p4c800 deluxe promise raid 0 disk failed ?

  • Thread starter Thread starter Itech
  • Start date Start date
I

Itech

I turned off APM in BIOS trying to fix standby issues in XP (system won't go
into standby per settings in power options). No luck with auto standby so
went to shutdown and hit standby to see if it would. It did but there was a
nasty "grunch" sound like a disk stopping suddenly instead of spinning down
normally.
On next reboot Promise RAID FastBuild utility says a critical error, disk
failed or not responding, enter util and choose 3 to determine which disk
failed.
So I do and it shows both disks with no indication anything is wrong with
either.
Turned APM back on with no change.
No indication anything wrong from checking around in XP and Sandra.
How do I determine if / which drive is bad?
TIA for any assistance.
 
Oops, it's RAID 1 (mirror), not 0 (stripe).
Click to expand...

Sounds like an interesting problem :-) I've never used RAID before,
but in thinking about what I would do with it, the first thing is
to test the mirroring thing before there is any data on the disks.
I.e. Set up the mirror, remove one drive, see what happens. Basically,
do a "disaster walkthrough" before the disaster happens.

I think my main concern would be desynchronizing the disks by placing
different data on them. In principle, I would disconnect one drive,
and then test the other drive, by say, read verifying the entire surface
of the disk. If you do this on the RAID port, I expect the RAID BIOS
management software will treat the disk as a single volume virtual
device. The RAID BIOS will probably write the fact that the disk is now
"lonely" into the sector of the disk used to record the configuration.
Later, if you connect the other disk onto the RAID port, the management
software will probably do a sector by sector copy of the first disk
onto the second disk, after you tell it you want to make a mirror.
So, you had better do a surface verify before allowing the BIOS to
make a copy like that.

You can test both drives one at a time, and then make a decision as
to which one you will use to "clone" the second one. My guess is
the RAID BIOS will insist on doing things this way, in order to
keep the disk contents synchronized. Otherwise, if you play with
each disk separately in Windows, then put them back in mirror mode,
the disk structures will forever diverge and be screwed up.

So, my _theory_ is, "split the mirror and you'll be forced to reclone".
Make sure you identify which disk is the "good" one before cloning.
If you are at all unsure about this, find a third drive and back up
all the data before even touching the RAID array. I learned my lesson
about recovery procedures going awry before, so the backup is now the
first thing I do, on any compromised disks. (I once had a disk with
duplicate directory structures on it - the system said one structure
was damaged - I used a respected utility to repair it, and the utility
proceeded to copy the bad directory over top of the good one!!! Kaboom.)

Post back how your testing turns out.

HTH,
Paul
 
One thing you could do is go to the hard drive
manufactures web site and download a diagnostics
program. You may have to install the drive
to be tested on one of the normal IDE ports
to test it.
 
Thanks for replies.
Slept on it & got up strategy: copy all files to other HD & then attempt to
rebuild array.
So after cleaned up to make space, copied all and rebooted and GUESS WHAT?
Right, RAID now shown as "functional", no "Critical Error" warning, system
runs thru and boots normally.
GGGGRRRRRRRRR...
Check for BIOS & FastTrak updates, all seems up to date but some confusion
at ASUS download site, new 1014 BIOS available? Must update "AsusUpdate"
util in order to get new BIOS yet running new util [v5.22.02] does not find
it so stayed w/ current 1011.
So WTF, go Shutdown > Standby to see if array gets wacked again: Nope. Still
got nasty noise as something ground to rapid halt though...
Crap. The whole point of the RAID 1 is for protection of most important data
but now I don't trust it. Tells me a drive is wacked but not which one and
then decides it's OK next day after I waste hours rebooting, checking
settings, researching, cleaning and copying files. GGRRRRRRRRRRR!!!!!!!!
Anyone seen this sort of "possible false alarm"?
Guess I'll look for utils to test health of RAID and drives......
Suggestions / comments?
 
When I first started using RAID on this MB
on boot I would intermittently get a missing
drive message on boot. And would have to
rebuild the array. This happened about once
a week. I replace a power cable that had a loose
connection and updated the RAID driver and
haven't seen the problem for awhile now.
I'm using driver 1.0.1.37 dated 8/6/2003

Itech said:
Thanks for replies.
Slept on it & got up strategy: copy all files to other HD & then attempt to
rebuild array.
So after cleaned up to make space, copied all and rebooted and GUESS WHAT?
Right, RAID now shown as "functional", no "Critical Error" warning, system
runs thru and boots normally.
GGGGRRRRRRRRR...
Check for BIOS & FastTrak updates, all seems up to date but some confusion
at ASUS download site, new 1014 BIOS available? Must update "AsusUpdate"
util in order to get new BIOS yet running new util [v5.22.02] does not find
it so stayed w/ current 1011.
So WTF, go Shutdown > Standby to see if array gets wacked again: Nope. Still
got nasty noise as something ground to rapid halt though...
Crap. The whole point of the RAID 1 is for protection of most important data
but now I don't trust it. Tells me a drive is wacked but not which one and
then decides it's OK next day after I waste hours rebooting, checking
settings, researching, cleaning and copying files. GGRRRRRRRRRRR!!!!!!!!
Anyone seen this sort of "possible false alarm"?
Guess I'll look for utils to test health of RAID and drives......
Suggestions / comments?

