randrepeat false for ssds

randrepeat false for ssds

[Slow bucks]

Does it make sense to disable randrepeat on SSDs with the way data is 
written (always write to new block with empty pages)?

Re: randrepeat false for ssds

[Sitsofe Wheeler]

Depends on your objective (are you trying to dodge tricks your SSD
might do?) but bear in you can't control where the SSD ultimately
chooses to put the data or how much it chooses to stuff into each
erase block size. Additionally randrepeat only has an impact when an
entire run is repeated. Some modern SSDs do compression/de-duplication
so if you want to try defeat that you might want to look at the
refill_buffers/scramble_buffers parameters
(http://fio.readthedocs.io/en/latest/fio_doc.html#cmdoption-arg-refill_buffers
).

On 6 February 2017 at 15:54, Slow bucks <frankwhitebe@gmail.com> wrote:
> Does it make sense to disable randrepeat on SSDs with the way data is
> written (always write to new block with empty pages)?
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe fio" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



-- 
Sitsofe | http://sucs.org/~sits/

Re: randrepeat false for ssds

[Slow bucks]

My primary objective is to really have consistent results with the same 
jobfile with any ssd drive test.
It sounds like setting randrepeat to false makes sense for ssd drives

On 2017-02-06 20:14:18 +0000, Sitsofe Wheeler said:

> Depends on your objective (are you trying to dodge tricks your SSD
> might do?) but bear in you can't control where the SSD ultimately
> chooses to put the data or how much it chooses to stuff into each
> erase block size. Additionally randrepeat only has an impact when an
> entire run is repeated. Some modern SSDs do compression/de-duplication
> so if you want to try defeat that you might want to look at the
> refill_buffers/scramble_buffers parameters
> (http://fio.readthedocs.io/en/latest/fio_doc.html#cmdoption-arg-refill_buffers
> ).
> 
> On 6 February 2017 at 15:54, Slow bucks 
> <frankwhitebe@gmail.com> wrote:
>> Does it make sense to disable randrepeat on SSDs with the way data is
>> written (always write to new block with empty pages)?
>> 
>> 
>> --
>> To unsubscribe from this list: send the line "unsubscribe fio" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html