> You will find that the limit of the disk access is not whether you are
> using udma66 or udma33 or scsi but the maximum data rate on for the disk
> drive.
that's the case for older-model disks, but not current generation ones
(15G/platter, especially if 7200rpm, which deliver ~32-35 MB/s.)
> over that past year with all the disk drive problems that I have read
> and helped with, I don't use anything but ibm drives now). Neither the
I have had excellent experience with Maxtor and IBM disks. I stay
away from WD for everything. older Samsung, NEC and Fujitsu's were
flakey, though have improved recently. Seagate IDE's have a reputation
for not playing well with other drives. Quantum has hit some foul balls,
too.
> drive, if you can find it on maxtor's website, you will be delighted to
> discover that you have deluded yourself (this is not an insult but a
I'm not sure what delusion you're talking about; there's nothing like that
in the message.
> performance. And until the udma66 drivers are completely debugged for
> linux, you will have a lot more reliable system if you stick with the
> udma33 interface for now.
I see no reason to believe this: udma66 is a minor change from udma33,
especially on well-established controllers like the Intel PIIX line.
> > /var/log on hda too). The fs I use is reiser-fs, except for /boot which is
> > ext2 (as said necessary in the manual).
you should definitely be looking for reiser updates, then.
> > The problem is that if I set the dma on the ide hard disks using
> >
> > hdparm -d1 /dev/hda
modern (2.4) kernels don't need this kind of silliness.
> > If I remove the udma setting, everything works fine.
this doesn't implicate udma.
> > What can I do ? Buy a bunch of SCSI ?
grotesque. you system has shown a problem, probably one of the many
kernel/reiserfs bugs. it is bizarre to blame it on udma.
regards, mark hahn.
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/dmentre/smp-howto/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
