Gigabyte with RAID 5 crashed

I had a RAID 5 with 3 x 500GB harddisks (SATA) on a Gigabyte P35-DQ6 crash and sent it on a RMA and got it back today but they couldn't find anything wrong with the motherboard but they tested with one ATA harddisk and it ran through stresstest and windows xp installed fine and have now returned it, will get it back probably tomorrow



This is what I had done before the RMA

before crash
http://img529.imageshack.us/img529/6...atrixsmgi2.jpg

after crash
http://img98.imageshack.us/img98/8783/84raid5.jpg

tried to install windows xp but it crashed
http://img231.imageshack.us/img231/2430/83bsod.jpg


How do proceed to rescue the harddisks?

Another RMA or something else?



PS this is related to the motherboards and its RAID controller so the right sub forum for this thread is here or if mod finds a better place move it DS


Facts in sig. (put it here as the sig is newer and got other raid installed)

Edit: 2014-01-16 0-2:3x500GB Seagate Barracuda 7200.10 16MB cache firmware AEK on Intel Matrix Storage Manager with 128GB RAID 0 128k stripe and 850GB RAID 5 gone bad

How do I rescue the data and get it to work?

I bought one extra 1TB SATA harddisk but it crashed when I tried to install windows XP

BSOD
http://img231.imageshack.us/img231/2430/83bsod.jpg

You need to press F6 when windows starts the reinstall to load raid drivers

From the second image it looks like two of the three hard drives failed. I have had far to many Seagate "AS" aka ASS drives fail recently and most likely nothing wrong with the motherboard, just really crappy drives.

You need the same model drive and swap out the PCB first and hope the drive comes back online and spins. If the drive spins and does not come back online put the original controller on it and pick up a Disk Jockey Pro

The DJ Pro will move the data onto the new drive and if all goes well it will fire up on raid like there was nothing wrong, if the drive has physical damage to the platters or the read heads it will fail and prompt you to skip or abort the process, you can skip those block reads until it gets to a good area and hope the data you need is still intact.

I have read about your experience with SSD, raid 5 and Murphys Law as I´m a longtime XS reader, I don´t write very much though, but you made it come back to life, thanks

The weird thing is that had 2-3 day before this raid 5 crash happened I had upgraded my nforce drivers for PhysX support and had tested it in a few applications, gamedemos, screensavers etc that had support for it, downloaded all these from nvidia and selected about 4 GB from the 4.6 GB Graphic Plus Power Packs

...some screensavers worked great and some games but some other didn't nothing out of the ordinary but one application, Crazy Machines II PhysXFluids (CM2), had you to enter your name before actually start using it´s PhysX effects etc but the mouse cursor to it was gone and the only way to get out of the program was to press the power button on the computer for 4 seconds restarted computer tried again, same thing, uninstalled it, then tried a later revision of it but exactly the the same result

...but this time the desktops icons in the taskbar had changed size, not a good sign, but it was easy to change back and then I tried Firefox etc it all looked like it worked but when I started Winamp the computer froze only the mouse moved, did the 4-second restart thing again and thought it was something with winamp and deleted its local settings/temp/folder and tried it but i froze but this time when windows xp had started the Intel Storage Matrix Manager ran a Verification and Repair on d: the raid 5 volume, this took 3 hours

it found nothing wrong and did nothing, didn't report that any of the 3 harddisks had failed or anything, tried to see log reports of any evidence of some errors etc but found nothing

computer restarted and looked like it worked anduninstalled CM2 and ran antivirus & antispyware but it found nothing

Then tried Media Player Classic (MPC) and it worked great playing up music and videos then I tried Winamp again and it froze but this time after the 4 sec powerbutton thing it just shows the

http://img98.imageshack.us/img98/8783/84raid5.jpg

and the freezes, you cant even get into the raid bios or the motherboard bios, execpt if you disconnect the raid harddisks but them you wont see anything concering them anyway but did a BIOS default reload and have tried many things but all failed

Bought one of four 1TB SATA harddisk just to see if the motherboard was ok or not but got a BSOD when installing windows XP, this was in March so I dont remember every step that I did, but I know from the past that at some point you press F6 to enter the Raid-drivers,

well, its too late now 4am, will think about buying the
Disk Jockey Pro IT - Hard Drive Duplicator with SATA (DJ), or sending the computer to some computerexperts or something but its probably more expensive than the $500 DJ



How can two drives or more fail at the same time?


It annoys me a bit that the RMA guy didn't put in a SATA harddisk to make sure the SATA controller is ok or not, now I don't now, and it might be a bad idea if the SATA-controller is bad and it scrambles the raid 5 drives even more, well idk, will try again and see if I can find some used parts on the internet as I always buy new stuff




I searched some webshops around here but they didn´t have the 7200.10 any longer as that was may plan to put in other one

Hey Hell fire!

You were over clocking on the P35 witch has an ICH9 intel chip, correct? On my X48-DQ6 the ICH9 would drop off line under heavy over clocking, figured I'd put that out there for you. Same same with Gigabyte USA the problem there is they will not over clock the system to stress the ICH9 chip. So no warranty, ever sense they tried it I just been using a straight up OS no more raid.
May or may not be your problem.

I'm quite suspicious of RAID on Gigabyte boards these days I found it to be quite flaky on my last machine (using your board) with a couple of 7200.10s in RAID 0, and RAID 5 has been a nightmare on my MA790GP-DS4H.

Quote:

---

Originally Posted by jaggerwild

Hey Hell fire!

You were over clocking on the P35 witch has an ICH9 intel chip, correct? On my X48-DQ6 the ICH9 would drop off line under heavy over clocking, figured I'd put that out there for you. Same same with Gigabyte USA the problem there is they will not over clock the system to stress the ICH9 chip. So no warranty, ever sense they tried it I just been using a straight up OS no more raid.
May or may not be your problem.

