Author Topic: Hard disk woes (Read 360 times)

Grayarea · « **on:** June 02, 2002, 04:23:28 AM »

Hi all

I have a problem with my win2k system;

Abit KT7-RAID
2xIBM deskstar 38Gb ATA100 In RAID0 on Highpoint controller
Creative Labs Live Platinum
2x128Mb Crucial 133 CAS2
1x256Mb Cruical 133 CAS2
Phillips DVD

When saving large files to disk the system micro freezes and I get the following in the system alert log. Also I get these messages when I do a scandisk.

The device, \Device\Scsi\HPT3xx2K1, did not respond within the timeout period.

This would suggest a bad sector but scandisk does not make any differance. It does not seem to be getting any worse, but I am not sure how to proceed.

I read a some time ago that someone had to lowlevel format IBM disks when they had system lockups. This could tally because I had some system instability untill relaxing the memory timeings recently.

However I was under the impression that you could not low level IDE drives without killing them, or haveing he correct IBM software and a serial link to the drive.

The drive is still covered by warrenty but I don't really want to rebuild the system if I can fix he problem without doing this.

Any ideas?

Thanks,

MrRiplEy[H] · « **Reply #1 on:** June 02, 2002, 05:13:34 AM »

The problem is because your motherboard has a crappy chipset.

This is a typical problem of the 686B bridge. It's nice trying to make a backup of your HD.. LOL.

This and several other problems convinced me to change the motherboard to a KT266A one.

I had 30 GB of data corrupted in a raid-stripe that I couldn't back up because of this nice 'feature.'

Does your progress counter also go nuts after 50 or so MB transfer? Mine started to show 10031030103 hours untill the whole system bluescreened and transfer got corrupted.

Lephturn · « **Reply #2 on:** June 02, 2002, 06:25:35 AM »

Make sure you have the latest VIA 4-in-1 drivers. http://www.viahardware.com

Also, flash the latest BIOS... which should flash your highpoint controller's BIOS as well.

Also make sure you have the latest service pack and all updates/patches for your OS.

That device is your second onboard IDE controller made by HighPoint. If the above steps don't fix the problem, I'd try to return the board because of a faulty IDE controller. Worst case, replace the board as Mr. Ripley suggests.

MrRiplEy[H] · « **Reply #3 on:** June 02, 2002, 07:28:51 AM »

Lephturn what he described is part of the known bugs list of the 686B northbridge.. There's nothing he can do to fix it I'm afraid.

http://www.viahardware.com has a q&a about KT7 and it's problems.

Grayarea · « **Reply #4 on:** June 02, 2002, 08:00:34 AM »

Quote

Originally posted by MrRiplEy[H]
Lephturn what he described is part of the known bugs list of the 686B northbridge.. There's nothing he can do to fix it I'm afraid.

http://www.viahardware.com has a q&a about KT7 and it's problems.

Thanks for the responce guys.

MrRiplEy,

I have had a look around the FAQ you pointed me to but cannot find this specific issue. Could you post a link to the area where my problem is described please?

Thanks,

MrRiplEy[H] · « **Reply #5 on:** June 02, 2002, 08:17:02 AM »

The Q&A doesn't directly address your problem but let me assure you it's there because the 686B can't handle high PCI activity during the transfers. I did extensive research when I had big stability problems with the KT7.

Some useful clues can be found here anyway: http://www.viahardware.com/faq/kt7/kt7faq.htm

Animal · « **Reply #6 on:** June 02, 2002, 09:17:40 AM »

Dude, you are running two Deskstars on Raid 0

Thats just asking for trouble. Have you any idea how prone these drives are to failure? And having two of them depending on one another, you are doubling your chances of losing all your data.

Are you hearing any different noise from the disks sometimes? any kind of clicking or scratching? if you ever do, make a backup inmediately.

I dont recomend Raid 0, not to mention on Deskstars wich are very failure prone hard drives.

MrRiplEy[H] · « **Reply #7 on:** June 02, 2002, 09:19:35 AM »

I have to second that Animal.

Grayarea · « **Reply #8 on:** June 02, 2002, 09:53:10 AM »

Guys,

Yes they are in a stripe set, for the speed, and it is fast

Deskstars have a recent bad rep because they decided to use a glass platter, this has turned out to be less then durable, however the 38Gb drives I have do not use this technology.

I work as a Oracle DBA so I am aware of the pros and cons of differant RAID configs, I spend a fair amount of time explaining this to new customers for hardware scoping

The disks do click when the problem occurs, but if it is the disks and not the hipoint controller then why does scandisk not turn up and mark as bad the dodgy sectors?

Also all modern disks spare faulty tracks transparently, and assuming it has run out of spare tracks to spare then the problem, I would think, should be getting rapidly worse.

However the problem seems to be getting no worse and no better.

I am confused. I have built the last 3 systems I have used by hand so I know the ropes, but as MrRipley says the Abit board has been troublesome from day one.

If I can identify the issue 100% then I can move forward, but as it is I just don't know what the problem is, and more importantly I don't seem to be getting any nearer to finding the problem.