Unclear with Container garbage collection configuration

[T-Lakshmi]

What happened?

Iam trying to disable maxPerPodContainer feature in my cluster. To achieve this followed the docs Container garbage collection and gone through code related to garbage collection.

Configured Container garbage collection parameters in the kubelet config file and restarted the kubelet sucessfully.
After that if i check, values in kubelet configuration file are persist but those are actually not reflected in cluster.

What did you expect to happen?

Container garbage collection parameters which i added in config file should reflect in cluster.

minAge:
maxPerPodContainer:
maxContainers:

How can we reproduce it (as minimally and precisely as possible)?

Ex:

1. Add below values in kubelet configuration file: /var/lib/kubelet/config.yaml

minAge: "5m"
maxPerPodContainer: 1
maxContainers: 50

2. Restart kubelet service:
   sudo systemctl restart kubelet
   Make sure kubelet restarted sucessfully: sudo systemctl status kubelet

3. Run: kubectl proxy

4. Open another terminal and check whether added parameters are present in configuration or not:
   curl http://localhost:8001/api/v1/nodes/<hostname/proxy/configz

Anything else we need to know?

I tried many ways to add these parameters but no use. even these are not listed in api-resources.
https://kubernetes.io/docs/reference/config-api/kubelet-config.v1beta1/
https://kubernetes.io/docs/reference/command-line-tools-reference/kubelet/

Is it possible to define these values in kubelet as mentioned in docs? if possible how can we configure them?

Kubernetes version

$ kubectl version
Client Version: v1.30.3
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.30.3

Cloud provider

OS version

# On Linux:
$ cat /etc/os-release
# paste output here
$ uname -a
# paste output here

# On Windows:
C:\> wmic os get Caption, Version, BuildNumber, OSArchitecture
# paste output here

Install tools

Container runtime (CRI) and version (if applicable)

Related plugins (CNI, CSI, ...) and versions (if applicable)

[T-Lakshmi]

/remove-kind bug
/sig node
/area kubelet

[saschagrunert]

It looks like we don't populate those values in the kubelet configuration any more after #53088:

minAge: "5m"
maxPerPodContainer: 1
maxContainers: 50

It's possible to set them using the CLI, though:

--minimum-container-ttl-duration duration                  Minimum age for a finished container before it is garbage collected.  Examples: '300ms', '10s' or '2h45m' (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)
--maximum-dead-containers int32                            Maximum number of old instances of containers to retain globally.  Each container takes up some disk space. To disable, set to a negative number. (default -1) (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)
--maximum-dead-containers-per-container int32              Maximum number of old instances to retain per container.  Each container takes up some disk space. (default 1) (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)

I think the whole container GC configuration is deprecated in favor of the eviction feature and I'm not sure we should do anything here.

[T-Lakshmi]

@saschagrunert,
Thanks for responding.

--minimum-container-ttl-duration duration                  Minimum age for a finished container before it is garbage collected.  Examples: '300ms', '10s' or '2h45m' (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)
--maximum-dead-containers int32                            Maximum number of old instances of containers to retain globally.  Each container takes up some disk space. To disable, set to a negative number. (default -1) (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)
--maximum-dead-containers-per-container int32              Maximum number of old instances to retain per container.  Each container takes up some disk space. (default 1) (DEPRECATED: Use --eviction-hard or --eviction-soft instead. Will be removed in a future version.)
I think the whole container GC configuration is deprecated in favor of the eviction feature and I'm not sure we should do anything here.

Yes, i was going through these values before. There was no information these parameters are direct related with container GC configuration. Is these are for container GC configuration?

[haircommander]

TBH I think we probably should drop these CLI flags now. We just hadn't gotten around to it yet, but they're marked as DEPRECATED. We should also update the documentation to deter folks from using them

/triage accepted
/priority important-longterm
/kind documentation

[haircommander]

/help

[k8s-ci-robot]

@haircommander:
This request has been marked as needing help from a contributor.

Guidelines

Please ensure that the issue body includes answers to the following questions:

• Why are we solving this issue?
• To address this issue, are there any code changes? If there are code changes, what needs to be done in the code and what places can the assignee treat as reference points?
• Does this issue have zero to low barrier of entry?
• How can the assignee reach out to you for help?

For more details on the requirements of such an issue, please see here and ensure that they are met.

If this request no longer meets these requirements, the label can be removed
by commenting with the /remove-help command.

In response to this:

/help

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[haircommander]

/assign @esotsal

[T-Lakshmi]

Raised PR in website repo for docs fix.
PTAL!

@haircommander @esotsal