Why can a single worker process use multiple cores in a k8s environment?

[fighterhit]

Why can a single worker process use multiple cores in a k8s environment? According to my understanding, one python process can only use one core?

I test the program using the following config in k8s cluster:

# my pod command
gunicorn --workers 1 --worker-class uvicorn.workers.UvicornWorker --bind 0.0.0.0:8000 --keep-alive 60 main:app

# pod resource yaml

resources:
  limits:
    cpu: "3"
    memory: 8Gi
  requests:
    cpu: "3"
    memory: 8Gi

# cat /sys/fs/cgroup/cpuset/cpuset.cpus 
11,58-59

[benoitc]

Hi, uivicorn loop is probably using 1 core. Gunicorn with 1 worker may use 2 (1 for the arbiter , another for the worker). This depends on how the system will balance the tasks.

[fighterhit]

Hi, uivicorn loop is probably using 1 core. Gunicorn with 1 worker may use 2 (1 for the arbiter , another for the worker). This depends on how the system will balance the tasks.

Thanks @benoitc. Do you mean that the uivicorn loop uses 1 of the 3 cpu cores, and Gunicorn with 1 worker uses the remaining 2 cores? Does that mean that Gunicorn with 1 worker needs at least 2 CPUs?