---

I had only a moderate overclock q6600 2400 @ 3200 Mhz
It which was 100% stable in all games, prime etc
Crysis maxed out, GRID fluidly running, etc

I have not been able to run SATA harddisks in SATA-mode (AHCI) and got a BSOD while installing windows xp, as I mentioned before, and was thinking of switching mode later on so I have had it in IDE-mode

Today I found a good guide (link) and got it working in SATA-mode again and the Intel Matrix Storage Manager for the ICH9R seem to work, will so some testing to make sure it works 100%

Maybe I will try to see if it works in RAID 0+5 mode again but that might be a moot point as it the RAID 5 has already failed, to make sure the ICH9R works 100% and then use RAID 0 + 1 instead, offloading the heavy parity bits calculation's from it so it wont heat up as much.

The motherboard P35-DQ6 has the CrazyCool heatpipes etc which should be able etc transfer the heat unless something went wrong, don´t know yet what really happened.

Thinking of changing windows xp to run the SATA harddisks in SATA RAID-mode (RAID), there is a registry good guide (link) for it too but as I havn´t bought a spare 500GB Seagate Barracuda 7200.10 of the exact same type I might have to wait with it.

Going to calibrate the RealTemp with some serious overclock but thats another thread, don´t know why as I´m not a "serious" overclocker, moderate overclocking with a stable computer is fine for me.

I had planned to replace the 3 x 500 GB Seagate Barracuda 7200.10 RAID 0+5 as I had ran out of space when the crash happened and the harddisk which I purchased and have running in SATA (AHCI) is a 1TB Samsung Spinpoint F1

iMatrixSM-ICH9-SATA-AHCI
http://img200.imageshack.us/img200/7...h9sataahci.jpg

I was going to buy another three to get a 3 x 1TB and a spare but now are bigger 2 TB out and I might go away from the orginal plan and RAID 0+5 and buy 2 of them or something, there are some OS-limitations with very big harddrives? Haven't done any research of what makes sense to get right now.

I'm not sure if you are having the same problem as I did in this post. http://www.xtremesystems.org/forums/...d.php?t=225184 Mine was freezing in the exact same spot as your second picture. After erasing both drives with SeaTools I haven't had an issue since.

If it is the same problem, enable RAID in BIOS with the drives unplugged then save and exit. Once the CTRL+I option appears, enter it and plug the drives in then open and exit a few of the menus in the utility until it says the array is complete. It usually only took one or two menus before it was complete for me. If it works save all your important data before restarting again as it will stop in the same spot again until you erase the drives.

I have used the P35-DQ6 and RAID, and several times have had RAID arrays fail on both the P35-DQ6 and EP35-DS4 boards that I tested. One was a RAID 10 array, the other a RAID 0. In the case of the RAID 10 array I was preforming a warm reboot, I rebooted, and suddenly the AHCI RAID controller could not detect the RAID array and said two of the drives had failed. Didn't matter if I reset the system back to stock, or replaced the motherboard with the DS4 but it couldn't find the array.

After that I started testing, and began noticing that even mild overclocks of ~3.2GHz on my Q6600 would start to see bad data written to the RAID array. For awhile I thought it could be anything from a unstable overclock to unstable RAM, but it wasn't. Have you run the hard drive array data verification & repair test under Intel's Matrix Storage controller? It was finding anywhere from 10-100 errors every few days... Worried about my RAID array I reset to stock but then even at stock it started finding errors! Those were also Seagate hard drives, four 7200.10 320GB drives.

After losing a second array and getting nothing but new data verification errors on a 3rd RAID array I gave up Intel Matrix RAID, it offers no protection whatsoever and all it takes is a single reboot to go from a fully working RAID 10 array to losing all your data. Much safer to buy a NAS with RAID, use just a single drive for overclocked systems or an SSD. Much fewer headaches, frustration, and risk that way.

Yes!
I think with a Quad core the OC does not have to be as high, as the board seems to perform better with a core 2 duo. It may be the added voltage of a Quad or just more stress maybe. I forget who told me about the ICH9 dropping off line, but they were correct.
After that point I just went with a straight OS on one hard drive, as rebuilding and saving my raid was too nerve stressing on me. I think I was Raid 5 but not certain.
I do not think the factory cooling is much good on it really, and I could not put a water block on mine as I was running crossfire-X on the board(no room).:down:

I sent the harddisks for the Raid 0 + 5 away today to a place specialized in rescuing crashed computers.

...btw I have got a single running on the ICH9 controller, a 1TB Samsung (SATA) running the AHCI (SATA mode) woriking again so it's probably some of the 500GB Seagate 7200.10 harddisks that have failed.
...still, two drives failing at the same time doesn´t fit in my world of probability, except for some extreme other thing going on, failed fans making som the harddisk overheat etc, but I haven't got any fans that died but since they are kind of software/hardware driven with the T-Balancer XL it can't be ruled out.

I have seen temperature curves the T-balancer that somehow been altered, but as far I can remember it has never been the harddisk- , cpu- or case out fans.

This image shows the 80mm doorfan that can blow some more air into the gpu intake area, but it was mostly ruin in manual-mode and turned off anyway as it lowered the the gpu by maybe 2C at 100%

http://img13.imageshack.us/img13/393...curveerror.jpg

Quote:

---

Originally Posted by Kougar

After losing a second array and getting nothing but new data verification errors on a 3rd RAID array I gave up Intel Matrix RAID, it offers no protection whatsoever and all it takes is a single reboot to go from a fully working RAID 10 array to losing all your data. Much safer to buy a NAS with RAID, use just a single drive for overclocked systems or an SSD. Much fewer headaches, frustration, and risk that way.

---

Ouch!

After having some fun with a RAID 5 on Nforce 4 I too have decided to go RAID 10 on the Intel boards: first mirroring and then striping your data. So, 1-disk errors should be no problem and your data should be complete and restorable?
Dropping 50% of your disk capacity for nothing? Too bad ...

