Based on Jon Gardner's post in response to my initial query about 
a speed discrepancy between a DDS-2 drive on a slow SCSI external bus 
and and the VXA drive on the Fast Internal bus (on a PowerTower Pro 

>Any I/O bus, be it SCSI, network, or whatever, has a limit to the amount of
>data that can be passing through it at any given moment. This is a bit of a
>simplification, but if you're trying to read data at full speed from your
>HD, you cannot simultaneously write at full speed to a tape drive on the
>same bus, because the bus is saturated with the read. If the tape drive is
>on another bus, you can read at full speed on one bus and write at full
>speed on the other bus at the same time.

I was able to resolve the slow down problem by putting the VXA drive 
on its own SCSI Card (Adaptec 2930 card, although I think in 
principle the less expensive 2906 card would have solved the problem 
as well). After moving the VXA drive to the 2930 card speed of backup 
went from average of 40 MB/min to 133 MB/min. In normal use I was 
seeing backup speeds approaching 180 MB/min across the network when 
backing up G3 clients (during the compare phaste).

I also did a brief test. At first the VXA drive was the only device 
on the card and I did a test back up of an external hard drive and 
got around 133 MB/min, then moved that external drive to the external 
port of the 2930 card and backed up the same files and did indeed get 
slightly lower numbers (although still perfectly acceptable).

So in conclusion, the VXA drive (when installed internally) works 
fastest when it is not on the same bus as the boot drive and if 
you're concerned with absolutely the highest speed in backing up 
local devices, it should be on it's own card all by itself... if 
you're only concerned with maximum network backup speeds it should 
just be on a different (preferably Fast SCSI 2 or faster) bus from 
the boot drive.

I'd also like to acknowledge the excellent customer service from 
Ecrix on this matter. I received rapid responses to my queries and 
Bob Zegarelli even sent me an active inline terminator so that I 
could properly test the drive with the 2930 card.

Thanks to everyone,
-Greg Morin

Gregory Morin, Ph.D.  ~~~~~~~IS Manager~~~~~~~~~~~~~~~~~~~
Seachem Laboratories, Inc.      www.seachem.com     888-SEACHEM

To subscribe:    [EMAIL PROTECTED]
To unsubscribe:  [EMAIL PROTECTED]
Archives:        <http://list.working-dogs.com/lists/retro-talk/>
Problems?:       [EMAIL PROTECTED]

Reply via email to