kube-ovn-controller: fix workqueue metrics

[zhangzujian]

• Make sure you have followed Kube-OVN Code Style.

What type of this PR

• Bug fixes

Which issue(s) this PR fixes:

Do not import k8s.io/component-base/metrics since it sets the default workqueue metrics provider.

$ go mod why k8s.io/component-base/metrics                                                                            ✔  02:33:49 PM 
# k8s.io/component-base/metrics
github.com/kubeovn/kube-ovn/cmd/controller
k8s.io/apiserver/pkg/server
k8s.io/apiserver/pkg/audit
k8s.io/component-base/metrics

WHAT

🤖 Generated by Copilot at a09280e

Refactor workqueue creation in the controller. Extract common logic to a new function newNamedRateLimitingQueue in pkg/controller/controller.go.

🤖 Generated by Copilot at a09280e

Oh, we're the brave controllers of the Kubernetes fleet
We manage all the workqueues with our code so neat
We've refactored and improved the newNamedRateLimitingQueue
So pull the line and sing along, we'll make it run like new

HOW

🤖 Generated by Copilot at a09280e

• Add a new function newNamedRateLimitingQueue to create workqueues with custom names, rate limiters, and metrics providers (link)
• Refactor the Run function to use newNamedRateLimitingQueue instead of workqueue.NewNamedRateLimitingQueue for creating the podQueue, subnetQueue, ipQueue, vpcQueue, serviceQueue, and endpointQueue workqueues (link)