Quote:

---

Originally Posted by -X-hellfire

I sent the harddisks for the Raid 0 + 5 away today to a place specialized in rescuing crashed computers.

---

It turns out that two of the three harddisks has cluster problems etc and is being cloned right now to other more reliable harddisks for further analysis as what steps need to be done to recover the them (or as much data as possible). I will know more tomorrow as they need more time to asses the damage. It would be nice to know the cause of the breakdown of the RAID 5 as it right know seams as a the extra data security of RAID 5 is pointless if it fails in this way.

Well I taked to the recueoperator and he he was abit concerned as the Raid wasn't found with his hardware as a RAID 0+5 but when talked some more with him it sounded like he didn´t have any experience with Intel Matrix RAID with both RAID 0 and RAID 5 on the same three harddisks
http://www.ocforums.com/showthread.php?t=467848
http://img269.imageshack.us/img269/5912/matrix2k.png

...had to explain it to him which is not a good sign when you want your data recued by "experts"

Maybe I´ll cancel his recue effort and do it myself and a Disc Juggler Pro

After four days I got a mail today and it looks like it would be too expensive to let this data recovery expert do it. He wrote that he only could see the directory view of the files but not open them, yet, and it would take a long time, very long time to recover some of the data. As a first estimate it would cost $1300 to recover the data (don't know how much in % that is lost)

I think I will try to do it myself, buy a Disk Jockey Pro IT for $459, or more as it doesn't look available in sweden, wait found it now and it cost about $1000.

Ouch, that is twice as much, for something I will only use for this data recovery and never be used again as I'm done with Intel Matrix Raid 5. Although it could be the 500 GB Seagate Barracuda 7200.10 and some flaky firmware that caused this mess, I don´t know.

I will probably get a new haircolor, gray will look great, by the time I´m finished with it.

I decided to bring my crashed RAID 0 + 5 home from the "Experts" and paid a fee for them had been analying them for a week. When I got there the guy who I been deling with before wasn't there and it was some other guy who was on the verge to close the shop at 2pm. ****ing clowns.

Anyway I got the 3 x harddisk home and put them back in and started the computer and it halted in the same way as before, only showing 1 drive.

http://img98.imageshack.us/img98/8783/84raid5.jpg


Disconnected harddisk #1 and tried again, this time though the screen looked a little different and said something like "Controlled by Raid Bios" and found one harddisk, didn't take a photo of it.



Reconnected harddisk #1 and disconnected harddisk #2 and restarted and suddenly the RAID volumes was listed and showing the status of them






Raid Volume MatrixRAID0winXP has FAILED
Thats the Raid 0 (stripe) with c: which has Windows XP on it

Raid Volume MatrixRAID5wPrg has been DEGRADED
Thats the Raid 5 (parity) with d: which has all data on it, and some ghost files that I didn't backup on other media


DSC13283-RAID-disk13ok-disk2bad
http://img44.imageshack.us/img44/517...isk13okdis.jpg



Pressed Ctrl I and entered the Raid bios
DSC13284-RAIDbios-disk13
http://img188.imageshack.us/img188/9...biosdisk13.jpg



It's hard to read some lines/menus from the photos taken, tried to zoom in
DSC13286-RAIDbios-create
http://img188.imageshack.us/img188/6...bioscreate.jpg







Well it looks like harddisk #2 has died, so I almost went the nearby computer store to buy a "spare" 500GB harddisk to put in, but they had closed for vacation but anyway found one in the a webshop, they had only one 500GB in store so I drove to town and got it and have now put it in the computer



DSC13288-RAIDbios-disk2spare
http://img81.imageshack.us/img81/287...iosdisk2sp.jpg



I pressed ENTER so now it is in REBUILD status, and will start to rebuild the Raid 5 once I get Windows XP up and running again.

DSC13290-RAIDbios-Rebuild
http://img525.imageshack.us/img525/7...iosrebuild.jpg








I'm wondering how to proceed here.

...should it be something like this?
1) Delete the FAILED volume on c:
2) Create a new volume MatrixRAID0winXP c:
3) Disconnect all usb 1.1 devices
4) Connect an external usb 2 harddisk (Mybook-1) with older ghost images on it

5) Start ghost from floppy and load a ghost from Mybook-1 to c:
6) Disconnect Mybook-1, reconnect usb1.1 and restart computer
7) Hope windows starts and then it will start the REBUILD of d: (takes 2-4 hours)

8) Do a backup of d: to my other Mybook-2 (takes 5-15 hours)

9) Disconnect the RAID harddisks #1 #2 #3


10) Think about a better RAID-configuration :shrug:

Intel® Matrix Storage Manager
Failed RAID 0 Volume
http://www.intel.com/support/chipset.../cs-020992.htm

Quote:

---

Failed Hard Drive Member

Note: When a hard drive member of a RAID 0 volume has failed, the RAID 0 volume cannot be recovered and any data on the RAID 0 volume is lost. This procedure deletes the failed RAID 0 volume.

1. Make sure the system is powered off.
2. Replace the failed hard drive with a new one that is of equal or greater capacity.
3. Power on the system. During the system startup, the Intel Matrix Storage Manager option ROM user interface will display the RAID 0 volume status as "Failed," prompting you to press Ctrl-I to enter the user interface.
4. Press Ctrl-I to enter the "MAIN MENU."
5. Select option 2 in the main menu to "Delete RAID Volume."
6. In the "DELETE VOLUME MENU," select the failed RAID 0 volume using the up and down arrow keys.
7. Press the Delete key to delete the volume.
8. Press Y to confirm the deletion.
9. Create a new RAID 0 volume. If the failed hard drive was a part of the system volume, you will also need to reinstall the operating system.

---

now that I can see the harddisks and Before I start the the lenghty rebuild I wen´t the Seagate.com site and tried to get some information of the latest firmware for the 500GB Seagate Barracuda 7200.10 (ST3500630AS)

I can't find anything on that site, but it is probably hidden deep deep in there somewhere.

