Andrew Lentvorski wrote:
> Neil Schneider wrote:
>> It's that simple, but I would do dd if=/dev/hda of=/dev/hdc so they
>> are both on different channels. If they are on the same IDE channel
>> it
>> will be really slow and take a long time.
>
>
>> It will already be slow because of them being IDE drives.
>
> Back up your statement, sir!
>
> Internal transfer rates max out at about 100 Megabytes/second. This
> maxes out at about 150 Megabytes/second for anything reasonable.
>
> The defining characteristic: RPM. It doesn't matter whether the
> interface is SCSI or IDE, the difference is all in the spindle speed.
Well that may be what the specs say, but real life experience says
different. In my experience, transfering IDE drives using dd on the
same controller will take overnight to complete, and there may be
multiple errors. Comparable transfers using scsi will take hours.
Drive size and speed all enter into the equation.
> Even a parallel ATA interface is quite capable of keeping up with all
> but the newest and fastest drives.
Interesting, because in testing we did running VPN across "the
internet" (actually a crossover cable) we could saturate a 100MB
connection with SCSI drives on the receiving side. Not true with IDE
drives. But these were real life tests, not theoretical.
> If the system has a serial ATA interface, then it is capable of
> handling
> effectively *all* existing drives just fine.
Speed continues to improve but the drive sizes increase at the same
time. I can't prove what I wrote, I can only report my experience.
As always, YMMV.
--
Neil Schneider pacneil_at_linuxgeek_dot_net
http://www.paccomp.com
Key fingerprint = 67F0 E493 FCC0 0A8C 769B 8209 32D7 1DB1 8460 C47D
The more you read and observe about this Politics thing, you got to
admit that each party is worse than the other. The one that's out
always looks the best. - Will Rogers
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
