Parity error in System Event log nvgts

[Novox]

Hello Sir,

I have using nForce IDE Driver version 10.3.0.27 in Windows Server 2003. I see the error report in the System Event log. It said "A parity error was detected on \Device\Scsi\nvgts1." I don't know how to fix it. This is abnormal?

Code: Select all

0000: 0f 00 10 00 01 00 64 00   ......d.
0008: 00 00 00 00 05 00 04 c0   .......À
0010: 02 00 00 00 00 00 00 00   ........
0018: 2c 22 12 00 00 00 00 00   ,"......
0020: 00 00 00 00 00 00 00 00   ........
0028: 01 00 00 00 01 00 00 00   ........
0030: 00 00 00 00 01 00 00 00   ........

Thanks!

[Fernando 1]

If you do a Google search for "parity error +nvgts", you will get a lot of links.
I don't believe, that the reported error was induced by the nForce IDE drivers v10.3.0.27. Nevertheless you may try to update the drivers to v10.3.0.42, which are part of NVIDIA's brandnew 15.23 driverpacks.

Another thought:
Do you have enabled NCQ? If yes, try to disable it and look, if the error message comes again.

[cbjaust]

I also get this same error in my event logs and have not figured a way to fix it. My driver version is 10.3.0.42 from the 15.23 driver pack.

I only have one hard drive (and it is on the NVIDIA nForce serial ATA controller) so I don't understand how I can be getting parity errors.

Windows Server 2003 SP2 x86
Asus M2N-SLI Deluxe (nForce 570 SLI)
AMD Athlon 6000+ (ADX6000IAA6CZ)
2GB OCZ Reaper HPC PC8500 (OCZ2RPR10662GK)
Western Digital 640GB Drive (WD6400AAKS)
NVIDIA GeForce 8800GTS 640

any more ideas?

[Novox]

If you do a Google search for "parity error +nvgts", you will get a lot of links.
I don't believe, that the reported error was induced by the nForce IDE drivers v10.3.0.27. Nevertheless you may try to update the drivers to v10.3.0.42, which are part of NVIDIA's brandnew 15.23 driverpacks.

Another thought:
Do you have enabled NCQ? If yes, try to disable it and look, if the error message comes again.

Hi Fernando,

You must believe it. Here is the picture. Clink to enlarge.

2551 is equal to 2008

[cbjaust]

If you do a Google search for "parity error +nvgts", you will get a lot of links.
I don't believe, that the reported error was induced by the nForce IDE drivers v10.3.0.27. Nevertheless you may try to update the drivers to v10.3.0.42, which are part of NVIDIA's brandnew 15.23 driverpacks.

Another thought:
Do you have enabled NCQ? If yes, try to disable it and look, if the error message comes again.

Hi Fernando,

You must believe it. Here is the picture. Clink to enlarge.

2551 is equal to 2008

http://img444.imageshack.us/img444/8397/nvgtsae7.th.jpg

Yep, I get the same errors and I'm using the newer driver v10.3.0.42



I have not been able to find very much info on this and I am sure that the drive is OK (command queuing in Enabled), plus it's wrong for it to be having parity errors because it's not set to be in RAID mode - the drive is clearly on the nForce serial ATA controller and not the RAID class controller.

I would like to ignore it but every time I boot up I get an error message about "one or more devices" but I guess that could be because Server 2003 doesn't wait too long for stuff to start up...

Anyway I might double check the BIOS and make sure none of the SATA channels are set to RAID mode.

Cheers

EDIT: I just disabled command queuing and rebooted and there is still an error at startup and still nvgts errors in the event log:



EDIT II: I got rid of the above error; the solution was to switch Performance Logs and Alerts service to run under the local machine account. KB article here. nvgts errors still there - a total of four after startup.

EDIT III: Checked the controller options in the BIOS and I had 4 of the 6 SATA channels enabled for RAID. I left the RAID controller enabled but I disabled the option for RAID for each channel. Rebooted and similar to EDIT III above, I still have four nvgts errors in the event log at startup. Time to disable the RAID controller in the BIOS (I'm not using it at the moment anyway).