If it is not to risky doing a firmware update, it might be a good idea doing it.

edit: ...found a thread about firmware AAK vs AAE but it seems not possible to update the firmware
http://www.fluffles.net/forum/storage/34.2

Quote:

---

I just got off the phone with a Level 2 tech at Seagate. I have 8 of these 500GB ST3500630AS (7200.10) drives, all with the 3.AAK firmware. The AAK firmware is on the newest controller boards manufactured in China, and there is no way to change/flash the firmware. He said they could not even do it in house, the board does not have flash capability. He downplayed the performance difference between AAK and the previous firmware revisions, but I saw in the post above the AAK firmware is apparently optimized for 64k block size. The only thing I can do is RMA the drives, and I would get back an identical drive, which may or may not have the AAK firmware. Since I am running these drives in a RAID5 setup (2 servers with 4 drives each), I want the same firmware for all the drives, so I'm going to stick with what I have. I don't know whether I would even see a performance difference with an older firmware version/drive in a file server, the network is usually the bottleneck even with GbE, and these servers are lightly loaded anyway. But I am going to change the block size on the filesystem to 64k, it can't hurt, since most of my files are huge (>100MB). Also I am going to remove the jumper to force SATA II mode, since I have a SATA II interface in each server.
- Chris

---

Need some help here "RAID GURU:s"

...I have tried to run ghost but it freezes when it tries to access d: the RAID 5 volume MatrixRAID5wPrg which is in Rebuild-mode which will start to REBUILD as soon as Windows with the Intel Matrix Storage Manager installed sees it, but as the c: RAID 0 volume MatrixRAID0WinXP FAILED and I done the these things
1) Delete the FAILED volume on c: in the ICH9 bios
2) Create a new volume MatrixRAID0winXP c: in the ICH9 bios
3) Formated c: with NTFS i Partition Magic

I conneceted external usb Mybook 1TB and started Ghost but it freezes after a while when I still had the FAILED c: then tried a another smaller one 320 GB, same thing but the it worked for a while in the menues, tried disconnect mouse and minimal stuff connected, same thing, then after I deleted the c: Ghost never starts up, just freeze when it founds the d: which is in REBUILD status.


It looks like I can´t use Ghost as at all for now.

I could backup my internal c: 1 TB Samsung SATA harddisk, which runs fine in AHCI (SATA-mode) on the Intel ICH9 controller now, and start windows from it and the let the Intel Matrix Storage Manager start the REBUILD of d: (which needs a to be run in Windows)

I don´t know if it can coexist with the RAID 5 harddisks on that same controller?

What is the possibilities when the c: has to be created some how, I guess if I start a windows install it will freeze as soon as it finds the REBUILD d: volume as it hasn't got the Intel Matrix Storage Manager installed yet.


There are a some other possibillities
1) Connect an internal ATA 300 GB on the Gigabyte2 controller ATA-mode (needs to a backup of it first)
2) Connect an internal Samsung 1TB SATA on the Gigabyte2 controller in AHCI (SATA-mode)
3) buy a usb Flash drive, which support booting from it, and install Windoes XP and on it
...
some other options?

2) Connect an internal Samsung 1TB SATA on the Gigabyte2 controller in AHCI (SATA-mode)
which has Windows XP on it and

Connect 3 x 500GB Seagate (SATA) in RAID 5 on the Intel ICH9R controller (RAID-mode in bios)

Will try this approach and start the rebuild from it with Intel Matrix Storage Manager.

Here is an update on this... mess



I bought 5 x 1 TB and configured as following...

RAID 10 on Intel Matrix ICHR9 controller
4 x 1 TB Samsung F3 (SATA/RAID)

Photoshop Scratchdisk, and hot spare if raid fails
1 x 1 TB Samsung F3 (SATA/non raid)

Gigabyte Jmicron SATA2 or called Gigabyte2 (SATA/non-raid) ...2nd controller (36x)
WinXP and lots of data
1 x 1 TB SAmsung F1 (SATA/non raid)
c: WinXP-127
d: wProg-826

created two volumes
c: WinXP R10
d: wProg R10

loaded ghost files for c and d but WinXP didn't like going from non-raid to raid, long story and skipped loading ghost on c and installed WinXP-64

created two volumes
c: WinXP-64 R10 ...... 128 GB ...installed OS
d: wProg-64 R10 ....... 1.7 TB ....loaded the ghost again (8 hours from a external usb 2 harddisk)



It worked great, then rebooted and started the c: WinXP-127 on Jmicron 36X controller, had to install drivers for it and Intel Matrix ICH9R, and got into the RAID 10 from it, tried several thing and suddenly we had a blackout and it was enough to get the RAID 10 to rebuild for 2-3 hours. Anyway it worked out great so after that I disconnected the RAID 10 array and connected............







Ta- da... the old RAID 5

RAID 5 on Intel Matrix ICHR9 controller
3 x 500 GB Seagate 7200.10 (SATA/RAID) but one exchanged for a
1 x 500 GB Seagate 7200.12

...and to my surprise it started to rebuild it :shocked: and after it was done 3-5 hours later I got a normal RAID 5 system running



I could then copy all my photos, drawings, work and most important stuff I thought I lost to an external usb harddisk ( 300 GB or so until it was almost filled )




Then I was just checking out folder by folder for more stuff to copy and got tired, took a break for some hours, continued and got a old Ghost folder and it suddenly said ACCESS DENIED

WUT? :(

Tried several files in that folder and even a .txt file got the same response.



Rebooted the computer but I got a BSOD with some NTFS error message, tried several times same thing.
(It shows that the RAID 5 is normal i RAID-bios)


Disconnected the RAID 5 an put RAID 10 in and it worked right away.


