Add alert checking for Disk I/O or Disk status

This week, we had a issue where, in one of our ceph nodes, a disk went offline and we only noticed way to late. Looking at our monitoring/Observium, we realized that the disk going offline wasn't really being registered anywhere. The quick fix we want to try is to create a alert for zero disk I/O, which should not occur on a online disk.

However, It seems that there is no way to create an alert for Disk I/O?

Any better alternative to catch such an event (/dev/sda going offline) would be appreciated as well.