All of lore.kernel.org
 help / color / mirror / Atom feed
* IDE drives
@ 2004-03-24 22:33 David Becker
  2004-03-24 22:49 ` Ian Pratt
  0 siblings, 1 reply; 4+ messages in thread
From: David Becker @ 2004-03-24 22:33 UTC (permalink / raw)
  To: xen-devel


I'm seeing quite slow IDE write performance on Xenolinux-1.2 
Native linux writes at 30MB/s to an ext2 fs, and Xenolinux DOM0 gets
about 2MB/s (this is not a virtual disk extent).   hdparm doesn't work
under xenolinux, but that is only part of the problem.  Native linux
writes at 6MB/s without any hdparm tuning.

The SCSI hosts seem fine.  native linux writes to scsi at 36MB/s and
xen at 28MB/s.

Is the IDE speed about what I should expect for xen-1.2 or did I screw
up my configuration?





-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

^ permalink raw reply	[flat|nested] 4+ messages in thread

* RE: IDE drives
@ 2004-03-24 22:45 Neugebauer, Rolf
  0 siblings, 0 replies; 4+ messages in thread
From: Neugebauer, Rolf @ 2004-03-24 22:45 UTC (permalink / raw)
  To: David Becker, xen-devel; +Cc: Neugebauer, Rolf


This seems quiet slow.

Can you check the xen boot message for a warning if PIO is used? That
might explain. I think there was something on the xen-devel list a while
back. There might be two mentioning of PIO (or at least I think that
used to be the case) one of them might be wrong. 

Rolf


> -----Original Message-----
> From: xen-devel-admin@lists.sourceforge.net [mailto:xen-devel-
> admin@lists.sourceforge.net] On Behalf Of David Becker
> Sent: 24 March 2004 14:34
> To: xen-devel@lists.sourceforge.net
> Subject: [Xen-devel] IDE drives
> 
> 
> I'm seeing quite slow IDE write performance on Xenolinux-1.2
> Native linux writes at 30MB/s to an ext2 fs, and Xenolinux DOM0 gets
> about 2MB/s (this is not a virtual disk extent).   hdparm doesn't work
> under xenolinux, but that is only part of the problem.  Native linux
> writes at 6MB/s without any hdparm tuning.
> 
> The SCSI hosts seem fine.  native linux writes to scsi at 36MB/s and
> xen at 28MB/s.
> 
> Is the IDE speed about what I should expect for xen-1.2 or did I screw
> up my configuration?
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/xen-devel



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: IDE drives
  2004-03-24 22:33 IDE drives David Becker
@ 2004-03-24 22:49 ` Ian Pratt
  2004-03-24 23:18   ` David Becker
  0 siblings, 1 reply; 4+ messages in thread
From: Ian Pratt @ 2004-03-24 22:49 UTC (permalink / raw)
  To: David Becker; +Cc: xen-devel, Ian.Pratt


> 
> I'm seeing quite slow IDE write performance on Xenolinux-1.2 
> Native linux writes at 30MB/s to an ext2 fs, and Xenolinux DOM0 gets
> about 2MB/s (this is not a virtual disk extent).   hdparm doesn't work
> under xenolinux, but that is only part of the problem.  Native linux
> writes at 6MB/s without any hdparm tuning.

Take a look at the boot messages -- I suspect Xen hasn't got a
proper driver for your IDE chipset and is falling back to PIO.

You could try porting the driver from Linux (which I expect will
be easy), or wait until the new IO stuff is ready.
 
> The SCSI hosts seem fine.  native linux writes to scsi at 36MB/s and
> xen at 28MB/s.

I'm surprised you're seeing such a drop. With our aacraid
PERC3/Di cards we see (MB/s):
xenolinux	Write 28.3	Read 46.7
linux		Write 29.2	Read 47.2

Are you sure you're using the same part of the disk? It could be
a driver version issue, but I find it surprising you're
seeing such a performance drop.


Cheers,
Ian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: IDE drives
  2004-03-24 22:49 ` Ian Pratt
@ 2004-03-24 23:18   ` David Becker
  0 siblings, 0 replies; 4+ messages in thread
From: David Becker @ 2004-03-24 23:18 UTC (permalink / raw)
  To: xen-devel


" Take a look at the boot messages -- I suspect Xen hasn't got a
" proper driver for your IDE chipset and is falling back to PIO.

Yep, you are right.  Xenolinux says PIO.  I'll add this chipset to my todo list.

    ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
    ServerWorks CSB5: IDE controller on PCI bus 00 dev 79
    ServerWorks CSB5: detected chipset, but driver not compiled in!
    ServerWorks CSB5: chipset revision 147
    ServerWorks CSB5: not 100% native mode: will probe irqs later
        ide0: BM-DMA at 0x0700-0x0707, BIOS settings: hda:DMA, hdb:DMA
        ide1: BM-DMA at 0x0708-0x070f, BIOS settings: hdc:DMA, hdd:DMA
    hda: ST340016A, ATA DISK drive
    hdc: LG CD-ROM CRN-8245B, ATAPI CD/DVD-ROM drive
    ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
    ide1 at 0x170-0x177,0x376 on irq 15
    hdc: ATAPI 24X CD-ROM drive, 128kB Cache
    Uniform CD-ROM driver Revision: 3.12
    hda: 78156288 sectors (40016 MB) w/2048KiB Cache, CHS=4865/255/63 PIO (slow!)


" Are you sure you're using the same part of the disk?

most likely not.  I ran the scsi writes on two different hosts just as a
sanity check, and not as a rigorous check.  Both were the same hardware
and may even have had other users on them.  It was close enough to be labeled
'the same' for my purposes.


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2004-03-24 23:18 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-03-24 22:33 IDE drives David Becker
2004-03-24 22:49 ` Ian Pratt
2004-03-24 23:18   ` David Becker
  -- strict thread matches above, loose matches on Subject: below --
2004-03-24 22:45 Neugebauer, Rolf

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.