Re: standard performance (write speed 20Mb/s)

On 7/27/2011 5:26 AM, John Robinson wrote:

On 27/07/2011 11:22, Stan Hoeppner wrote:

quoted

On 7/27/2011 12:42 AM, Simon Matthews wrote:

quoted

On Sun, Jul 17, 2011 at 5:11 AM, John Robinson
[off-list ref]  wrote:

quoted

Pretty poor. CentOS 5, Intel ICH10, md RAID 6 over 5 7200rpm 1TB
drives,
then LVM, then ext3:
# dd if=/dev/zero of=test bs=4096 count=262144
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 2.5253 seconds, 425 MB/s

What hard drive offers a sustained data rate of 425 MB/s or even half
that?

425 MB/s / 3 spindles = 142 MB/s per spindle

That's not poor, it's excellent.  Which drives are these?  WD Black,
Seagate, Hitachi?

Gentlemen, we've been round this loop before about 10 days ago. Pol's 20
MB/s was poor because he was testing on an array with unaligned
partitions and a resync was running, my 425 MB/s was a bad test because
it didn't use fdatasync or direct and I said dd was a bad test anyway,
etc etc.

Of course the 425 number is higher than real world.  But this is the dd
"test" method for which many post results, so the number is still
somewhat valid for very coarse comparison purposes among systems.  A
10GB run would have been better than 1GB obviously.

The Linpack benchmark is very similar to dd in this regard.  It's a
horrible measure of supercomputer performance, but it's what everyone
uses.  It's what allows US Government labs to go to Congress with and
say "The Chinese have passed us in the Supercomputer race.  We need more
money to take the lead".  (And, stupidly, Congress gives them the money)

-- 
Stan