求助：卡死问题排查

**Describe the bug (描述bug)**
adx服务在启动的时候大概率会出现卡死。vars接口显示创建了大量bthread和channel_connections，并且不断增长，最后导致OOM。如果能成功启动，服务运行还是比较稳定，高峰时bthread最多1800个，channel_connections最多2500个。

brpc服务接收到请求后会广播给下游dsp。协议基本都是http。我们对每一个dsp请求都会启一个bthread，bthread里同步请求下游服务，设定的超时时间为500ms左右。处理服务请求的bthread会join每个dsp的bthread。

服务最高并发不到100。dsp120个左右。

gdb调试的时候，worker线程堆栈目前看到有两种：

1. worker都在发送dsp请求，暂停在dsp的CallMethod。看堆栈貌似把pthread卡住了

2. worker都在建立ssl链接

有没有排查思路？

**To Reproduce (复现方法)**

**Expected behavior (期望行为)**

**Versions (各种版本)**
OS: ubuntu 22.04.3
Compiler: g++ 11.4
brpc: 1.6
protobuf: 3.20.1

**Additional context/screenshots (更多上下文/截图)**

Start with the vars output and the two GDB worker-stack patterns described: blocked DSP CallMethod calls and SSL connection setup. Inspect how the service creates and joins bthreads for each downstream DSP request, and how channel_connections grow during startup. Done means identifying the cause of the unbounded growth or documenting a reproducible diagnosis and concrete mitigation.

Written by the indexing model from the issue text.