while sending multiple failing port-forwarding requests, crio memory usage increases until the node dies

OCP 4.15.2
cri-o-1.28.4-4.rhaos4.15.git92d1839.el9.x86_64

all the time

apiVersion: v1
kind: Pod
metadata:
  name: myclient
spec:
  restartPolicy: Never
  containers:
  - name: myclient
    image: quay.io/centos/centos:stream9
    ports:
    - containerPort: 80
    args: ["bash", "-c", "sleep 12000" ]
    resources:
      requests:
        cpu: 200m
    imagePullPolicy: IfNotPresent
    readinessProbe:
      failureThreshold: 3000
      httpGet:
        path: /ready
        port: 8080
        scheme: HTTP
      initialDelaySeconds: 1
      periodSeconds: 1
      successThreshold: 1
      timeoutSeconds: 1 

    1. Create this pod
    2. while sleep 1 ; do kubectl port-forward myclient 15000:8080 & sleep 1 && curl localhost:15000/ && wait; done     

    

crio memory increases as long as the while loop above is running until there is no memory left

crio memory remains stable