optimize provider network

[fanriming]

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

What type of this PR

Examples of user facing changes:

• Bug fixes

Currently, the same provider-network resource will be updated by all kube-ovn-cni pods. When the number of nodes is large and there are several inconsistent states, update events will trigger all kube-ovn-cni to patch provider-network again and again. In this case, the apiserver will receive a large number of patch requests and the cpu will be high load, and the apiserver will become very slow to respond.

The state of provider-network does not need to be changed frequently. provider-network should be read-only to the kube-ovn-cni. Update operations of provider-network should be minimized to avoid conflicts and inconsistencies. The optimization is as follows:

1. Removes the code that performs updates to provider-network in kube-ovn-cni.
2. Write error messages to kube-ovn-cni pod annotations.
3. Optimize controller timing synchronization code to update provider-network status based on node and kube-ov-cni annotations.
4. Add the notReadyNodes field to the status for clearer display of configuration results.

[fanriming]

@zhangzujian It has been modified as suggested, please review it again, thanks.

[zhangzujian]

Please run go fmt ./... to fix gofmt issues.