Search the metrics should be added for Linux file system

[J-lena]

Is your feature request related to a problem? Please describe.
Pool application performance can cause by several reasons: network jitter, application complex process logic, operator system jitter, and so on. Now kindling provides lots of network metrics to help to identify network problems and other valuable features also, but now it lacks metrics to identify file system problems and for example, can not lead us to identify if slow i/o device results in pool application performance.

Describe the solution you'd like
Hoping kindling will support more metrics about the Linux file system including VFS and block I/O performance.