Itech said:
Oops, it's RAID 1 (mirror), not 0 (stripe).
Click to expand...
Click to expand...
 
I've got 1.0.1.28., 6/3/2003, according to Windows XP.
I see one called Promise FastTrak 378 RAID Driver V1.00.1.37 for Windows
98SE/NT4/ME/2000/XP/2003 on ASUS download site.
Is this what you are using on Asus P4C800 Deluxe?
TIA...

JBM said:
When I first started using RAID on this MB
on boot I would intermittently get a missing
drive message on boot. And would have to
rebuild the array. This happened about once
a week. I replace a power cable that had a loose
connection and updated the RAID driver and
haven't seen the problem for awhile now.
I'm using driver 1.0.1.37 dated 8/6/2003

Itech said:
Thanks for replies.
Slept on it & got up strategy: copy all files to other HD & then attempt to
rebuild array.
So after cleaned up to make space, copied all and rebooted and GUESS WHAT?
Right, RAID now shown as "functional", no "Critical Error" warning, system
runs thru and boots normally.
GGGGRRRRRRRRR...
Check for BIOS & FastTrak updates, all seems up to date but some confusion
at ASUS download site, new 1014 BIOS available? Must update "AsusUpdate"
util in order to get new BIOS yet running new util [v5.22.02] does not find
it so stayed w/ current 1011.
So WTF, go Shutdown > Standby to see if array gets wacked again: Nope. Still
got nasty noise as something ground to rapid halt though...
Crap. The whole point of the RAID 1 is for protection of most important data
but now I don't trust it. Tells me a drive is wacked but not which one and
then decides it's OK next day after I waste hours rebooting, checking
settings, researching, cleaning and copying files. GGRRRRRRRRRRR!!!!!!!!
Anyone seen this sort of "possible false alarm"?
Guess I'll look for utils to test health of RAID and drives......
Suggestions / comments?

Itech said:
Oops, it's RAID 1 (mirror), not 0 (stripe).

I turned off APM in BIOS trying to fix standby issues in XP (system won't go
into standby per settings in power options). No luck with auto
Click to expand...
standby
so
went to shutdown and hit standby to see if it would. It did but
Click to expand...
there
was a
nasty "grunch" sound like a disk stopping suddenly instead of
Click to expand...
spinning
down
normally.
On next reboot Promise RAID FastBuild utility says a critical error, disk
failed or not responding, enter util and choose 3 to determine which disk
failed.
So I do and it shows both disks with no indication anything is wrong with
either.
Turned APM back on with no change.
No indication anything wrong from checking around in XP and Sandra.
How do I determine if / which drive is bad?
TIA for any assistance.
Click to expand...
Click to expand...
Click to expand...
 
yes

Itech said:
I've got 1.0.1.28., 6/3/2003, according to Windows XP.
I see one called Promise FastTrak 378 RAID Driver V1.00.1.37 for Windows
98SE/NT4/ME/2000/XP/2003 on ASUS download site.
Is this what you are using on Asus P4C800 Deluxe?
TIA...

JBM said:
When I first started using RAID on this MB
on boot I would intermittently get a missing
drive message on boot. And would have to
rebuild the array. This happened about once
a week. I replace a power cable that had a loose
connection and updated the RAID driver and
haven't seen the problem for awhile now.
I'm using driver 1.0.1.37 dated 8/6/2003

Itech said:
Thanks for replies.
Slept on it & got up strategy: copy all files to other HD & then
Click to expand...
attempt
to
rebuild array.
So after cleaned up to make space, copied all and rebooted and GUESS WHAT?
Right, RAID now shown as "functional", no "Critical Error" warning, system
runs thru and boots normally.
GGGGRRRRRRRRR...
Check for BIOS & FastTrak updates, all seems up to date but some confusion
at ASUS download site, new 1014 BIOS available? Must update "AsusUpdate"
util in order to get new BIOS yet running new util [v5.22.02] does not find
it so stayed w/ current 1011.
So WTF, go Shutdown > Standby to see if array gets wacked again: Nope. Still
got nasty noise as something ground to rapid halt though...
Crap. The whole point of the RAID 1 is for protection of most
Click to expand...
important
data
but now I don't trust it. Tells me a drive is wacked but not which one and
then decides it's OK next day after I waste hours rebooting, checking
settings, researching, cleaning and copying files. GGRRRRRRRRRRR!!!!!!!!
Anyone seen this sort of "possible false alarm"?
Guess I'll look for utils to test health of RAID and drives......
Suggestions / comments?

