It depends on the hardware that you use.
DELL has iDrac which is a helpful management platform for server engineers. With it you can check all faults, hardware error messages and overall health of the server.
IPMIview, HP iLo and OpenBMC are technologies similar to iDrac that can be used to monitor the server's overall health. Other than that you can check the server physically for any red or amber lights on the hard drives depicting their failure state.
RAID is only about uptime, nothing else. No longer about performance because SSDs, unless maybe you're talking hundreds of terabytes.
When a drive fails it starts hammering the other drives to do a rebuild, if any of those drives suffers a URE (the chances of which are quite high on a large drive) it fails that drive and then hammers the remaining drives twice as hard to rebuild, which greatly increases the chances of hitting another URE. If it does hit another URE that drive also fails, and the entire array fails.
Unless you're using all SSD drives, do NOT use RAID5 or RAID6. It's not worth the risk of losing the entire array during a rebuild. RAID10 should be your default.
