BentoML

Without any logs this would be hard to diagnose

<@U03NQ3FCY6A> There wasn't anything in the logs..
I added prints to the webserver startup and configuration phase to understand what happens. These are the logs I get:
```configuring server logging mode                                                                                                                                                                                                                                                                                              Starting server in quiet mode...```
And these are logs I have generated myself. Basically it shows me all the startup of the application all over again

I can add prints at certain places if you want to generate those logs

It didn't show any worker timeout or anything like that. Just suddenly restarted

It might be worth to check the kubernetes logs

Sometimes it be `OOMKilled` for example which might not be apparent in the logs of the program but you would immediately notice when you do `kubectl describe po...`

I am not using kubernetes
I deploy it simply on a server

If I knew all the reasons a worker or runner can crush I would inspect the reasons and see if that's the case