Check whether the kubelet on the node is running properly.
If the kubelet malfunctions, the node is unavailable. Restore the node and check again. For details, see What Should I Do If a Cluster Is Available But Some Nodes Are Unavailable?
The version of the pause container image on which kubelet depends is not cce-pause:3.1. If you continue the upgrade, pods will restart in batches. Currently, the upgrade is not supported. Contact technical support.