Kevin Bucknum wrote:

The fastest any device will ever run on that channel will be whatever
the slowest device is. So even if you never use the CD/DVD drive, your hard drive will be limited to the speed of the CD/DVD drive.




I don't believe this is true any more.

hda - ATA133
hdb - ATA66
hdc - ATA133
hdd - PIONEER DVD-RW  DVR-105 (/dev/scsi/host1/bus0/target0/lun0/cd)

hdparm -t /dev/hda
/dev/hda:
Timing buffered disk reads:  124 MB in  3.03 seconds =  40.92 MB/sec

hdparm -t /dev/hdb
/dev/hdb:
Timing buffered disk reads:   38 MB in  3.14 seconds =  12.10 MB/sec

hdparm -t /dev/hdc
/dev/hdc:
Timing buffered disk reads:  126 MB in  3.02 seconds =  41.72 MB/sec

Peter
--



hmmm - I wonder if that is a recent change? I'm still on 2.4.20 and this is my hdparm -v -i

I think that changed some time ago (I think it was in the times of ATA33).


UDMA modes: udma0 udma1 udma2 udma3 *udma4 UDMA modes: udma0 udma1 udma2 udma3 *udma4 udma5

First is my 13 gig primary and second is my 30 gig secondary. I'll have to
triple check the cable when I get home to make sure that's not the bottle
neck.  I know the board will do ata/100 since I had that drive on the other
channel before I bought a dvd drive and it was doing it there.

I think the point her is that the slave cannot be faster than the master.

One note as to the way to connect your devices. I have 2 disks both ATA100, 1 DVD and 1 CD-RW (both ATA33)
In my former system I had the disks as masters on the 2 IDE channels and the DVD/CD as slaves. Now, due to cable length problems I have both disks on one channel and the same for DVD/CD.
Copying between both disks is noticably slower (due to the fact that IDE can not address both drives at the same time).
Before I had my partitions split , in a hopefully intelligent way, across both drives, this doesn't speed up my system now anymore.


Christian




-- [EMAIL PROTECTED] mailing list



Reply via email to