[kwlug-disc] SMART for SSDs

[Khalid Baheyeldin]

On Fri, Jul 31, 2015 at 12:23 PM, B.S. <bs27975 at yahoo.ca> wrote:

> The point taken from the phrasing of your message, though, is ... THERE'S
> A KERNEL BUG IN THE SSD CODE ... OMG my drive is about to die!!!!!
>

> Evidently not what you intended, but how I took it ... (-:
>

No, it is mainly about the blacklisting of some models, and conflicting
information re: discard or fstrim with cron, ....etc.

*    Data Set Management TRIM supported (limit 1 block)
>

I have that line, but with 16 blocks instead of 1.


> *    Deterministic read data after TRIM
>

I don't have the above line though.


> >This only runs on Intel and Samsung SSDs as some SSDs ...
>
> Yes, as your links note, there is some sensitivity to models, and settings
> changes made when found. Didn't know that, thank you. Good to know.
>

With that info, one has to wonder if fstrim-all does anything at all. This
is the default way Ubuntu 14.04 trims SSDs. It may run and does nothing at
all on non-Samsung/Intel SSDs ...


> Mind you ... does one typically have to think about such things in a cloud
> server? Isn't the hardware / uptime / trim / so on their problem that you
> have no control over?
>

Most hosting providers will be using enterprise class drives, usually from
Samsung or Intel. So those will be supported and better used on Linux.

We are using consumer stuff.

I gathered some courage, took a deep breath, and ran fstrim manually, using:

# time fstrim -v /
/: 219199045632 bytes were trimmed

It took 20.5 seconds.

-- 
Khalid M. Baheyeldin
2bits.com, Inc.
Fast Reliable Drupal
Drupal optimization, development, customization and consulting.
Simplicity is prerequisite for reliability. --  Edsger W.Dijkstra
Simplicity is the ultimate sophistication. --   Leonardo da Vinci
For every complex problem, there is an answer that is clear, simple, and
wrong." -- H.L. Mencken
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://kwlug.org/pipermail/kwlug-disc_kwlug.org/attachments/20150731/102b7b07/attachment.htm>