Andrew Morton wrote:
Lincoln Dale <[EMAIL PROTECTED]> wrote:
sg_dd uses a window into a kernel DMA window. as such, two of the four
memory acccesses are cut out (1. DMA from HBA to RAM, 2. userspace
accessing data).
1.6Gbps / 2 = 800MB/s -- or roughly what Ian was seeing with sg_dd.
Right.
Andrew Morton wrote:
Lincoln Dale [EMAIL PROTECTED] wrote:
sg_dd uses a window into a kernel DMA window. as such, two of the four
memory acccesses are cut out (1. DMA from HBA to RAM, 2. userspace
accessing data).
1.6Gbps / 2 = 800MB/s -- or roughly what Ian was seeing with sg_dd.
Right.
Lincoln Dale <[EMAIL PROTECTED]> wrote:
>
> sg_dd uses a window into a kernel DMA window. as such, two of the four
> memory acccesses are cut out (1. DMA from HBA to RAM, 2. userspace
> accessing data).
> 1.6Gbps / 2 = 800MB/s -- or roughly what Ian was seeing with sg_dd.
Right. That's a
At 08:32 PM 4/02/2005, Andrew Morton wrote:
Something funny is happening here - it looks like there's plenty of CPU
capacity left over.
[..]
Could you monitor the CPU load during the various tests? If the `dd'
workload isn't pegging the CPU then it could be that there's something
wrong with the
On Thu, Feb 03, 2005 at 07:03:48PM +, Paulo Marques wrote:
> FYI there was a patch running around last April that made a new option
> for "dd" to make it use O_DIRECT. You can get it here:
>
> http://marc.theaimsgroup.com/?l=linux-kernel=108135935629589=2
>
> Unfortunately this hasn't made
Ian Godin <[EMAIL PROTECTED]> wrote:
>
>
>I am trying to get very fast disk drive performance and I am seeing
> some interesting bottlenecks. We are trying to get 800 MB/sec or more
> (yes, that is megabytes per second). We are currently using
> PCI-Express with a 16 drive raid card
Ian Godin [EMAIL PROTECTED] wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or more
(yes, that is megabytes per second). We are currently using
PCI-Express with a 16 drive raid card (SATA
On Thu, Feb 03, 2005 at 07:03:48PM +, Paulo Marques wrote:
FYI there was a patch running around last April that made a new option
for dd to make it use O_DIRECT. You can get it here:
http://marc.theaimsgroup.com/?l=linux-kernelm=108135935629589w=2
Unfortunately this hasn't made it
At 08:32 PM 4/02/2005, Andrew Morton wrote:
Something funny is happening here - it looks like there's plenty of CPU
capacity left over.
[..]
Could you monitor the CPU load during the various tests? If the `dd'
workload isn't pegging the CPU then it could be that there's something
wrong with the
Lincoln Dale [EMAIL PROTECTED] wrote:
sg_dd uses a window into a kernel DMA window. as such, two of the four
memory acccesses are cut out (1. DMA from HBA to RAM, 2. userspace
accessing data).
1.6Gbps / 2 = 800MB/s -- or roughly what Ian was seeing with sg_dd.
Right. That's a fancy way
Ian Godin wrote:
[...]
Definitely have been able to repeat that here, so the SG driver
definitely appears to be broken. At least I'm glad I am not going
insane, I was starting to wonder :)
I'll run some more tests with O_DIRECT and such things, see if I can
figure out what the REAL max
Ian Godin wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or more
(yes, that is megabytes per second). We are currently using PCI-Express
with a 16 drive raid card (SATA drives). We have achieved that
On Feb 3, 2005, at 9:40 AM, Nuno Silva wrote:
Ian Godin wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or
more (yes, that is megabytes per second). We are currently using
PCI-Express with a 16 drive
On Feb 2, 2005, at 7:56 PM, Bernd Eckenfels wrote:
In article <[EMAIL PROTECTED]> you
wrote:
Below is an oprofile (truncated) of (the same) dd running on
/dev/sdb.
do you also have the oprofile of the sg_dd handy?
Greetings
Bernd
Just ran it on the sg_dd (using /dev/sg1):
CPU: P4 / Xeon,
On Wed, Feb 02 2005, Ian Godin wrote:
>
> I am trying to get very fast disk drive performance and I am seeing
> some interesting bottlenecks. We are trying to get 800 MB/sec or more
> (yes, that is megabytes per second). We are currently using
> PCI-Express with a 16 drive raid card (SATA
On Wed, Feb 02 2005, Ian Godin wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or more
(yes, that is megabytes per second). We are currently using
PCI-Express with a 16 drive raid card (SATA
On Feb 2, 2005, at 7:56 PM, Bernd Eckenfels wrote:
In article [EMAIL PROTECTED] you
wrote:
Below is an oprofile (truncated) of (the same) dd running on
/dev/sdb.
do you also have the oprofile of the sg_dd handy?
Greetings
Bernd
Just ran it on the sg_dd (using /dev/sg1):
CPU: P4 / Xeon, speed
On Feb 3, 2005, at 9:40 AM, Nuno Silva wrote:
Ian Godin wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or
more (yes, that is megabytes per second). We are currently using
PCI-Express with a 16 drive
Ian Godin wrote:
I am trying to get very fast disk drive performance and I am seeing
some interesting bottlenecks. We are trying to get 800 MB/sec or more
(yes, that is megabytes per second). We are currently using PCI-Express
with a 16 drive raid card (SATA drives). We have achieved that
Ian Godin wrote:
[...]
Definitely have been able to repeat that here, so the SG driver
definitely appears to be broken. At least I'm glad I am not going
insane, I was starting to wonder :)
I'll run some more tests with O_DIRECT and such things, see if I can
figure out what the REAL max
In article <[EMAIL PROTECTED]> you wrote:
> Below is an oprofile (truncated) of (the same) dd running on /dev/sdb.
do you also have the oprofile of the sg_dd handy?
Greetings
Bernd
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL
In article [EMAIL PROTECTED] you wrote:
Below is an oprofile (truncated) of (the same) dd running on /dev/sdb.
do you also have the oprofile of the sg_dd handy?
Greetings
Bernd
-
To unsubscribe from this list: send the line unsubscribe linux-kernel in
the body of a message to [EMAIL
22 matches
Mail list logo