[lug] RH Linux download + hdparm

[D. Stimits]

John Karns wrote:
> 
> On Thu, 18 Oct 2001, Sexton, George said:
> 
> > Which brings up a really under appreciated point. Just because the interface
> > is rated at a certain speed it doesn't mean the drive can do anything like
> > that kind of speed on sustained reads or writes.
> >
> > A good example is Ultra-160 SCSI drives. Generally, the drives cannot read
> > or write more than 20-25MB second. Period. End of sentence. Look at the
> > sustained throughput rate specification if you want to find out what you can
> > realistically do with a drive.
> 
> My understanding is that the theoretical maximum bw on a PCI controller is
> around 33 MB/s, due to the design spec of the PCI bus itself.  Someone
> please correct me if I'm mistaken.  Thus the main advantage in U-160 SCSI
> drives is in a RAID configuration where the data passes from hdd to hdd
> via the controller.  My assumption is that in this situation, the SCSI
> controller would not need to use the mobo's DMA controller chip.  Again,
> someone please correct me if I'm wrong.  For this reason I figure I'm just
> as well off with the cheaper Ultra-Wide 40 MB/s drives.
> 
> ----------------------------------------------------------------
> John Karns                                        jkarns at csd.net
> 
> _______________________________________________
> Web Page:  http://lug.boulder.co.us
> Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug

I forget how it is figured out, but a regular PCI bus is 33 MHz, not 33
MB/sec. The bus is 4 bytes/octets wide (32 bit), so that means at least
4 * 33 = 132 MB/sec. In my case, I'm operating on a 66 MHz, 64 bit bus,
so the minimum rate would be 4 times that. U160 is nice even on a
regular 33 MHz PCI bus, but peak rates or the rate of a good raid
controller could saturate the bus under normal PCI.

D. Stimits, stimits at idcomm.com