Vista chkdsk on every reboot?

[nath]

Hiya all,

I just got back from a client who has a custom built* AMD system. It's been working fine for a year but as of a few days ago it's been performing a chkdsk every time it boots. It seems that vista triggers the "dirty bit" that I've seen certain forums refer to as the chkdsk runs, completes then reboots the system and vista boots in fine. Everything seems fine in there, and I've not heard any clicks of death so I'm not sure the hard drive is actually faulty or not. It's not on SP1 yet, but when I did some updates there appeared to be a SATA interface driver update that didn't fix the problem.

I'm going to be talking with this guy this evening to discuss where to go from here, I've a feeling it could be a replacement hard drive job but I don't want to do anything unnecessary. Anyone got any ideas of other things it could be? I'm sort of hoping that service pack 1 will resolve the issue too, but I think that's wishful thinking.

Any thoughts (relating to this subject.. and to boobs if you feel like veering off topic) are greatly appreciated.








*Not by me

 

[nath]

Grr - edit limit expired.
Just thought of something else. After installing the updates - the first reboot it didn't do a chkdsk, then proceeded to continue as usual on subsequent reboots. So it doesn't appear to be a permanent fixture, but something windows is telling the system to do most of the time.

 

[Kryten]

I've seen that before somewhere, but it was easily fixed with a look in msconfig , there was an entry somewhere as if it had been enabled by an option in Vista itself somewhere. I'd long have imagined you'd already thought of that/had the chap involved check it, so I guess it'd be something else, unrelated.

Quick google around, this comes up:

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.

Another look brings this up, try this first:
Open command prompt with admin priviledges (run as, right click/run as admin)
chkntfs /d
That resets the dirty bit counter and if the drive's alright it shouldn't chkdsk every time.

 

[nath]

I did the chkntfs /d

The dirty query thing I didn't try, and msconfig I'm ashamed to say I didn't either. I always do my startup checks directly in the registry - localmachine run and currentuser run folders.

Given that the machine has been fine for a year or so, how likely is it that the motherboard has developed a fault? If it's a software issue that's fine, I'll get it sorted, if it's the hard drive, easily replaced but replacing the motherboard isn't likely to be cost effective. My thoughts are that it's highly unlikely to be the mobo, but any thoughts on that?

 

[Kryten]

Difficult to say, it's more likely to be the drive due to the mechanical aspect of it's operation. Electrics don't just tend to fail without reason behind it: but then again there can be lots of reasons - someone fitting something into the computer, knocking off a tiny yet seemingly important piece of the board whilst doing it, spilling a coffee into your PC (ahem, *) or kids inserting unwanted objects into places you shouldn't be sticking them.

But on the same note, what about the bit in-between - the cable? SATA or IDE, doesn't really matter, but well worth finding a spare somewhere and at least ruling that out.


*
I coughed somewhat there, as last night, I accidentally poured a small cup of orange juice into the top mesh of my PC.
I can be *extremely* thankful that it was a small cup and only a little bit of liquid, and *even more thankful* that the fan in said mesh was exhausting air, keeping all the moisture out of the PC Spot of good luck there

 

[nath]

Eep, close call there

Yup, I've thought about replacing the cable but it's looking to be a back to base fix, one I get it here I can take my time working out exactly what's wrong. I just wanted to try and rule out as many things that would make it either not worth fixing or not needing a back to base fix.

 

[nath]

Just thought of another thing, I couldn't get the system to do this in the short time I was there, but the client said that on a number of occasions it would just switch itself off while they were in the middle of browsing the web. From the way they described it, sounded like a graceful shutdown, not *flick* dead.

 

[Kryten]

A full, initiated shutdown would imply a software issue of some description but it's not unheard of for some of the better motherboards (and using some 3rd party software) to do a shutdown once the system hits a threshold temperature. Assuming temps are all OK currently though, I'd be doing the obvious anti-virus and other associated nasties checks, and looking in the event log. If it was graceful as you say, I'd imagine it would say what initiated it in there.

 

[Mazling]

I think I fixed a problem aaaagggeeesss ago by removing all alternate data streams from all files. Nothing was broken at all on the comp, but chkdsk would always report errors and 'fix' them on reboot for them only to 'reappear' straight away. I'm certain I used a different tool than this, but streams.exe in the SysInternalsSuite seems to do the same thing. FileForum | Sysinternals Suite
A quick google seems to give a lot of alternatives and info ;-)