Oops, it's RAID 1 (mirror), not 0 (stripe).

I turned off APM in BIOS trying to fix standby issues in XP (system
won't go
into standby per settings in power options). No luck with auto standby
so
went to shutdown and hit standby to see if it would. It did but there
was a
nasty "grunch" sound like a disk stopping suddenly instead of spinning
down
normally.
On next reboot Promise RAID FastBuild utility says a critical error,
disk
failed or not responding, enter util and choose 3 to determine which
disk
failed.
So I do and it shows both disks with no indication anything is wrong
with
either.
Turned APM back on with no change.
No indication anything wrong from checking around in XP and Sandra.
How do I determine if / which drive is bad?
TIA for any assistance.
Click to expand...
Click to expand...
Click to expand...
 
Interesting: in Windows it says 1.0.1.28, but during boot it says 1.00.1.37.
I'll try installing .37 in Winblows...
Thanks for reply.
 
Just gets better:
Now XP says it's 1.0.1.37 in Driver tab of "WinXP Promise FastTrak 378 (tm)
Controller" but in "Driver Details" (didn't bother to look b4) it has 2
files:
c\w\sys32\drivers\Fasttx2k.sys = v1.00.1.37
c\w\sys32\ptipbmf.dl = v1.00.0.3
This is for "WinXP Promise FastTrak 378 (tm) Controller" under "SCSI &RAID
Controllers" in Device Mgr., the one I had before. OK!

But now I have a new device under "SCSI & RAID controllers":
"WinXP Promise Promise RAID Console SCSI Processor Device".
It also shows 1.0.1.37 in Driver tab, but hitting "Driver Details" button
results in "No driver files are required or have been loaded for this
device".

Clear as mud: installing new drivers results in new device that has no
drivers.

Perhaps new one is for SATA (not used yet) and the original was for ATA 133
(in use)?

Oh well, any way the winblows.

Suggestions, comments, enlightenment appreciated,

duh
 
Thanks for the reply.
Reassuring to say the least.
Happy Holidays, you get Karma points, hope it pays!
duh
 
When I first started using RAID on this MB
on boot I would intermittently get a missing
drive message on boot. And would have to
rebuild the array. This happened about once
a week. I replace a power cable that had a loose
connection and updated the RAID driver and
haven't seen the problem for awhile now.
I'm using driver 1.0.1.37 dated 8/6/2003
Click to expand...

I had exactly the same problem on the same board, except that I'm
running Raid 0. It was always Chan 1 in the Raid 0 array that would
fail. New cables (power and data) and drivers didn't help. I
couldn't trust that the machine would start and I got tired of
rebuilding the array. I was pretty sure it was the Promise controller
on the motherboard, but wasn't willing to send the board back to ASUS
for repair/replacement (2-3 weeks). I went out and bought the Promise
FastTrak S150 TX4 for about $60 and haven't had a speck of trouble
since.

Pete
 
Thanks.
Have seen several similar comments to the effect of "just get a real
controller" for various reasons.
I'm going to see how it goes now with new drivers.
And hope to pick up some big SATA drives soon so will try the SATA RAID.
But yes, a card is a serious possibility.
 
duh said:
Thanks.
Have seen several similar comments to the effect of "just get a real
controller" for various reasons.
I'm going to see how it goes now with new drivers.
And hope to pick up some big SATA drives soon so will try the SATA RAID.
But yes, a card is a serious possibility.
Click to expand...

I've been thinking about this and just realised it
couldn't be new drivers that fixed my problem.
Because the error occurs before windows is even.
loaded. The 1011 bios had an update for the
promise raid boot rom. I forgot I'd updated the
BIOS too, so that's what probable fixed my problem, or the cable.
 
No more RAID trouble here yet.
Before I updated Windows drivers Windows was showing 1.0.1.28 but BIOS said
1.00.1.37. Now after updating drivers BIOS says 1.00.1.37 and Windows says
1.00.1.37 and there's no more trouble (yet, knock on wood).
Makes me wonder if not having them match could cause trouble, IE Windows had
it going one way then BIOS tried to change back on reboot?
I'll top post back if trouble returns.
duh
 
Back
Top