Experienced multiple times now, that the startup process gets stuck here, but without any error message.
Normally the container should answer back within 1 minute worst case. Anything above this time basically means that the container is not recoverable and will end in an error, but after several minutes of waiting.
If the waiting time is over 2 minutes, the process should abandon the container and start fresh.
That sounds harsh, but since we don't have persistent containers anymore, it'll end up the same (error -> new task).
Is there an existing issue for the same bug?
Describe the bug
Method
_wait_until_alive
in clientruntime
retry loop can get stuck for too long on unrecoverable container. https://github.com/OpenDevin/OpenDevin/blob/9cb0bf97c1212c801037a39a9222da098d3b53d7/opendevin/runtime/client/runtime.py#L195Experienced multiple times now, that the startup process gets stuck here, but without any error message. Normally the container should answer back within 1 minute worst case. Anything above this time basically means that the container is not recoverable and will end in an error, but after several minutes of waiting.
If the waiting time is over 2 minutes, the process should abandon the container and start fresh. That sounds harsh, but since we don't have persistent containers anymore, it'll end up the same (error -> new task).
Current OpenDevin version
Installation and Configuration
Model and Agent
Operating System
WSL
Reproduction Steps
No specific steps found.
Logs, Errors, Screenshots, and Additional Context
No response