By default every Nx System has preconfigured rules "On Storage Failure" which will trigger actions "Show notifications to all users" and "Send email to admin user". Make sure that "admin" user has correctly assigned email address. Add more users if necessary.
When everything is set up correctly you'll see different emails/notifications in different cases.
Note! Nx Witness can only detect drives failures if these drives are used for recording.
Local drive case
The most likely scenario here is that users will receive email notifications saying "I/O errors occurred" right before the failure. After that it's expected that storage failure will be processed by the OS and it will become unavailable in Nx as well.
NAS(Network Attached Storage) drive case
When either NAS becomes inaccessible via network or hangs OS won't get information about this right away. The most typical configuration considers very long timeout before OS accepts the fact that failure happened. During that timeout Nx Witness will keep sending "write" commands and all of them will be stacked into a queue. Usually it takes server around 30-50 seconds to overflow the default queue limit.
Right after that Nx Server starts to generate tons of "HDD/SSD slow detected" errors. More specifically, by default alarm email will be generated and sent per each camera every 30 seconds.
As soon as OS gets information that NAS is no longer in a working state user will start getting "I/O errors occurred" emails.
RAID array failure(local or shared)
Nx Witness Server currently has no integrations with RAID controllers. So Nx VMS will keep recording to the RAID drive while it's possible. Nx Media Server will detect that something bad happened only when the read/write requests to the drive fail. At this point user will again start receiving "I/O errors occurred" messages.
E.g. if there's RAID5 array of 4 HDDs is set up user won't know anything about single HDD failure. But as soon as any of the rest drives fails the errors will be sent.