HPT371 Raid 0 problem

[jon]

I´ve mounted a raid 0 with 2 80 Gb maxtor hard disks in a highpoint
HPT371 and results are very disappointing. It performs worse than a
single drive (HDtach test). I have 3 identical harddisks so can chech
easyly. Harddisks in HPT371 are connected in master-slave mode.
Motherboard is VIA KT400 based. OS Windows XP.

I have surf for reasons but till now I don´t have a clue. Things that
I´ve tried are:

- Array Block size 16K and format NTFS 4k cluster size.
- Array Block size 16K and format NTFS 32k cluster size.
- PCI latency patches. VIA raid patch (VIAPFD105) and George Breeze.

I have not found any more hints in internet. Have you got some idea?

Thanks

 

[Will Dormann]

I´ve mounted a raid 0 with 2 80 Gb maxtor hard disks in a highpoint
HPT371 and results are very disappointing. It performs worse than a
single drive (HDtach test). I have 3 identical harddisks so can chech
easyly. Harddisks in HPT371 are connected in master-slave mode.

If both drives are sharing the same channel, then this is expected.



-WD

 

[Folkert Rienstra]

If both drives are sharing the same channel, then this is expected.
Nonsense.




-WD

 

[jon]

If both drives are sharing the same channel, then this is expected.

HPT371 is always a raid controller with single channel, so are you
stating highpoint sells a worthless product?

I don´t expect very much performance but at least 2 array disks should
perform as a single I think. HDtach speed for my raid drops heavily
from 70 Mb/s to 10 Mb/s doing peaks while single drive works slower
but steadily.

 

[Folkert Rienstra]

HPT371 is always a raid controller with single channel,

No, it isn't. It is not intended for RAID although you can use it like
that in Software RAID like any other supported IDE controller.

 

[jon]

If both drives are sharing the same channel, then this is expected.

No, it isn't. It is not intended for RAID although you can use it like
that in Software RAID like any other supported IDE controller.

I think you´re wrong. HPT371 maybe a software raid controller but it´s
a raid. I mean it´s not the same as an IDE windows software raid
because you can boot with a HPT371 raid but not with a IDE software
raid.

 

[jon]

Changing Maxtor jumper from master-slave mode to cable select has been
the solution. The problem was in the harddisks.

 

[Folkert Rienstra]

Changing Maxtor jumper from master-slave mode to cable select has been
the solution. The problem was in the harddisks.

Or maybe you had the master jumpered wrong.

 

[jon]

Or maybe you had the master jumpered wrong.

Not possible. If bad jumpered disks are not found by controller. There
is only a jumper for master-slave, and same for cable select so
mistake is rather difficult.

 

[Folkert Rienstra]

Not possible. If bad jumpered disks are not found by controller.
There is only a jumper for master-slave, and same for cable select so
mistake is rather difficult.

Oh?
Last time I looked, Maxtor still use 3 jumpers for 'Master', 'Slave'
and 'Master with Slave present'.

The last one isn't what it appears to be.

 

[Folkert Rienstra]

I think you´re wrong.

That's entirely possible. Why don't you just PROVE me wrong, OK?

HPT371 maybe a software raid controller but it´s a raid.

No, it's not. It's a chip.

Unfortunately http://www.highpoint-tech.com/USA/ataraid.htm
doesn't mention any single channel raid cards.

The only single channel card using that *chip* is the Rocket 133S
http://www.highpoint-tech.com/USA/r133SB.htm