Filesystem issues should be treated as array failures

[spamalam]

I've recently discovered that if there is anything other than a hardware failure, such as XFS filesystem failing, unraid silently ignores this and files dissappear from the array:

I believe that unraid should really treat this no differently to a hardware issue and the apps and array should be stopped if the filesystem goes into read/write failure or input/output error.

 

In the linked example, this caused cascading deletes across an entire cluster of servers because files disappeared from the master share.