The Old Joel on Software Forum: Part 4 (of 5) - Persistent errors on NTFS Volumes

Persistent errors on NTFS Volumes

I have three computers configured with NTFS. A WinXP Home machine and two running Win2Ksp4. Every time I run chkdsk on any of these machines, errors are found, and then supposedly fixed at next bootup, but they return immediately. I've taken to running the Recovery Console, where I can run chkdsk in repair mode multiple times in a row. The first time it will find errors, but not on subsequent runs. Then as soon as I exit the console into the OS proper chkdsk again finds errors.

I've done a complete system scan for Viruses with the latest virus defs from Symantec, and I've run the latest Ad-Aware defs to search for malware.

All machines are different makers, with different HD manufacturers. One is a laptop. And I never hit reset, or "just power down", I always use the Shutdown command.

Any ideas? Much obliged.

Ken Klose
Wednesday, February 18, 2004

I had an Athlon machine with a newish motherboard (at the time) that did this every time I booted Win2k, regardless of service pack. Finally updated the machine to XP and *poof* it all went away - XP obviously had some better integrated disk controller support for that particular motherboard than did Win2k. I have no idea if you have the same problem but it sure sounds familiar.

Ivan Bootski
Wednesday, February 18, 2004

Problems with your power?

Flamebait Sr.
Thursday, February 19, 2004

The was a Via southbridge chip (which includes the IDE controllers) which caused disk corruption. I think it's the 686B. Yes, it was for the Athlons.

Brad Wilson (dotnetguy.techieswithcats.com)
Thursday, February 19, 2004

I had a problem some time back with an overclocked chipset.

It was basically the same swindle that PC Chips did with their motherboards, though the make was PC Partner.

The Cyrix processor had more or less disappeared at the time (early 2000) but there were still a load of chipsets around that had been built for it and ran at 83Mhz. What PC Partner did was buy a load and put them on a Motherboard with an adjustable memory bus speed, so that officially you could run the memory and FSB at 66Mhz, 100Mhz or 133Mhz. I had a new Pentium III 733Mhz (still working on the machine I'm using now) that ran at 133Mhz. For two months I had unaccountable freeze ups, and disk corruption; the machine went back to the shop three times but hours of checking found nothing (except that they had used the wrong cable for the HD's so they ran on PIO3 instead of UDMA).

It was only when I looked in detail at the results from Sandra, and saw that the PCI bus was running at %3Mhz and the AGP at 106Mhz, that I realized what had happened. And of course the IDE controller runs through the PCI bus.

Changed the motherboard and the problem went away.

If all three machines have the same hardware then you can be pretty certain it's a hardware problem. Check it's not the disk controller by swapping out the machines. Check they haven't overclocked the chipset (bus speeds are changing all the time now so there must be slower busspeed chipsets around for unscrupulous manufacturers to rebadge), using Sandra http://www.sisoftware.co.uk

Stephen Jones
Friday, February 20, 2004