MCP-Bridge sometimes stucks with no response

[baryhuang]

I'm not sure if you or anyone has seen this.

Once a while, I have to reboot my EC2 and restart the mcp-bridge to get things back to working.

I'm using a large EC2, running two mcp-bridge service.

top - 19:09:12 up 23 min,  1 user,  load average: 0.00, 0.00, 0.00
Tasks: 133 total,   1 running, 132 sleeping,   0 stopped,   0 zombie
%Cpu(s):  0.3 us,  0.2 sy,  0.0 ni, 99.4 id,  0.0 wa,  0.0 hi,  0.0 si,  0.2 st 
MiB Mem :   7820.9 total,   6236.0 free,   1079.7 used,    759.9 buff/cache     
MiB Swap:      0.0 total,      0.0 free,      0.0 used.   6741.2 avail Mem 

I FEEL that something if a service got stuck for whatever reason, the mcp-bridge doesn't respond any more. The logs doesn't show obvious clues.

Is there some on-going work related to this area? Would it be worth I take another look and spend some effort to try to improve the debugability and robustness?

[SecretiveShell]

without logging or some kind of error message this is almost impossible to look into. Make sure you are running with debug log level in future, and list all processes on the machine when it seems to stop responding (it may be one of your MCP servers has an issue and died, though they should restart if the ping fails)

Would it be worth I take another look and spend some effort to try to improve the debugability and robustness?

I recently added a vscode debug profile, but otherwise logging is not a high priority case right now (I want to focus on openrouter support, as people keep opening issues about it) I would love to see a PR though if you are interested.