SSD Endurance

[Matthew Dillon]

Not necessarily.  Data degradation is a natural process so it depends what
the controller decides to do with it.  It will probably reallocate the
block and bump the Reallocated_Event_Count counter in the smart status, but
that doesn't mean that the SSD is going bad.  If the counter starts
increasing regularly though that could be an indication of future problems
brewing.

The biggest indicator of problems is the Offline_Uncorrectable counter...
that means data went bad that the SSD controller couldn't recover.  Flash
wear characteristics are such that controllers are supposed to be able to
detect and rewrite weak data before it actually goes bad so if data does
actually go bad it's probably a good idea to replace the SSD.

-Matt

On Thu, Mar 12, 2015 at 2:00 PM, Carsten Mattner <carstenmattner at gmail.com>
wrote:

> On Thu, Mar 12, 2015 at 9:42 PM, Matthew Dillon
> <dillon at apollo.backplane.com> wrote:
> >     When left powered, SSDs do internal scans and will slowly fix up
> cells
> >     before the data becomes unrecoverable.
>
> Does that mean those cells are marked dead?
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dragonflybsd.org/pipermail/users/attachments/20150312/1dbe19c0/attachment-0002.htm>