>>106250359
>Then it randomly switched and a DIFFERENT drive started failing after the old one
Classic cascading array failure
3.5 years is all you can reliably expect from a drive
We replace all prod drives at 3 years
Since all the drives in an array are typically the same model and even the same batch / lot they will die sequentially
Have been experiencing the same thing with multiple clients.
>6 drive array
>1 drive dies, R&R
>2 more drives die the following week, R&R
>1 more drive dies the following week, R&R
> R&R last 2 old drives is scheduled

Bad cards throw errors on all the drives
Bad cables throw errors on all the connected drives