RE: filmscanners: Second Hard Drive/Umax scanner

2000-12-14 Thread Mark Edmonds

Because, (if my understanding is correct), the RAID card improves over
normal SCSI error correction (1 bit) and ups this to 4 bit. It is nothing to
do with RAID parity.

Mark

> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Bill Ross
> Sent: 11 December 2000 20:50
> To: [EMAIL PROTECTED]
> Subject: RE: filmscanners: Second Hard Drive/Umax scanner
>
>
>   ... The RAID port is populated with an Adaptec ARO1130U2 RAID
>   card on which hang two IBM Ultrastar 18ES 7200rpm 18GB drives
>   running a RAID0 (striped, no parity) array.
>
>   ... Clearly, writing to SCSI outperforms IDE and has the added
>   advantage in that the error correction is better as the RAID
>   controller error correction is better than a standalone drive
>
> How could the RAID controller could improve error correction
> w/out parity?
>
> Bill Ross




RE: filmscanners: Second Hard Drive/Umax scanner

2000-12-11 Thread Bill Ross

... The RAID port is populated with an Adaptec ARO1130U2 RAID 
card on which hang two IBM Ultrastar 18ES 7200rpm 18GB drives 
running a RAID0 (striped, no parity) array. 

... Clearly, writing to SCSI outperforms IDE and has the added 
advantage in that the error correction is better as the RAID 
controller error correction is better than a standalone drive 

How could the RAID controller could improve error correction
w/out parity?

Bill Ross



RE: filmscanners: Second Hard Drive/Umax scanner

2000-12-11 Thread Mark Edmonds

> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Mike Kersenbrock

> P.S - Magazines like P.C. Magazine has done benchmarking of
> servers using IDE vs
>   SCSI disks, and I recall their conclusions to be that they
> were very surprised to
>   find that it didn't make much difference in the actual
> system performance.

As a newbie to the list, this is my first post but maybe I can add some
first hand experience and opinions to this debate...

The system I run is as follows:

The motherboard is dual PIII with built in U2W SCSI and RAID port. The RAID
port is populated with an Adaptec ARO1130U2 RAID card on which hang two IBM
Ultrastar 18ES 7200rpm 18GB drives running a RAID0 (striped, no parity)
array. This array forms the basis of my main system drives and additionally,
I have a ~30GB UDMA33 7200rpm IDE drive which I use as a dumping ground for
files I only need to use once in a blue moon. The OS is NT4.0.

Some example timings:

Saving a 50MB cpt file to the SCSI drives takes about 5 seconds and to the
IDE drive, about 10 seconds.
Reading a 50MB cpt file from SCSI and IDE takes about 1 or 2 seconds.

Clearly, writing to SCSI outperforms IDE and has the added advantage in that
the error correction is better as the RAID controller error correction is
better than a standalone drive (I should also point out that the IDE channel
is dedicated to the harddrive - if you have another IDE drive on the same
channel which is being accessed, the IDE timings will get a lot worse).

You pays your money and makes your choice. I would always go for SCSI if the
budget allows - but IDE offers way more space for your money.

Don't know if that helped or not!

Mark




Re: filmscanners: Second Hard Drive/Umax scanner

2000-11-27 Thread Mike Kersenbrock

Rob Geraghty wrote:
> 
> Bill Ross wrote:
> >Just a thought - getting a small 10k rpm drive for current
> >work plus a cheap large 5400 to park stuff might make for
> >faster overall workflow at comparable price.
> 
> >From what I've seen, 7200rpm drives are cheap enough that it's hardly worth
> getting a 5400rpm drive.  However as far as speed is concerned, the 16MB/s
> limit of IDE is the pain - even if an IDE drive is 10K rpm it can't do sustained
> transfers at over 16MB/s.  That means (cache affects aside) a 30MB scan

I thought that the UDMA/100 IDE interface was good for up to 100MB/s which
according to reviews is much faster than the speed of the data coming off
of the physical platters within the disk drives (and that even the older
UDMA/66 is fast enough (66 MB/s) to be faster than the physical disk speed).

I don't think "plain" IDE has been what drives have used for quite a few years.

To get faster, I've read that one just needs to get faster (spin) disks
with faster track access times to get the overall speed up, so it's not the
IDE interface that's limiting factor, but the disk "underneath".  

Because high-end systems use SCSI, the high end fast-spin (> 7200 rpm) and 
fast access time disks (< 8ms )are SCSI ones.  So for really fast speed one does
need SCSI disks, but not because of the interface.

There are some other theoretical advantages to SCSI in a multi-user server 
application but I don't know if they apply to the users in this newsgroup
(not me anyway :-).

Mike K.

P.S - Magazines like P.C. Magazine has done benchmarking of servers using IDE vs
  SCSI disks, and I recall their conclusions to be that they were very surprised to
  find that it didn't make much difference in the actual system performance.



Re: filmscanners: Second Hard Drive/Umax scanner

2000-11-26 Thread Rob Geraghty

Bill Ross wrote:
>Just a thought - getting a small 10k rpm drive for current
>work plus a cheap large 5400 to park stuff might make for
>faster overall workflow at comparable price.

>From what I've seen, 7200rpm drives are cheap enough that it's hardly worth
getting a 5400rpm drive.  However as far as speed is concerned, the 16MB/s
limit of IDE is the pain - even if an IDE drive is 10K rpm it can't do sustained
transfers at over 16MB/s.  That means (cache affects aside) a 30MB scan
takes 2 seconds to load or save, and a 50MB scan takes at least 3 seconds
to load or save.  A two drive IDE RAID array could nearly halve the times,
and although they don't sound like much, having to wait a few seconds every
time you save a file gets very tedious.  Striping doesn't lose drive capacity
- if your two drives are 20GB then your array is 40GB.  However, a simple
RAID stripe does NOT give the safety of parity.  With 3 or more drives in
an array you can have parity, which means that if one drive fails, no data
is lost.  Of course, parity also loses drive capacity. :)

Also, for people whose motherboard may not support UDMA100, it gives you
that without having to replace the motherboard.  I just wish UDMA100 RAID
was around when I was deciding on which motherboard to buy so I didn't waste
the dollars on a board with built-in U2W SCSI.

As with all these things, YMMV.  As Bill suggested, 16MB/s may be just fine
for many people.  Hell, I've put up with 5400rpm for quite a while now -
but since buying the film scanner it's getting annoying!

Tony, I hope this isn't OT - I find the speed of loading and saving scans
one of the most annoying aspects of using the film scanner.

Rob


Rob Geraghty [EMAIL PROTECTED]
http://wordweb.com






Re: filmscanners: Second Hard Drive/Umax scanner

2000-11-26 Thread Bill Ross

Just a thought - getting a small 10k rpm drive for current
work plus a cheap large 5400 to park stuff might make for
faster overall workflow at comparable price.

Bill Ross



Re: filmscanners: Second Hard Drive/Umax scanner

2000-11-25 Thread Tony Sleep

>  Does anyone have any experience with the Maxtor Diamond Max 81.9 (EIDE)
> GB 5400 rpm hard drive?
>  Does anyone have any experience with the Maxtor Diamond Max 81.9 (EIDE)
> GB 5400 rpm hard drive?

I've had (still have) quite a few Maxtor drives and all have been good, apart from one 
5Gb drive which died after ~2yrs. The 7200rpm versions are significantly faster, 
though 
not quite as quick as the IBM 7200rpm Deskstars, which is what I use now, for 
preference.

Regards 

Tony Sleep
http://www.halftone.co.uk - Online portfolio & exhibit; + film scanner info & 
comparisons