I got at least the most important stuff back but it would be nice to get it running again :)

Here are some screencaps and photos to the post above




1 TB Samsung F1 on the Intel Matrix ICH9R controller in SATA-mode

http://img708.imageshack.us/img708/4784/hdd1tb.jpg



Installed drivers for it in WinXP
Gigabyte Jmicron SATA2 (SATA) 36X controller and
Intel Matrix ICH9R RAID controller

http://img19.imageshack.us/img19/365...icronsata2.jpg



First I installed a RAID 1 and tried it when I had 3 x 1TB Samsung F3

intel matrix raid1 with F3 array
http://img638.imageshack.us/img638/3...id1withf3a.jpg

intel matrix raid1 with F3 WinProg -811 volume
http://img638.imageshack.us/img638/6...withf38.th.jpg

intelmatrix raid1 with F3 RAID harddisk info
http://img638.imageshack.us/img638/3...withf3h.th.jpg







I tested th RAID 1 and noticed the RAID 10 capability in RAID-bios I decided to buy two more F3


edit: 5:07am I was going to upload some more recent photos/screencaps of the raid from my Sony DSC-F707 camera but it wasn't recognized by WinXP-64 and then tried to download a new driver

* Not Supported:
o The Microsoft Windows XP 64-bit operating system is not supported.

Have to reboot and start WinXP to transfer them :rolleyes:

edit: 5:30am Have now transfered the photos and I´m back to WinXP-64 and can continue this update

It was just a couple of photos on the camera and they are
not uptodate as the RAID 10 volume WinXP-R10 120
is only 120GB and WinXP



dsc13470 raidbios 5 x 1 TB Samsung F3 with non-raid

http://img693.imageshack.us/img693/2...ios5x1tbf3.jpg



dsc13471 raidbios 4 x 1TB Samsung F3 with RAID 10

http://img693.imageshack.us/img693/8...ios4x1tbf3.jpg


edit: 5:14am will take some current photos but then it is back
to the dualboot routine, and I who thought going to WinXP-64 would
be a trouble free move ;)










edit: 7:17am

Here are some current photos of the RAID 10 volume WinXP-R10 128
and is corrected to the 128 GB limit (137 GiB) and WinXP-64,
I thought I changed the volume-name to something with 64-bit
but apparently I didn't :shakes: but as I recall now it was because of
the 11 character filename limit, and 10 in Ghost 2003, that I decided
to keep it cleaner


DSC13511 raidbios 4 x 1TB Samsung F3 with RAID 10

http://img682.imageshack.us/img682/6...ios4x1tbf3.jpg



DSC13513 raidbios 4 x 1TB Samsung F3 with RAID 10

http://img52.imageshack.us/img52/603...ios4x1tbf3.jpg

Tried to get the colors right with the these photos but the
"normal" text should be green and the other green text
should be yellow... bummer :(


DSC13512 gigabyte2 36x 1TB F1 with non-raid

http://img52.imageshack.us/img52/862...236x1tb.th.jpg

I am having trouble rebooting the computer. It "freeze" randomly when the windows startup screen has done 2 - 6 bars.

It is totally stable once I get it up and running, then I can run all heavy games, test etc and its stable for weeks even month.

...however it´s very unstable once I reboot or turn it off for the night, next reboot it "freezes".


Windows xp pro (64-bit)
This is with the current raid 10 as seen in my sig.
...so when it crashes it can take 20 min for c: or 6-8 hours for d: to verify the raid 10.

I have been dual booting by going into bios and selecting a 1 TB Samsung Spinpoint F1 ( UJ ) with windows xp pro (32-bit) and usually it has been stable while booting but the last couple of weeks it has also "freeze" randomly as the (64-bit)



Monday I spent 3 hours trying to start the computer until I went to bed and turned it of for the night.
Tuesday, I had read that sometimes usb hardware can cause crashes and my printer is on a almost 5 meter cable (3+2m or something). When windows is running it from time to time the "pling plong" sound is heard as it find the printer. I disconnected the printer and next the time I started the computer it worked like a charm.

I have since had it on, until today when I turned it off and when it rebooted it freeze again again and again in both 32-bit and 64-bit windows. I could not even get into "Safe mode".

I can conclude that the usb-printer is not the culprit for the "freezes"





Then got into bios and set the 2400 @ 3200 Mhz oc to 2400 and the 266 @ 400 Mhz FSB to 266.

...restarted and after a 20 min raid verify it started without a problem.






Question: Is this the "cold bug" or some other thing?

How can I make the stable 2400 @ 3200 overclock stable during booting windows?

Booted and selected windows xp pro (32-bit) in bios ...still with default clock cpu, no oc
...computer freeze after the "bars" with a "black screen with prompt of death" both in Safe Mode and Normal.

The second attempt with Normal it started, weird...
...and the raid 10 started to verify c: (but I aborted the operation)

Did some windows update (WU) restarted and it worked.



Restarted and selected windows xp pro (64-bit) in bios ...it worked
...did the same WU restarted, worked again.

Restarted and applied the oc I had before in bios, 2400 @ 3200 Mhz ...it worked



Maybe it´s my Gigabyte motherboard that´s flaky, idk?

Here is the beginning of the log file for windows xp pro (64-bit)

ntbtlog.txt

PHP Code:

