On 06/02/2020 11:54, Scott wrote:
On Thu, 6 Feb 2020 11:16:49 +0000, Martin Brown
wrote:

On 04/02/2020 13:11, wrote:
Andrew wrote:
But since SSD are the storage eeuivalent of a Minister Of Defence
(here today, gone tomorrow), you back up the SSD on something
slightly more reliable, like a conventional hard disk (or 2,3)

Have you got any reference or citation to back this assertion up?

I think he is being unduly pessimistic. The odd early SDD had controller
faults that could lose all data. I have had one brick on me.

The one thing that does tend to happen differently with SSDs is that
with spinning rust you generally get plenty of warning of impending doom
with weird noises, slow downs, CRC errors and SMART diagnostics.

IME an SSD fails catastrophically with total loss of all data and
without any warning. Once bricked all content is long gone unless you
use very specialised recovery services. Spinning rust you have a fair
chance of getting most of the data back safely yourself (very slowly).

I thought the write function failed first and the read data remained
available to access. Is this totally wrong?

In an ideal world it might be true, but the only example of SSD failure
I have had the thing died so completely that no utilities could even
recognise it as a hard drive after it declined to boot up the machine.

--
Regards,
Martin Brown