I'm wondering which trigger this operation.

Could you please provide us more information about your cluster setup like: • Harvester version • cluster size • where did you find the error log Furthermore, it would be nice if you could generate a support-bundle file for us to look into.