---

 Service Pack 1 3  1 2010 02:12:49.109
Loaded driver \WINDOWS\system32\ntoskrnl.exe
Loaded driver \WINDOWS\system32\hal.dll
Loaded driver \WINDOWS\system32\KDCOM.DLL
Loaded driver \WINDOWS\system32\BOOTVID.dll
Loaded driver ACPI.sys
Loaded driver \WINDOWS\system32\DRIVERS\WMILIB.SYS
Loaded driver pci.sys
Loaded driver isapnp.sys
Loaded driver pciide.sys
Loaded driver \WINDOWS\system32\DRIVERS\PCIIDEX.SYS
Loaded driver MountMgr.sys
Loaded driver ftdisk.sys
Loaded driver dmload.sys
Loaded driver dmio.sys
Loaded driver volsnap.sys
Loaded driver PartMgr.sys
Loaded driver atapi.sys
Loaded driver iaStor.sys
Loaded driver jraid.sys
Loaded driver \WINDOWS\system32\drivers\SCSIPORT.SYS
Loaded driver disk.sys
Loaded driver \WINDOWS\system32\DRIVERS\CLASSPNP.SYS
Loaded driver fltMgr.sys
Loaded driver sr.sys
Loaded driver dmboot.sys
Loaded driver KSecDD.sys
Loaded driver Ntfs.sys
Loaded driver NDIS.sys
Loaded driver Mup.sys
Loaded driver crcdisk.sys
[COLOR="#FF0000"]* * * * * * * * * * * * * * [/COLOR]
Did not load driver ACPI Multiprocessor x64-based PC
Did not load driver Audio Codecs
Did not load driver Legacy Audio Drivers
Did not load driver Media Control Devices
Did not load driver Legacy Video Capture Devices
Did not load driver Video Codecs
Did not load driver WAN Miniport (L2TP)
Did not load driver WAN Miniport (IP)
Did not load driver WAN Miniport (PPPOE)
Did not load driver WAN Miniport (PPTP)
Did not load driver Packet Scheduler Miniport
Did not load driver Direct Parallel
Did not load driver Audio Codecs
Did not load driver Legacy Audio Drivers
Did not load driver Media Control Devices
Did not load driver Legacy Video Capture Devices
Did not load driver Video Codecs
Did not load driver WAN Miniport (L2TP)
Did not load driver WAN Miniport (IP)
Did not load driver WAN Miniport (PPPOE)
Did not load driver WAN Miniport (PPTP)
Did not load driver Packet Scheduler Miniport
Did not load driver Direct Parallel
Did not load driver Intel Processor
Did not load driver Intel Processor
Did not load driver Intel Processor
Did not load driver Intel Processor
Did not load driver Audio Codecs
Did not load driver Legacy Audio Drivers
Did not load driver Media Control Devices
Did not load driver Legacy Video Capture Devices
Did not load driver Video Codecs


...

[I]und zu weiter...[/I] 


---

Is it the ACPI that´s flaky?

Quote:

---

Did not load driver ACPI Multiprocessor x64-based PC

---

I will try to attach the whole 2.6 MB file but I guess it´s too big, edit, I was right as the .txt filesize limit is 200 kb, but it worked to .zip it to 42 kb

I googled error message above and found something, maybe relevant or not, as it a laptop. In that post it seams like it could be related to memory, thus with one stick of memory it works.

http://vip.asus.com/forum/view.aspx?...Language=en-us

Quote:

---

Well I finally went back to Windows XP 64bit and eventually figured out how to get it to boot up without hanging - removed the two 2Gb DIMMS and replaced with original 512Mb DIMM. Windows then booted up no problem with the 9600GT card in place. Then could install full Geforce video drivers and reboot. Then tried replacing the 512Mb DIMM with a single 2Gb DIMM - windows booted up fine again with this. Then tried putting the second 2Gb DIMM back in (to make 4Gb total) - windows hangs on startup - to be more specific while it loads the ACPI - multiprocessor HAL Halaacpi.dll See here for more info on hardware abstraction layer - http://support.microsoft.com/kb/309283. Any idea how to debug this problem so i can get the second DIMM in operation. p.s i don't think there is anything wrong with the DIMM - think it's some sort of memory access/addressing issue between GPU, CPU and bus system.

---

I got 2x2GB DDR2 system memory, see sig, in two slots for a total of 4GB, and have been looking the last days to buy two more sticks for at total of 8GB but maybe it will get worse if it´s the memory controller that´s flakey during boot.



Well I try some more google and maybe Ill find something more specific for my computer.

I have rebooted a couple of times today without any problems (crash etc) then I upgraded my 4 GB DDR2 memory and installed another 4 GB (2 x 2 GB) and there was some problems.

The computer started and it looked like it was okay but I wanted to make sure it was really stable and test with AIDA64 portable (installed on the Liberkey in a folder on d: ).

Started AIDA64 and looked at the summary of memory and it showed it was populated with DDR2 in slot 0,2 : 2 x 2 GB and slot 1,3 : 2 + 2 GB.

Then I checked a few more summaries and then the computer froze (screen froze).

Rebooted into bios and set the memory to "auto" instead of the +0.20V DDR2 that I had before (imagined that it made the computer more stable when rebooting, not so sure any more).

before I set it to "auto"
http://img254.imageshack.us/img254/7814/dsc14276.jpg


Rebooted into Windows XP Pro 64-bit in "safe mode" and it found both the raid 10 volumes c: d: had to be verified. I set d: to "normal" and let it verify c: (done in 15 minutes without any errors).

Rebooted normal start of windows, it started okay but then the Liberkey alerted that it had crashed and had a corrupt file (usually Liberkey fixes the it if Firefox etc crashes but this time it offered no help).

Started raid 10 verify of volume d: (which would take 6 hours and usually takes 3 hours but this time it took 6 hours and found no errors at all)



I was thinking of unplug the raid (all cables to eliminate raid problems) and rebooting into bios and start with the 1 TB Samsung SATA NON-RAID with Windows XP Pro 32-bit and the run the AIDA64 and test the memory, that was my plan.

or some other method to "unplug" raid without physically unplug the cables.



Then I suddenly remembered that it only supports 4GB DDR2 (about 3.5 GB accessible) thus making it impassible to test the full power of 8 GB DDR2.

Damn, all this writing to conclude that I have to live with the "raid delay" if something goes south.



