Good SSDs fails far less often than mechanical HDD. In 2 years I had only 1 drive fail (100 deployed devices - it was a firmware related failure).
There is a popular alternative solution: install a second, independent drive and schedule a periodic rsync from the primary one. If primary drive fails, there will be some data loss and downtime. But:
- no hw or sw RAID is involved (one less thing that can go wrong).
- to recover the system turn off the server, pull out both drives, put the secondary drive in the first drive slot, turn on the system again and do a few post-boot adjustements. Total downtime is short. Data loss may be not so nasty afterall (most contents does not change from one rsync to the next). No sw RAID issues due to half-dead disks - after a drive starts to fail, it is removed from the system.
- the secondary drive doubles as "poor's man backup" solution (provided you schedule the rsync once or twice a day). It is NOT a real backup, of course, but it will catch some common issues (fat finger errors etc.)
I was sceptical of this arrangement at first, my initial thought was the same as @
Munzy. But, but... It is common now on industrial control systems, and the RAID card failure rate is almost the same as a SSD drive, so putting two SSDs on HW RAID 1 to enhance reliability does not make a lot of sense anyway.