Re: Question on ultra160 hard drives (performances, pinout...)
- From: "Folkert Rienstra" <see_reply-to@xxxxxxxx>
- Date: Tue, 20 Sep 2005 00:42:37 +0200
"Dr. Anton T. Squeegee" <SpammersAreVermin@xxxxxxxx> wrote in message news:MPG.1d9513d08b5dacbb98969f@xxxxxxxxxxxxxx
> In article 432b41a7$0$7922$892e7fe2@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx, see_reply-to@xxxxxxxx says...
>
> <snippety>
>
> > Nonsense, you don't know that since he never mentioned any part number.
>
> I see you're still working on those diplomacy skills, Folkert. ;-)
Yes, I am.
>
> You have a point. I would probably have been more accurate in
> saying that "many such drives under the Compaq label were designed for
> use in their servers." However, in my defense, the fact that the drive
> is labeled as 'Compaq,' and is an Ultra-160, makes it pretty likely to
> my mind that it's a retired server drive.
Of course, since it is SCSI, not IDE.
Just not for what you said in the part that you so conveniently snipped.
One server is not an other server. Drives in a small office desktop server
may be different from drives used in a mainframe type configuration.
And even 'stock' SCSI drives from 'Seagate, IBM, or whoever' may be setup differently from oneanother.
>
> > > As such, their firmware is different from stock Seagate, IBM, or whoever
> > > made the actual drive. While it may work, it may not have the same level of
> > > performance that a stock off-the-shelf drive may have.
> >
> > Which usually can be remedied with changing a few settings.
>
> "A few settings" where? And what specific settings are you referring to?
Cache management mainly. Bad sector management, possibly.
The settings that are a no-no for servers but essential for desktop use.
>
> > > SCSI truly shines in the area of supporting multiple
> > > devices, and multiple types of devices, on one bus.
> >
> > Nonsense.
> > That 's not a positive for SCSI, that is a negative for IDE which is
> > easily circumvented. SCSI shines on random access performance
> > which helps on systems that do parallel IO, like WinXP.
>
> Excuse me? Since when has multiple devices and multiple types of
> devices NOT been a positive for SCSI?
Only in theory.
>
> I don't understand what you're referring to with "easily
> circumvented" either. What, exactly, is "easily circumvented?"
One device per channel, buy more channels (if not already available
on the MB). IDE channels are very cheap, SCSI channels are not.
A scsi channel only supports 4 devices simultaniously anyway.
For non-simultaniously used devices you can use more but then
that also applies to IDE (2 devices per channel).
>
> On SCSI's random access performance: That's a given to my mind.
> That's why I didn't mention it.
How strange then that you didn't see that as a positive for a single drive.
>
> As for 'U-IDE,' I was referring to 'UltraIDE'
That is known as Ultra-ATA in Trade quarters.
> (I think I got that name right).
You could have checked.
> This is the one that uses an 80-conductor cable, where each
> line is paralleled (sp?) with a corresponding ground.
That's the ones that support UDMA modes over mode 2.
>
> Keep the peace(es).
Nope, I glue them together and hand them back again.
.
- References:
- Re: Question on ultra160 hard drives (performances, pinout...)
- From: Dr . Anton T . Squeegee
- Re: Question on ultra160 hard drives (performances, pinout...)
- From: Folkert Rienstra
- Re: Question on ultra160 hard drives (performances, pinout...)
- From: Dr . Anton T . Squeegee
- Re: Question on ultra160 hard drives (performances, pinout...)
- Prev by Date: Re: More on SCSI cables
- Next by Date: Hopefully simple SCSI problem with 8 bay CDR Tower
- Previous by thread: Re: Question on ultra160 hard drives (performances, pinout...)
- Next by thread: Re: Question on ultra160 hard drives (performances, pinout...)
- Index(es):
Relevant Pages
|