As noted the computer has been stable for 6 hours with Firefox 10 and 12 tabs, VLC music and video on H: ...some memory stability, but I didn't want to stress it with a game like UT2004.


edit 1: rebooted normal start of windows and the chkdsk started to verify d: and found... damn, my sony F707 has to be run with windows xp 32-bit drivers. I'll post the photos later, anyway it found several errors which it fixed. Liberkey started without any "alert-ballons".

edit 2: When I checked Liberkey it has lost the "recent" list and my user installed apps on it. Damn I did a ERUNT and a "restore point" after the crash and installation of DDR2 (not before, stupidme). Going to check if there are any auto created but I doubt it.

edit 3: The ERUNT had a backup from today, 6 april 2012, 18:58:39, before the upgrade but that will not help the Liberkey as it is portable and has its settings not in the registry.
The "System restore" had a restore point from earlier today, 6 april 2012, 11:03:48, but will that help?


edit 4: Going to test the 8 GB DDR2 with memtest86+ version 2.11
...forgot how, but I think it was run from a floppy, before windows was started. Hmm, wait, does it support 8GB?

edit 5: I researched and found out that there are never versions available that supports 8 GB and multi core like my Q6600

Memtest86 wiki
Version 3.5a corrects a bug in 3.5 that prevented testing more than 4 GB of memory. Version 3.5a now works with up to 64 GB of memory.
http://en.wikipedia.org/wiki/Memtest86#Known_problems

Memtest86 Release 4.0a Multi-threaded with up to 16 cores, tests up to 64 GB of memory
A number of conditions existed that potentially cause problems when testing more than 4 GB of memory. Most if not all of these conditions have been identified and corrected.
Ver 4.0a - Windows Floppy disk package
Ver 4.0a - USB key image (Only installable from Linux)
...It seams there is no usb-key support?
http://www.memtest86.com/

memtest86+ 4.20
there are a long changelog list but nothing about more than 4 GB memory but
New installer for USB Key
...great, usb-key support
http://www.memtest.org/

I have 4 usb-keys, 3 x 2 GB and a 16 GB. I wanted to use the 16 GB but it's not bootable. I remember correctly only one of the 2 GB is not used but Its a pity to waste a whole usb-key for a small memtest program (don't know if it's then useable for other stuff, I recall that there were some program that could do a boot-menu for a usb-key thus making it possible to have several ISO-images on it etc)

edit 6: update to edit 1: as I copied the photos but it's to late now at 07:00 AM.

edit 7: the computer is still running 15 hours later. Here are the blurry photos that has been processed in PS

http://img235.imageshack.us/img235/926/takapicyc7.gif

http://img17.imageshack.us/img17/378...nrexpq801b.jpg


The "myapps" and "recent" probably belongs to the Liberkey that crashed.
http://img193.imageshack.us/img193/6...26usmq801b.jpg

There were also a stage 3 screen, but I wasn't quick enough with the camera, but what I remember it was no further errors.


PS I posting this while the computer is still running and will edit this post for more info. Gaaah, I thought I could fix this tonight and now at 07:00 AM the whole night is gone. Google is a black hole where the time disappears.

I'm wondering if a rollback with "restore point" is the way to go now and then create the memtest86 floppy and usb-key and then do the testing?

edit 1: no the rollback didn't help for the lost data in Liberkey, but it wasn't much so I leave it this stable configuration.

edit 2: created a memtest86+ 4.20 floppy. Rebooted and ran the test for an hour and it passed the first pass with no errors. Will post photos later. Feel more confident now that the DDR2 is stable with 8 GB.
http://i51.photobucket.com/albums/f3...r/DSC14328.jpg
http://i51.photobucket.com/albums/f3...r/DSC14329.jpg
http://i51.photobucket.com/albums/f3...r/DSC14330.jpg

memtest86+ Pass 1, 1 hour, No errors
http://i51.photobucket.com/albums/f3...r/DSC14331.jpg

Have also played some games for several hours, UT2004, UT3, Crysis, Paradise City TUB, GRID etc. without any crashes, except once with UT3 and Tornado-PhysX map
...but I think it´s the game itself that crashed.
I wanted to try out the 8 GB and ran a test with PS CS5 but it only supported 3.2 GB :( as it was 32-bit, anyway I did some testing with a 400 MB .tif image and it worked great except it ran out of RAM several times.


edit 3: tried to create a memtest86 4.0a floppy but it didn't support 64-bit OS. Found out later on their homepage.

Calling All Developers!
A number of additions are planned for the Memtest86 diagnostic and also for the Memtest86 website. We need your help! Please contact us about contributing to the Memtest86 community and earn a little cash to boot. Listed below are a few of the planned projects.

Native 64 bit support with wider patterns. With 64 CPUs becoming common a native 64 version of Memtest86 is needed. Test algorithms need to be updated to use 64 and 128 bit patterns. A simple flat memory model can now be used that will eliminate a great deal of complexity and add efficiency.

The computer crashed ugly last week, 30/5, when it just died as I left it and was making a sandwich in the kitchen while I was listening to some music playing, VLC media player on repeat, when it was supposed to repeat it went silent. I came back some minutes later with the sandwich and both the tv-out and the monitor had lost signal and the computer was running but dead.

When the computer was restarted raid bios for the RAID 10 had found volume c: and d: DEGRADED

Started up into windows xp and IMSC (Intel Matrix Storage Console) which comfirmed it and the SATA port 0 had one of the four hard disk members in the RAID 10 gone bad.

Normally you then shut down the computer and replace that hard disk but there is a way to let IMSC recheck the "bad" hard disk.

I Right clicked on the port 0 hard disk and set to "marked as normal".

Then the rebuild starts, which took about 4 hours then it was fine again with both volume c: and d: had the status "normal" in IMSC.

Have had it run several programs, games etc last week and everything was fine.


Saturday, 2/6, it crashed while playing UT2004, and some music from youtube on repeat, I didnt notice if it was when the music restarted but I did notice that it was when I entered a Hellbender and the engine started plus some other sound effects.

It totally froze, dead.


Restarted the computer but the raid bios didn't show degraded status but when it started windows xp it had to do a "verify and repair" on volume d:

I then continued to play UT2004 for an hour and then let it finish which took at least 6 hours. It did find 6 errors which it repaired.

