CODEWITHSUNDEEP
pythonkubernetesfastapigunicorn
Saw
Saw

Reputation: 1

Kubernetes and Gunicorn health check issues

I currently have an application that uses fastapi + gunicorn on a production server.

It runs with 3 async workers, and it can only have that many workers.

This application receives hundreds of requests per second, and this causes the workers to be almost always busy with some task, so when the Kubernetes health check system tries to make an internal request to the application and is unable to do so (because all workers are busy), Kubernetes forces the container to restart because it thinks there is no availability.

Have you ever had a similar use case? What did you do to resolve this situation?

Upvotes: 0

Views: 47

Answers (0)

Related Questions