EDIT IV:: Well with the RAID function disable in BIOS and no RAID class contrller showing up in devmgmt.msc it's still with the four nvgts parity errors in the eventlog after start up. Guess it's time to try earlier drivers that don't use nvgts. Can you tell which are the best ones to try?

Cheers

[TwL]

It's the new try for create the RAID-5 style prity which is failing miserably. I'd change the drivers as soon as possible as this will occure as crash on later on.

the drivers fernando suggested maybe good, but I'd take a shot with v10.3.52.4 since that serie of drivers.

-edit-

If this does not start functioning flawless. I'd look for old drivers(Not serie v10.x) and simply update those because that unfixed or well bugging will as I said crash eventually and take data with it.

-edit 2-

Hmm, to stop scaring people with lost of data. I could also say that this might as well be hoax.

As this could also be the unused SATA ports in your BIOS. Meaning, if you got unused sata enabled that might cause the error. However, above post indicates directly to Serial ATA port which is in use so kinda kills this idea, but considering parity as two or more it could indicate 'to rest of ports...' which would mean the empty ports.

at least try to disable the unused ports and see if it goes away wouldn't be such a bad idea.

[cbjaust]

>>snip<<

-edit 2-

Hmm, to stop scaring people with lost of data. I could also say that this might as well be hoax.

>>snip<<

I have a feeling that I am inclined to agree with you there - My drive is functioning just fine with abslotely no hint of any problem.

Cheers

[BenPi]

For the record...

I just saw in the system event log of my Windows Server 2003 box at home that since mid-October 2008 I have this famous "A parity error was detected on \Device\Scsi\nvgts1.". (The last time I updated my drivers was at that time.)

Since then it throws me this error hourly...

I have only two SATA drives connected on the entire controller and they are not running in RAID mode.

The device connected on port0 (port0 = \Device\Scsi\nvgts1 ?) is running in SATA1 and the one on port1 in SATA2.

So I will assume that's only a bug of the driver. But that's only me...

Anyway, I'll verify my BIOS config and update my post.


I have the drivers v10.3.0.42 coming from an installation of Fernandos XP 32bit WHQL Performance Pack v4.1 or Fernandos XP 32bit NF4-7 Performance Pack v4.1. (Don't remember which one I installed back then.)

Motherboard : Gigabyte GA-K8N51GMF-9 / nForce 430

[Fernando 1]

@ BenPi:

Welcome at nForcersHQ Forum!

So I will assume that's only a bug of the driver. But that's only me...

I agree with you.

I have the drivers v10.3.0.42 coming from an installation of Fernandos XP 32bit WHQL Performance Pack v4.1 or Fernandos XP 32bit NF4-7 Performance Pack v4.1. (Don't remember which one I installed back then.)

That is impossible, because none of my driverpacks ever contained the nForce IDE drivers v10.3.0.42. All my XP Performance Packs had and still have the nForce IDE drivers v9.99.09.

[caffinebassman]

Fernando1 wrote:

BenPi wrote:
So I will assume that's only a bug of the driver. But that's only me...

I agree with you.

Hi all,

Fernando1, how confident are you with that statement? I want to believe it but ....

After having lost 3 raid mirrors with the nForce Raid (previous drivers navatax64 etc) I gave up and manually mirror between my raptor drives - the error with those drivers was "delayed write failed" - I find it hard to trust the nv controller.

Anyway, I too have this parity error issue "A parity error was detected on \Device\Scsi\nvgts643" .. I found an interesting thing though, I disabled "write caching" and "command queuing" and so far the error has not come back of course disk performance sucks now...

My system is an EVGA 680 SLI based box (no OC'ing) with Core2 Q6600 and 2xWD1500ADFD and running XP x64.

Any hints or updates would be greatfully recieved ... before I through this motherboard in the bin and go for an intel chipset...