It is now fine and "normal" again.




There must be something that ain't "normal" and stable in the computer...



Have tested the 8 GB of DDR2 with a memtest86+ 4.20 floppy. Rebooted and ran the test for almost 4 hours this night and it passed 2 pass with no errors.. The 8 GB DDR2 is totally stable I would say.



I can post S.M.A.R.T information from the raid 10 hard disks with CrystalDiskInfo.

The general health status says "Good" on all hard disks.


Which are the most important data to look for?



PS I didn't bother to take any photos or screenshots of the last week but it looks about the same as the previous posted photos and screenshots.

The raid 10 crashed tonight, 45 minutes ago, almost the same way as described in post #31


port 0 is red - Status: "the hard disk stopped working"
http://i51.photobucket.com/albums/f3...-error--1a.jpg

Crystal disk info S.M.A.R.T
http://i51.photobucket.com/albums/f3...sk-info-1b.jpg


I have also had another crash last week but it didn't take down the raid at all, not even a verify/rep. was needed.



Strange nonetheless that I get these computer crashes...

http://i51.photobucket.com/albums/f3...r--7-7--2a.jpg



http://i51.photobucket.com/albums/f3...r--7-7--2b.jpg


full size
http://i51.photobucket.com/albums/f3...r--7-7--2a.jpg

http://i51.photobucket.com/albums/f3...r--7-7--2b.jpg

The computer has been very stable for about 4 months, not a single crash which is vey good as it was something that happenend about every two weeks before.

...until today and it crashed the same way as previous post.

One thing though is that last week I ran defragler on d: (1.7 TB volume) which but it took too long, after 12 hours it said it was 1 > day left, I aborted it and installed PerfectDisk 12.5 Pro and defraged in about 5 hour. It has been runnig fine for a week and I hav e been playing many many hours of UT2004 without a single problem. Today the computer was running all day but only Firefox 20 and some lowbandwith videostreams. Nothing unusual and then.... freeze.


Im wondering if the harddisk stress last week has caused some sort of stability issues today?


I have done 41% of the rebuild of volume d: (its about 1 hour 22 minutes left to do)




edit: 2014-02-17

current sig.

After installing Win 8.1 64-bit I have had some crashes very similar to this -->post #20
https://communities.intel.com/thread...rt=15&tstart=0
Now it is Intel's business to find the reason for this shutdown bug and to solve it. post #15
"3. (more common) Windows shuts down, I can hear my HDDs shutting down one by one. When the last is supposed to shut down the system reboots (maybe just after it shuts down or just before, I'm not sure). When the RAID BIOS is supposed to show up on screen I can hear my 5 HDDs powering up SLOWLY one by one (the next HDD won't power up until the previous has complete the spin up). Then windows starts and from there I can shut down regularly."

When it tries to detect the Samsung F1 the text is blinking RED (not found or something) but when windows starts it is there and working fine.

I have had some crashes and when rebooting to WinXP or WinXP 64-bit there waw this...
http://www.pcreview.co.uk/forums/chk...ot-t72388.html
CHKDSK runs every time I boot!
"I was having this problem until today. I used this simple fix. It might
help you.

1. click on your start menu and open the run dialog.
2. type "cmd" and return (note: dont enter quotes)
3. Next type "fsutil dirty query e:"
4. If the return message indicates that the volume is dirty go to step 5
5. Next type "chkdsk e: /f /x"
6. After that finshes repeat step 3.
7. If it is no longer dirty then reboot and you should notice no more
ckdisk."

Tried the fsutil and it showed all hard disks and its partitions was DIRTY and then fixed it as described above.

Never had this happening before I installed Microsoft Win 8.1 64-bit and its raid driver.

Have downloaded the latest IRST but after reading about alot of problems I haven't installed it yet and not running Win 8.1 64-bit more than a few times a month since november 2013 and installed a backup program TrueImage 2014. I will do a backup when the OS is stable with a good raid driver.








Gigabyte P35-DQ6 - rev 1.0, F7 bios
Intel Q6600 G0 - 2.4 @ 3.200 Ghz, 400x8, Vcore 1.300V (Kentsfield)
ThermalRight Ultra EXTREME 120 + Noctua NF-P12 120mm fan
Corsair XMS DDR2 PC6400 5-5-5-18 2 x 2x2048 8GB kit @ 800MHz +( default )V in bios
Corsair HX-620W PSU | mb:Realtek HD audio 7.1 mb
Intel ICH9R - OROM 7.5.0.1017
Intel Matrix Storage Manager 8.9.0.1023 (IMSM) with volume c:128GB, d:1.7TB
SATA: 0-3:4x1TB Samsung Spinpoint F3 in RAID 10, 64k stripe
nonraid SATA: 4:250 GB Samsung SSD 840 EVO, 5:1TB Samsung Spinpoint F3
Gigabyte SATA2/Jmicron - OROM 1.06.59
nonraid SATA: 0:1TB Samsung Spinpoint F1 1:
usb3:Silverstone EC04P- (1x-pcie)
nonraid SATA:Rocket 620 (4x-pcie) 0: 1:
MSI Geforce 780 3 GB Twin FrozR Gaming (16x-pcie)
27" QNIX 2710LED @96Hz PLS 2560x1440 | case1: CM Stacker T01
OS: 1:Windows XP Pro, 2:64-bit 3:Win 8.1 64-bit
VMware7: 2a:Win 7 64-bit, 2b:Win XP Pro

Boot order

1. 0:WinXP-1TB, 0:Win64-R10, 0:wProg-R10, 5:PS R10-SP, 4:Win8.1-SSD
2. 0:Win64-R10, 0:wProg-R10, 5:PS R10-SP, 4:Win8.1-SSD, 0:WinXP-1TB
3. 4:Win8.1-SSD, 0:Win64-R10, 0:wProg-R10, 5:PS R10-SP, 0:WinXP-1TB