Question : Problem: $Mft Writeback errors on new WD160gig HD

After a few hours of usage I get this error. A reboot seems to correct and scandisk is also run each time.
I have searched the Western Digital support but no help. This is a new drive and the last Western Digital I will ever buy
again. I have all updates for mass stoage over the 138 gig limit. WD HD tools comes on a floppy disk and I have no access to such legacy hardware. The software must be installed via 3.5" floppy on boot-up, so it is worthless. To cheap to include a bootable CD. 3 days after my purchase they charged me $14.95 for tech support to answer a question that was omitted from the instruction manual which is rife with ambiquities.
I have Maxtor HD's and Seagates with never a problem. Can anyone tell me how to fix this recurring problem?

Answer : Problem: $Mft Writeback errors on new WD160gig HD

boopsie - The figure I gave of '50 power cycles' is an approximation. They have been as low as 25 on some drives, and as high as 60 on others. The factories decide (based on reported problems, returns, company policy, etc.) on-the-fly, so it's not even consistent between identical drives from the same plant. To make matters even more confusing, some OEMs can adjust those settings as well after they leave the factory.

reboots vs. power off/on : I honestly don't know. We've never got an engineer from one of the drive manufacturers to uncloak and give us some hints in any of the on-line discussions. And others, including myself, have played around with these supposedly 'failing' drives (an Hitachi/IBM Deskstar in my case) and seemingly got them past the factory-set forced SMART checks by power-cycling them with the drive's power connector (plugging in and unplugging the drive). I would NOT reccommend you do this - I was using a bench power supply and the drive was not connected to the computer's IDE cable. If you do this inside your computer you not only risk the drive, but the power supply and anything connected to it. Neither the power supply or drive is designed to be power-cycled in this fasion, and certainly not while installed in a system.

Others have speculated that the internal counter is only affected by an IDE initialize signal during boot, i.e., the drive has to be connected in a computer going through the boot process. The thought was that someone could write a little utility to send a series of these signals to the drive and advance the counter. None was ever produced that seemed to work consistently, but (theoretically) rebooting your computer would produce the same result. It's fairly easy to write a little program that will reboot your computer over and over, but it seems like a lot of wear and tear on the poor thing just to (possibly) advance some mystery counter inside a new drive.

Now (to us) the real solution would have been for Microsoft to fix XP. They deny any problem and the throttling-down of the ATA-speed is apparently their one and only effort to 'fix' IDE problems. Drive manufacturers could also have provided a utility to examine or turn off the behavior, but 1) everyone who buys their drives does not necessarily use XP, 2) they'll take the hit for some unnecessarily- returned drives because it probably prevents more problems in the long run than it causes, and 3) drive manufacturers have always been very secretive about SMART registers - they're afraid that consumers will interpret hundreds of remapped bad sectors on a new drive (which is quite normal) as 'junk' drives.

So.... If you normally leave your computer ON all the time, I would at a minimum turn it off at night so the counter advances. If you already do this, then it might be worth the effot to warm-boot it a few times as it's starting up. You don't have to let the O/S load - the computer just has to get through it's BIOS routines. In your case, this effectively has happened when you see the Promise controller's BIOS routines finish, but it still probably takes a few seconds for the drive to finish initializing. You could do something like F8 to interrupt XP before it starts loading.

Having said all of this, it's probably not worth the wear-and-tear on your computer to go overboard with rebooting. If you can play with any of the drive settings to eliminate the problem (even at the loss of a little speed) then you will be better off just letting the drive go through it's process through normal use. I would still suggest you get the WD utility to run in a few weeks so you can zero-fill the drive and start 'fresh'. If you still have the exact same problems after that, then I think it might be a better use of your time to swap brands in your computer and see if the WD will work in another one.  Keep in mind the problem is dependent on a number of factors, not just XP. It may work fine under XP in another computer because of some unknown timing or configuration differences. From the way you describe the problem though, I doubt the drive itself is 'bad'.

Good luck! Let us know how it all turns out.