Connection leak in cri-client #133183
Description
What happened?
creates a connection to the specified endpoint, howewer, it is not possible for the caller to later close it. CRI-O uses this method for its internal health check. As client connections dont get closed, nor do server ones, and then they'll eat up resources.What did you expect to happen?
No memory/connection leaks in client libraries too.
How can we reproduce it (as minimally and precisely as possible)?
See cri-o/cri-o#9360
Anything else we need to know?
No response
Kubernetes version
# kubectl version
Client Version: v1.33.1
Kustomize Version: v5.6.0
Server Version: v1.33.1Cloud provider
kvm
OS version
No response
Install tools
kubeadm
Container runtime (CRI) and version (if applicable)
crio 1.33
Related plugins (CNI, CSI, ...) and versions (if applicable)
No response
Metadata
Metadata
Assignees
Labels
Type
Projects
Status