My experience with HDDs over the last decade is that I rarely see them dying without warning, typically there are early indicators in the SMART data or failures in self tests. WD in North America has always exchanged drives with these symptoms for me, if still in warranty, without any hassle or issues. As I use their service to send out the replacement drive first, I usually can replace the failing drive at my leisure and have time to wipe it before returning it.
SSDs typically die on me in much greater numbers than HDDs and without any warnings. Great SMART data today, unreadable tomorrow. Brand does not appear to matter that much either. I had SSDs break from Adata, Intel, WD and more. Intel has been reasonable with warranty exchanges, WD great. Adata SSDs always broke just outside of warranty, I don't believe I have any survivors left.
I am not worried about data on the drives, as everything is encrpyted.
In any case, if at all possible, I run everything in some kind of RAID configuration. Preferably the system will run FreeBSD or be based on it and thus will run a ZFS mirror for boot SSDs or RAIDZ for storage drives (SSD/HDD),
Just last month, my pfsense router at home died with a failed SSD (some cheap off-brand Chinese SSD that came with the hardware), unfortunately the hardware supports only one M.SATA (not M2) SSD, so no mirror here, The new hardware has two M2 SSDs in a ZFS mirror (and will run OPNsense).
