Sata Performance

Mark Overby moverby at nvidia.com
Mon Jun 30 10:35:18 PDT 2008


Formatted message: <A HREF="r0806301_f.htm">HTML-formatted message</A>

You would need to go look at the implementation of Linux that you are
running on as these are all parameters that are not part of SCSI mode pages
as they are ATA controller specific, not command set specific, nor part of
any SCSI transport.
I¹d suggest checking with the maintainer of your SCSI->ATA storage driver.
On 6/30/08 6:34 AM, "Leonardo Sbaraini" <leonardo.sbaraini at orbisat.com.br>
wrote:
> * From the T10 Reflector (t10 at t10.org), posted by:
> * "Leonardo Sbaraini" <leonardo.sbaraini at orbisat.com.br>
> *
> Hi all, I'm list newbie.
> 
> Well, for a week I'm researching a way to boost sata data transfer.
> 
> Looking in http://linux.oreillynet.com/lpt/a/272 I found some
> parameters to change:
> 
>    * multcount: Short for multiple sector count. This controls how
> many sectors are fetched from the disk in a single I/O interrupt. When
>	   this feature is enabled, it typically reduces operating
> system overhead for disk I/O by 30-50%. On many systems, it also
> provides increased data throughput of anywhere from 5% to 50%.
> 
>    * I/O support: This is a big one. This flag controls how data is
> passed from the PCI bus to the controller. Almost all modern
> controller chipsets support mode 3, or 32-bit mode w/sync. Some even
> support 32-bit async.
> 
>    * unmaskirq: Turning this on will allow Linux to unmask other
> interrupts while processing a disk interrupt. What does that mean? It
> lets Linux attend to other interrupt-related tasks (i.e., network
> traffic) while waiting for your disk to return with the data it asked
> for. It should improve overall system response time, but be warned:
> Not all hardware configurations will be able to handle it.
> 
>    * using_dma: DMA can be a tricky business. If you can get your
> controller and drive using a DMA mode, do it. But I have seen more
> than one machine hang while playing with this option.
> 
> 
> I started using hdparm to get info:
> --------------------------
> hdparm /dev/hda
> 
> /dev/hda:
> multcount    =  0 (off)
> I/O support  =  0 (default 16-bit)
> ...
> --------------------------
> 
> I read http://sg.torque.net/sg/sdparm.html and the PDF "sbc3r15" wrote
> from T10 Vice-Chair Mark S. Evans and I still dont know wich parameter
> I need to modify with "sdparm" command.
> 
> Thanks in advance.
> *
> * For T10 Reflector information, send a message with
> * 'info t10' (no quotes) in the message body to majordomo at t10.org
> 
-----------------------------------------------------------------------------
------
This email message is for the sole use of the intended recipient(s) and may
contain
confidential information.  Any unauthorized review, use, disclosure or
distribution
is prohibited.	If you are not the intended recipient, please contact the
sender by
reply email and destroy all copies of the original message.
-----------------------------------------------------------------------------
------



More information about the T10 mailing list