IDE-SCSI RAID units

[Steffen Persvold]

Josip Loncaric wrote:
> 
snip...
> 
> Finally, some performance numbers: our 1999 vintage SCSI drives deliver
> read performance of ~19 MB/s, which software RAID0 improves to ~36 MB/s
> (using two drives) and ~47 MB/s (using three drives).  Similarly,
> software striping two 1999 vintage IDE drives (attached as master and
> slave to IDE0) increases read rates from ~13 MB/s to ~26 MB/s (but
> commodity IDE would be less reliable than high end SCSI).  The CPU load
> during these 'hdparm -t' tests appears to be about 7% (SCSI or IDE).
>

I have tested some Quantum 72GB Ultra160 10K disks in a software raid
environment. The mainboards I tested on was Supermicro 370DE6 with built-in
AIC7899 (Dual channel Ultra160). I got the following numbers with RAID0 (hdparm
-t) :

  1 drive on 1 channel   :   33 MB/sec
  2 drives on 1 channel  :   68 MB/sec
  3 drives on 1 channel  :  103 MB/sec
  2 drives on 2 channels :  108 MB/sec
  3 drives on 2 channels :  103 MB/sec

As you can see, using 2 drives on two separate channels (4 in total) and 3
drives on two separate channels (6 in total) doesn't give me much boost compared
to 3 drives on 1 channel. I am suspecting that the PCI bus is the bottleneck
here, because the 7899 chip is connected to the 32bit/33MHz onboard PCI bus.

It also seems like CPU intensive applications like burnP6 doesn't hurt the disk
I/O at all.

And BTW: I could still use my software RAID1 array while it was resyncing, isn't
this the case with hardware RAID1 ?

Best regards,
-- 
 Steffen Persvold                        Systems Engineer
 Email  : mailto:sp at scali.com            Scali AS (http://www.scali.com)
 Norway : Tlf  : (+47) 2262 8950         Olaf Helsets vei 6
          Fax  : (+47) 2262 8951         N-0621 Oslo, Norway

 USA    : Tlf  : (+1) 713 706 0544 (43)  Scali
                                         10500 Richmond Avenue, Suite 190
                                         Houston, Texas 77042, USA