I've tried to build an API application based on FastAPI to process some requests using AI models (Pytorch and Paddle framework).

Sometimes, I see those logs appearing continuously in the console. And at that time, I cannot reach any end-points anymore (response time out error). The only thing I can do is to restart the server and everything back to normal.

I don't know how to reproduce this error and have no idea where it came from.

How can I debug it? And which extra information that I need to provide for a closer view on this problem?

| W0308 23:13:21.249563    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:27.000713    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:31.084384    87 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:32.490388    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:40.359278    87 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:41.317720    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!                                                                        
| W0308 23:13:48.468175    87 sampler.cpp:139] bvar is busy at sampling for 2 seconds!
| W0308 23:13:52.974829    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!
| W0308 23:14:04.213120    40 sampler.cpp:139] bvar is busy at sampling for 2 seconds!