 123056 --- [ientScheduler-4] o.a.e.c.tcp.impl.SimpleSubClientImpl : <=heartbeat=>group:test_1_count_g ,isActive:true ,address:/10.19.239.145:63610 ,error:operation timeout, context.key=4613632557
```

### Are you willing to submit PR?

- [X] Yes I am willing to submit a PR!

### Code of Conduct

- [X] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct) *

Start at the SimpleSubClientImpl heartbeat path shown in the debug log and reproduce the TCP subscription scenario with five-minute sleeps and 1,000 messages. Trace why the heartbeat cannot be sent while message handling is busy; done means heartbeats continue reaching the server and the session remains active under this load.

Written by the indexing model from the issue text.