helm: support PodDisruptionBudget for resources

[a7i]

What type of PR is this?

/kind feature

What this PR does / why we need it:

Currently there's not a way to render PodDisruptionBudget for critical resources such as kube-apiserver

Which issue(s) this PR fixes:
Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

helm: support PodDisruptionBudget for resources

[RainbowMango]

/assign

Thanks @a7i I'll take a look ASAP.

[RainbowMango]

Does this mean all components will share the same budget?

[a7i]

Each resource has its own podDisruptionBudget configuration which can get an override. This was following the existing pattern of karmadaImageVersion on line 20 but happy to remove the anchor tag if you suggest though.

[RainbowMango]

Looks good to me.

[RainbowMango]

/lgtm
/approve

[karmada-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: RainbowMango

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

The pull request process is described here

Needs approval from an approver in each of these files:

• charts/OWNERS [RainbowMango]

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

[chaosi-zju]

hello @a7i, thank you for your important commit! Though the PR is merged, I wanna to bother you some questions:

1. you seems like added a empty PodDisruptionBudget config to chart template, why not add a config like minAvailable: 1? On the one hand it seems more meaningful, on the other hand it can provide a example to users, for users may not familiar with the format and rule of anchor.

2. didn't you test it by local helm install, and in your testing how did you configured the value.yaml?

Thank you again~

[RainbowMango]

why not add a config like minAvailable: 1? On the one hand it seems more meaningful,

I guess not all users need PDB, that's why take empty as default.

[a7i]

1. @RainbowMango is exactly right. If you have a single replica for example, minAvailable: 1 will block node rollout / cluster upgrades. As a result, PDB should only be defined for those who require high-availability and fault-tolerance.

2. It can be tested inline

  helm template host . --set controllerManager.podDisruptionBudget.maxUnavailable="20%" --set webhook.podDisruptionBudget.maxUnavailable="20%" --set apiServer.podDisruptionBudget.maxUnavailable="20%" --set kubeControllerManager.podDisruptionBudget.maxUnavailable="20%"

or by supplying your own values.yaml

[chaosi-zju]

ok, got it~