On Wed, 5 Apr 2000, Paul Barton-Davis wrote:
>>having multiple scsi bus's and segreatting the scsi h/w according to
>>speed/capabilities, to the different scsi chains, is a very good idea.
...
>But its not clear from the mobo docs, or from just staring at the mobo
>or from adaptecs web site, whether the U2/LVD connector on the mobo
>really represents its own "chain" or just a point to branch a U2/LVD
>cable from the one single chain thats connecting everything.
There is a special purpose chip that gates the non-LVD devices onto the LVD
bus and then to the controller. It is required to use LVD and non-LVD
devices at their full speeds -- assuming you plug everything onto the correct
connector.
I'm a little surprise Supermicro isn't proclaiming this loudly. That isolation
chip isn't free by a long shot.
>understood. its pretty dismal. some people have claimed to have gotten
>twice this performance from udma drives that cost probably 50% less. phooey.
The UDMA drives have newer servo hardware than your drives. But don't feel
too bad; their speed drops dramatically as the number of requestors goes
up. (Run hdparam once and it's impressive; run it 4 times in parallel, as
you'll see what I mean.) Also, I'd bet the UDMA drive was the only thing
on that IDE cable/bus; you can get full speed from SCSI drives up to bus
saturation.
--Ricky
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 21:00:17 EST