Does a slave drive suffer a perfomance hit ?

  • Thread starter Thread starter davexnet02
  • Start date Start date
D

davexnet02

Hello,
I've got 30GB Maxtor 53073h4 (Diamondmax Plus 45)
as Primary master, and
30GB IBM-dtla-307030 (IBM 75gxp series) primary slave.
Both running as UDMA 5.
AMD 1.2 Ghz on a KT133a MB.

Any operation on the IBM that involves reading and writing
to it at the same time is slow.

I used winrar to create a RAR archive made of a some small folders
and files, about 220MB.
Leaving the RAR archive on the IBM, and UNRARing it there,
it takes 65 seconds.
Copy the RAR file to the Maxtor, and UNRAR it there
takes about 30 seconds.

Both drives are defragged and have about 5GB space each.
The things (possibly) affecting the operation that I can see are:
The Maxtor is FAT32(8 kb cluster), IBM is NTFS (8kb cluster)
The IBM is slave
The IBM is getting old and/or it was always this way.

I thought the IBM 75gxp was supposed to be a good performer,
and the Maxtor average. Something seems suspicious here.

PS I've run full manufacturers diagnostics and no problems were
found.

TIA for any insights.
Dave
 
NTFS is slower than FAT32 in creating files.

I would check the smart counters on drives that old.
 
NTFS is slower than FAT32 in creating files.

I would check the smart counters on drives that old.
Click to expand...
Thanks for the info.
I did a little more experimenting using the ATTO disk benchmark.
Drive C (maxtor fat32) write=30 mb/s read 28 mb/s
Drive i: (IBM ntfs) write 17 mb/s read 29 mb/s

Interestingly enough
Drive h: (maxtor ntfs logical partition) write = 24 read = 22

Not sure what to make of this, except it's clear that the IBM
disk has slower writes. However, the small ntfs partition on
the maxtor while being overall slower than the fat32 on the
same drive has slightly faster writes than reads!

Dave
 
davexnet02 said:
Hello,
I've got 30GB Maxtor 53073h4 (Diamondmax Plus 45)
as Primary master, and
30GB IBM-dtla-307030 (IBM 75gxp series) primary slave.
Both running as UDMA 5.
AMD 1.2 Ghz on a KT133a MB.

Any operation on the IBM that involves reading and writing
to it at the same time is slow.

I used winrar to create a RAR archive made of a some small folders
and files, about 220MB.
Leaving the RAR archive on the IBM, and UNRARing it there,
it takes 65 seconds.
Copy the RAR file to the Maxtor, and UNRAR it there
takes about 30 seconds.

Both drives are defragged and have about 5GB space each.
The things (possibly) affecting the operation that I can see are:
The Maxtor is FAT32(8 kb cluster), IBM is NTFS (8kb cluster)
The IBM is slave
The IBM is getting old and/or it was always this way.

I thought the IBM 75gxp was supposed to be a good performer,
and the Maxtor average. Something seems suspicious here.
Click to expand...

Probably the VIA chipset mobo and deficient busmastering DMA drivers.
Get the latest of those.
 
davexnet02 said:
Hello,
I've got 30GB Maxtor 53073h4 (Diamondmax Plus 45)
as Primary master, and
30GB IBM-dtla-307030 (IBM 75gxp series) primary slave.
Both running as UDMA 5.
AMD 1.2 Ghz on a KT133a MB.

Any operation on the IBM that involves reading and writing
to it at the same time is slow.

I used winrar to create a RAR archive made of a some small folders
and files, about 220MB.
Leaving the RAR archive on the IBM, and UNRARing it there,
it takes 65 seconds.
Copy the RAR file to the Maxtor, and UNRAR it there
takes about 30 seconds.

Both drives are defragged and have about 5GB space each.
The things (possibly) affecting the operation that I can see are:
The Maxtor is FAT32(8 kb cluster), IBM is NTFS (8kb cluster)
The IBM is slave
The IBM is getting old and/or it was always this way.

I thought the IBM 75gxp was supposed to be a good performer,
and the Maxtor average. Something seems suspicious here.
Click to expand...

They are in the same league. The IBM is a bit faster on the fast end,
the Maxtor on the slow end, but only some 10%.
The Maxtor has the better accesstime but again only slightly.
However, they both have a big difference between the silent and fast setting.
 
Probably the VIA chipset mobo and deficient busmastering DMA drivers.
Get the latest of those.
Click to expand...
Hi Ron, interestingly enough all this time I have never installed
any VIA drivers. XP came out after my MB, so all stuff was supposed
to be included. Saying that I've just downloaded the 4in1 v4.43,
supposedly the best of the older set, recommended for my class of MB.
Cheers,
Dave
 
They are in the same league. The IBM is a bit faster on the fast end,
the Maxtor on the slow end, but only some 10%.
The Maxtor has the better accesstime but again only slightly.
However, they both have a big difference between the silent and fast setting.
Click to expand...
Thanks for the info. Using the various utilities, I've turned off
acoustic mgmt for both drives. Still the maxtor has the performance
edge as I mentioned in my first post.
UNraring a RAR archive takes twice as long on the IBM drive.
Dave
 
There have been several responses with theories as to why the drive is
slower. But the subject of the original post has not really been answered
(unless I missed something). Does a slave drive suffer a performance hit?
As I understand it, the slave drive has the potential to perform just as
well as the master, all other things equal (drive type, file system, free
space, fragged). Slave and master could have just as easily been A and B or
1 and 2. Does anybody have any input or comments on that? I appreciate any
input.

THANKS!
--Dan
 
dg said:
There have been several responses with theories as to why the drive is
slower. But the subject of the original post has not really been answered
(unless I missed something). Does a slave drive suffer a performance hit?
NO!

As I understand it, the slave drive has the potential to perform just as
well as the master, all other things equal (drive type, file system, free
space, fragged).
Click to expand...

Yes.

With a large maybe: expect on old VIA chipset mobos where anything seems to
have happened at one time or another regarding EIDE support
 
Thanks! You just further reinforced my previous thoughts. From you other
posts I have read, I gather you have some very in depth knowledge of storage
technology. I am curious what else you are into. I find I run into many
usenet users in other groups I frequent, as if there is something about us
that makes us like specific things. Wierd.

--Dan
 
I should have said "small files". You said you were unraring.
Thanks for the info.
I did a little more experimenting using the ATTO disk benchmark.
Drive C (maxtor fat32) write=30 mb/s read 28 mb/s
Drive i: (IBM ntfs) write 17 mb/s read 29 mb/s
Click to expand...
The low write WRT read speeds is unusual. If I: is less than 15% free, the
cause is likely allocation overhead (searching free space). ATTO does not
preallocate.
Interestingly enough
Drive h: (maxtor ntfs logical partition) write = 24 read = 22
Click to expand...
The difference between C and H is higher track density at the beginning of the
drive.
 
Back
Top