OK, so I built a new system with a 74GB SATA Raptor as the boot drive,
and two 160GB SATA Hitachi drive in RAID0 on the nvidia SATA chip.

I also have two more Hitachi 160GB drives in RAID 1 on the SI 3114.

I noted when I first created the RAID 1 array, the SATARAID5 utility
would start up on XP boot showing the array in yellow, and performance
on this volume was really poor.

I came to learn later that yellow means 'reduced', and the system is
"restoring redundancy" in the background. Since it is a new system, I
have not been leaving it on all the time. But after learning this I
left it on overnight and found that it took 5 freaking hours to go
back to a normal (green) condition.

So after this was done the array performance went back up to an
expected level.

Well, I have been leaving the system in standby mode. Yesterday, our
power blinked out for a few minutes, which cut the system off (from
standby). After rebooting, the f'ing array was back to yellow, and
task manager said "restoring redundancy"! Oh boy, I get to wait 5-6
hours with crappy performance on the raid array while this thing
rebuilds itself.

What the FUDGE is this thing doing? I'm using maybe 5gb on this drive
right now, why does the redundancy fail when it wasn't even doing
anything, and WHY IN GODS NAME does it take 5 fracking hours???? How
is the utility determining when the array is 'reduced'??

Is this chip just a POS? Did I set it up wrong??? I bought this board
specifically so I could have this raid config.

Sounds like a device driver issue. I suggest getting in touch with SI
/looking for a better driver, possibly firmware.

When a system goes into standby, the discs (logical) will have all queued
IO's completed. Running apps are suspended and their IO buffers are left as
is. If the power fails in this situation then you will have corrupt files,
but that is all you should have, not corrupt discs.

It sounds like the driver is not picking up the power status change
correctly and marking the RAID config as clean.

In the mean time, do not use standby or get a good UPS or both. Try
hybernate instead.

In the mean time you are risking getting and irreparable array.

Why 5 fracking hours? Hmmm = 300 minutes ~= 0.5GB sync / minute = OK to me.
They are always a tad slow when synching - be thankful the system is still
usable.



"BIOSMonkey" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>
> OK, so I built a new system with a 74GB SATA Raptor as the boot drive,
> and two 160GB SATA Hitachi drive in RAID0 on the nvidia SATA chip.
>
> I also have two more Hitachi 160GB drives in RAID 1 on the SI 3114.
>
> I noted when I first created the RAID 1 array, the SATARAID5 utility
> would start up on XP boot showing the array in yellow, and performance
> on this volume was really poor.
>
> I came to learn later that yellow means 'reduced', and the system is
> "restoring redundancy" in the background. Since it is a new system, I
> have not been leaving it on all the time. But after learning this I
> left it on overnight and found that it took 5 freaking hours to go
> back to a normal (green) condition.
>
> So after this was done the array performance went back up to an
> expected level.
>
> Well, I have been leaving the system in standby mode. Yesterday, our
> power blinked out for a few minutes, which cut the system off (from
> standby). After rebooting, the f'ing array was back to yellow, and
> task manager said "restoring redundancy"! Oh boy, I get to wait 5-6
> hours with crappy performance on the raid array while this thing
> rebuilds itself.
>
> What the FUDGE is this thing doing? I'm using maybe 5gb on this drive
> right now, why does the redundancy fail when it wasn't even doing
> anything, and WHY IN GODS NAME does it take 5 fracking hours???? How
> is the utility determining when the array is 'reduced'??
>
> Is this chip just a POS? Did I set it up wrong??? I bought this board
> specifically so I could have this raid config.
>

In article <d576sb$ru1$(E-Mail Removed)>, "Mercury" <(E-Mail Removed)> wrote:

> Sounds like a device driver issue. I suggest getting in touch with SI
> /looking for a better driver, possibly firmware.
>
> When a system goes into standby, the discs (logical) will have all queued
> IO's completed. Running apps are suspended and their IO buffers are left as
> is. If the power fails in this situation then you will have corrupt files,
> but that is all you should have, not corrupt discs.
>
> It sounds like the driver is not picking up the power status change
> correctly and marking the RAID config as clean.
>
> In the mean time, do not use standby or get a good UPS or both. Try
> hybernate instead.
>
> In the mean time you are risking getting and irreparable array.
>
> Why 5 fracking hours? Hmmm = 300 minutes ~= 0.5GB sync / minute = OK to me.
> They are always a tad slow when synching - be thankful the system is still
> usable.
>

If you use RAID, you should have a UPS. The next power failure
could occur while the computer is running, and even with the
world's best drivers, the disks could be de-synchronized when
the power fails. You may want a UPS with some kind of automatic
shutdown sequence, if you leave the machine in normal run state
a lot. In Suspend To RAM state, the current consumed by the
computer is low enough, that the UPS could stay up for quite a
while. In full Run state, the uptime will not be quite as long.

I don't use RAID and I still have whatever two computers I'm
currently using on the UPS. Just one less source of issues
eliminated, as I get way more one second power outages than
anything else.

Paul

>Why 5 fracking hours? Hmmm = 300 minutes ~= 0.5GB sync / minute = OK to me.
>They are always a tad slow when synching - be thankful the system is still
>usable.


It takes this long to compare and sync 155GB of blank data? Looks like sync
time should be proportional to used space.

The RAID Controller has no idea what a FAT, FAT32, EXT3, NTFS, or other file
system looks like so has no way of knowing what is used and what is not.


"Billy" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> >Why 5 fracking hours? Hmmm = 300 minutes ~= 0.5GB sync / minute = OK to
> >me.
>>They are always a tad slow when synching - be thankful the system is still
>>usable.
>
>
> It takes this long to compare and sync 155GB of blank data? Looks like
> sync
> time should be proportional to used space.

On Wed, 4 May 2005 09:45:56 +1200, "Tim" <(E-Mail Removed)> wrote:

>The RAID Controller has no idea what a FAT, FAT32, EXT3, NTFS, or other file
>system looks like so has no way of knowing what is used and what is not.

Hmm...that's a very good point.

"Billy" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> On Wed, 4 May 2005 09:45:56 +1200, "Tim" <(E-Mail Removed)> wrote:
>
>>The RAID Controller has no idea what a FAT, FAT32, EXT3, NTFS, or other
>>file
>>system looks like so has no way of knowing what is used and what is not.
>
> Hmm...that's a very good point.

However it should be able to compare the data very quickly and determine
which blocks are out of sync.

TC

I am having the same problem. I have the same motherboard and us
standby but have my system on a UPS and have not taken a power hit.
Is there a way to up the rebuild priority? It is now at 10. It jus
says reduced, I can not even tell if it is rebuilding or why my syste
is so slow

thank
Jo