Primary node: "STORAGE [xxxx] Failed to query filesystem disk stats (code: 2): No such file or directory"

We are now using PMM server with version 2.11.1, and pmm-agent with also 2.11.1. The log of 'STORAGE [xxx] Failed to query filesystem disk stats (code: 2): No such file or directory' only occurs on primary while enable monitoring on mongos, nothing unusual on secondary. When i remove monitoring from mongos, the logs disappear.

Besides, in the /var/log/message, something to be improved:

pmm-agent: #033[36mINFO#033[0m[2020-12-03T14:54:16.384+08:00] time="2020-12-03T14:54:16+08:00" level=error msg="cannot get replSetGetStatus: replSetGetStatus is not supported through mongos" #033[36magentID#033[0m=/agent_id/b050cc82-50b6-477a-a39c-bf4fcb0dcd8b #033[36mcomponent#033[0m=agent-process #033[36mtype#033[0m=mongodb_exporter

Any explanation here?