[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: disk i/o performance problem



On Sat, Jun 02, 2007 at 12:08:06PM -0400, Douglas Allan Tutty wrote:
> On Sat, Jun 02, 2007 at 11:58:00AM -0400, Roberto C. Sánchez wrote:
> > On Sat, Jun 02, 2007 at 10:54:51AM -0400, Douglas Allan Tutty wrote:
> > > On Thu, May 31, 2007 at 11:02:06PM -0400, WHIRLYCOTT wrote:
> > > > Etch: (output of hdparm)
> > > > 
> > > > 	/dev/sda:
> > > > 	Timing cached reads:   1588 MB in  2.00 seconds = 794.41 MB/sec
> > > > 	Timing buffered disk reads:  176 MB in  3.03 seconds =  58.15 MB/sec
> > >  
> > > > CentOS: (output of hdparm)
> > > > 
> > > > 	/dev/sda:
> > > > 	Timing cached reads:   4480 MB in  2.00 seconds = 2242.86 MB/sec
> > > > 	Timing buffered disk reads:  176 MB in  3.03 seconds =  58.16 MB/sec
> > > 
> > > The buffered disk reads are the same since that's the transfer rate from
> > > the drive's buffer.  The problem is the cache reads which is the time
> > > from you OS's cache.  
> > > 
> > I thought that the cached reads were from the disk drives onboard cache.
> > 
> 
> I could certainly be wrong, but that would mean that getting data from
> the drive cache was at 17 and 38 times faster than getting data from the
> OS (?) buffer and that getting such data was the same on both OSs
> (unlikely).  
> 
> I don't have hdparm installed to be able to check the man page.
> 
> If cached reads are from the drive's cache then it would suggest that
> the different OSs are able to configure the drive's firmware
> differently; 
> 
I guess I was wrong:

-T     Perform timings of cache reads for benchmark and comparison pur-
       poses.   For  meaningful  results,  this  operation  should   be
       repeated  2-3  times  on  an otherwise inactive system (no other
       active processes) with at least a couple of  megabytes  of  free
       memory.   This  displays  the speed of reading directly from the
       Linux buffer cache without disk  access.   This  measurement  is
       essentially  an  indication  of the throughput of the processor,
       cache, and memory of the system under test.  If the -t  flag  is
       also specified, then a correction factor based on the outcome of
       -T will be incorporated into the  result  reported  for  the  -t
       operation.

Regards,

-Roberto

-- 
Roberto C. Sánchez
http://people.connexer.com/~roberto
http://www.connexer.com

Attachment: signature.asc
Description: Digital signature


Reply to: