optimize: preserving needed fields for memory efficiency

[Colvin-Y]

fix #1737
In production scenarios, the fields of pods are usually quite complex. Testing has shown that in our case, the descheduler's memory usage has been reduced by 2/5.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign a7i for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[k8s-ci-robot]

Welcome @Colvin-Y!

It looks like this is your first PR to kubernetes-sigs/descheduler 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes-sigs/descheduler has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[k8s-ci-robot]

Hi @Colvin-Y. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Details

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.

[googs1025]

/ok-to-test

[googs1025]

Using SetTransform seems to be very helpful for memory optimization, especially in large clusters. We can delete the unused pod node field. 🤔

[googs1025]

/lgtm

cc @ingvagabund

[a7i]

should this move here instead?

[Colvin-Y]

should this move here instead?

ditto

[k8s-ci-robot]

New changes are detected. LGTM label has been removed.

[a7i]

Nice change @Colvin-Y - do you have any before/after results on memory footprint you can share?

@ingvagabund How does a global change like this impact a plugin? For example, if I write my custom Descheduler plugin which relies on some of the omitted fields, can I override this behavior?

[Colvin-Y]

不错的改变@Colvin-Y - 您能否分享关于内存占用的前后结果？

hi @a7i ,
In our prod env, save 47% memory on avg

[Colvin-Y]

If I write my custom Descheduler plugin which relies on some of the omitted fields, can I override this behavior?

hi @a7i

Before cache start, you can override it. But maybe best practice is to modify func preserveNeeded because there are many plugins but using same cache.

[ingvagabund]

Each plugin might tell which fields are necessary. I.e. through a new method in the similar fashion as scheduler's EventsToRegister. E.g. FieldsToPreserve or similar:

func (d *PodLifeTime) FieldsToPreserve() ([]fwk.ObjectFields, error) {
	return []fwk.ObjectFields{
		{Resource: fwk.Pod, Fields: {...}},
		{Resource: fwk.Node, Fields: {...}},
	}
}

This might be tricky to implement as each plugin will need different fields. So a simple clearing functions as in this PR will not suffice.

[Colvin-Y]

/retest-required

[ingvagabund]

/retest-required

[ingvagabund]

A naive solution would be to enumerate all fields. E.g.:

c.Command = nil
c.Args = nil
c.WorkingDir = ""
c.Ports = nil
...

into

const (
  POD_COMMAND = iota
  POD_ARGS
  POD_WORKING_DIR
  POD_PORTS
  ...

The same for nodes. Then, have each plugin to define a list of fields to preserve:

func func (pl *Plugin) PreservedField() [] uint {
  return []uint {POD_COMMAND, POD_ARGS, ...}
}

Then iterate through all plugins, make a union of all PreservedField() lists and use the list to tell which fields can be cleared:

if _, in := preservedFieldsMap[POD_COMMAND]; !in {
  c.Command = nil
}
if _, in := preservedFieldsMap[POD_ARGS]; !in {
  c.Args = nil
}
...

[Colvin-Y]

A naive solution would be to enumerate all fields. E.g.:

c.Command = nil
c.Args = nil
c.WorkingDir = ""
c.Ports = nil
...

into

const (
  POD_COMMAND = iota
  POD_ARGS
  POD_WORKING_DIR
  POD_PORTS
  ...

The same for nodes. Then, have each plugin to define a list of fields to preserve:

func func (pl *Plugin) PreservedField() [] uint {
  return []uint {POD_COMMAND, POD_ARGS, ...}
}

Then iterate through all plugins, make a union of all PreservedField() lists and use the list to tell which fields can be cleared:

if _, in := preservedFieldsMap[POD_COMMAND]; !in {
  c.Command = nil
}
if _, in := preservedFieldsMap[POD_ARGS]; !in {
  c.Args = nil
}
...

ditto

[fanhaouu]

I suggest not removing this c.Ports. The NodePorts plugin in the scheduler needs to use the container’s ports, and the related field values may be required during the node fit phase.

[Colvin-Y]

ack, I need some time to refact all as

func func (pl *Plugin) PreservedField() [] uint {
  return []uint {POD_COMMAND, POD_ARGS, ...}
}

[k8s-ci-robot]

PR needs rebase.

Details

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.

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all PRs.

This bot triages PRs according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the PR is closed

You can:

• Mark this PR as fresh with /remove-lifecycle stale
• Close this PR with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

[k8s-ci-robot]

@Colvin-Y: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
pull-descheduler-test-e2e-k8s-master-1-35	582a574	link	true	/test pull-descheduler-test-e2e-k8s-master-1-35
pull-descheduler-test-e2e-k8s-master-1-36	582a574	link	true	/test pull-descheduler-test-e2e-k8s-master-1-36

Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR.

Details

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. I understand the commands that are listed here.