The endpoint stopped responding after some time

The endpoint deployed successfully, and you previously verified it was working, but it is currently unresponsive. There are several possible reasons for this behavior:

1. Insufficient account balance. If your balance reaches zero or becomes negative, the server will be stopped and archived (Shelved).In this case: Top up your balance, then either Unshelve the corresponding virtual machines or Start the dedicated server.
2. Critical error in vLLM operation. This could be caused by too many concurrent requests, insufficient CUDA video memory (Out of Memory error), or an internal framework error in vLLM.

In this case, you can:

• View and save the error log from the virtual machine console;
• If the endpoint was created on a dedicated server, connect via SSH and view/save the log using docker logs;
• Restart the vLLM container;
• Delete the endpoint and create a new one;
• Contact technical support.