Intermittent boot failure and FIXMBR?

[Guest]

For a few weeks now, my PC almost always takes 2 attempts to boot: the first,
it stops at Verifying DMI Pool (which I think means it can't see the HDD);
press Reset and it boots fine. For the first time last night, I got the
dreaded “ntoskrnl.exe missing or corruptâ€, which usually means death...
pressed reset and it came back!

Now, normally at this point I’d say it was the disk, for sure. However,
neither Scandisk, nor Event Viewer, nor Western Digital’s diagnostics show
any disk problems... I’m stumped. It's worrying because I KNOW that one day
I’ll press Reset and it won’t... :-(((

I thought it might be a corrupt BIOS, so took the battery out and re-did all
the settings – then it worked. But it usually does on the 2nd attempt, so I
don’t think this proves anything. I’ll find out tonight, maybe.

I also thought maybe the MBR might be corrupt, so I started a recover
console, and tried FIXMBR... but it told me my MBR was either non-standard or
corrupt, and gave me dire warnings about proceeding, so I didn’t. Now, if
it's corrupt, great – I've found the problem. But if it’s “non-standardâ€,
perhaps due to my booting off a SATA drive (boot is C:, the first of 4
partitions on the SATA drive) rather than the IDE controllers, I might screw
everything up.

So my official MS-related question is – do you think I’ll wreck everything
if I do FIXMBR?

Unofficial responses to any other likely cause of the problem will be very
welcome too!

Thanks
Roger

 

[Malke]

For a few weeks now, my PC almost always takes 2 attempts to boot: the first,
it stops at Verifying DMI Pool (which I think means it can't see the HDD);
press Reset and it boots fine. For the first time last night, I got the
dreaded “ntoskrnl.exe missing or corruptâ€, which usually means death...
pressed reset and it came back!

Now, normally at this point I’d say it was the disk, for sure. However,
neither Scandisk, nor Event Viewer, nor Western Digital’s diagnostics show
any disk problems... I’m stumped. It's worrying because I KNOW that one day
I’ll press Reset and it won’t... :-(((

I thought it might be a corrupt BIOS, so took the battery out and re-did all
the settings – then it worked. But it usually does on the 2nd attempt, so I
don’t think this proves anything. I’ll find out tonight, maybe.

I also thought maybe the MBR might be corrupt, so I started a recover
console, and tried FIXMBR... but it told me my MBR was either non-standard or
corrupt, and gave me dire warnings about proceeding, so I didn’t. Now, if
it's corrupt, great – I've found the problem. But if it’s “non-standardâ€,
perhaps due to my booting off a SATA drive (boot is C:, the first of 4
partitions on the SATA drive) rather than the IDE controllers, I might screw
everything up.

So my official MS-related question is – do you think I’ll wreck everything
if I do FIXMBR?

Unofficial responses to any other likely cause of the problem will be very
welcome too!

Fixmbr is unlikely to help you since that is a software fix and your
problems are caused by hardware. The most probable cause of your issue
is a failing power supply. Swap out your power supply for a
known-working one. If that fixes things, replace the psu.


Malke

 

[Guest]

Thanks for your help – but... you say ARE caused by hardware, rather than
“may be†etc. So you're certain of this? Is that because it’s intermittent?

Thanks

 

[Rock]

Thanks for your help – but... you say ARE caused by hardware, rather than
“may be†etc. So you're certain of this? Is that because it’s
intermittent?

All the symptoms point to a power supply problem.

 

[Guest]

Wow… in all my years fixing PCs, I must have wiped or sometimes replaced hard
drives in over 200 Compaqs which wouldn’t boot, with this same or similar
error. Every time, it fixed the problem… and now it’s the power supply! I
can’t get my head round that one at all – I wish somebody would tell me why
all the symptoms point to the PSU, when everything I’ve ever done tells me
the symptoms point to the disk.

I grant you none of those other machines ever booted again until the disk
was wiped or replaced, and mine does, so I guess its intermittent nature is
why you diagnose the PSU? I just find it so hard to see why a dodgy PSU
causes something that looks like a disk problem. Or maybe it’s because it’s
had time to warm up a little between first and second boot?

Thanks

 

[Rock]

Wow… in all my years fixing PCs, I must have wiped or sometimes replaced
hard
drives in over 200 Compaqs which wouldn’t boot, with this same or similar
error. Every time, it fixed the problem… and now it’s the power supply! I
can’t get my head round that one at all – I wish somebody would tell me
why
all the symptoms point to the PSU, when everything I’ve ever done tells me
the symptoms point to the disk.

I grant you none of those other machines ever booted again until the disk
was wiped or replaced, and mine does, so I guess its intermittent nature
is
why you diagnose the PSU? I just find it so hard to see why a dodgy PSU
causes something that looks like a disk problem. Or maybe it’s because it’s
had time to warm up a little between first and second boot?

Your last paragraph hits it. Of course if you want to check before
replacing it get a multimeter and do a few checks. There were several posts
here over the past six months with detailed info on how to do this check.
Search for posts by, I think, it was Tom.

 

[Anna]

Fixmbr is unlikely to help you since that is a software fix and your
problems are caused by hardware. The most probable cause of your issue is
a failing power supply. Swap out your power supply for a known-working
one. If that fixes things, replace the psu.


Malke
--
Elephant Boy Computers
www.elephantboycomputers.com
"Don't Panic!"
MS-MVP Windows - Shell/User

Thanks for your help - but... you say ARE caused by hardware, rather than
"may be" etc. So you're certain of this? Is that because it's
intermittent?

Thanks

All the symptoms point to a power supply problem.

Wow. in all my years fixing PCs, I must have wiped or sometimes replaced
hard drives in over 200 Compaqs which wouldn't boot, with this same or
similar error. Every time, it fixed the problem. and now it's the power
supply! I can't get my head round that one at all - I wish somebody would
tell me why all the symptoms point to the PSU, when everything I've ever
done tells me the symptoms point to the disk.

I grant you none of those other machines ever booted again until the disk
was wiped or replaced, and mine does, so I guess its intermittent nature
is why you diagnose the PSU? I just find it so hard to see why a dodgy PSU
causes something that looks like a disk problem. Or maybe it's because it's
had time to warm up a little between first and second boot?

Thanks

Hogweed:
I'm inclined to agree with you in that it doesn't sound like a PSU problem
to me.

Our experience with that "dreaded ntoskrnl.exe missing or corrupt" Windows
message indicates that the cause of the problem could be either hardware or
software related. How's that for a start? Sometimes we've traced the problem
to a corrupt bios.ini file or some other corrupt system file; other times
we've encountered bad memory, or even a defective HDD, CPU or motherboard.
I'm hard-pressed to think of any incidents involving a defective PSU, but I
suppose that is a possibility.

I think at this stage you should consider a Repair install of the OS as an
initial step. I'm sure you know how to undertake that process. It's simple
enough to do, of course; the problem in doing so, as you again probably
know, is the annoying & time-consuming process of installing the various MS
critical updates. But we have found that a Repair install corrects the kind
of problem similar to what you're experiencing in a significant number of
cases.

Obviously if there's important data on the HDD involved, you should
disk-clone it as a backup so that you can retrieve the data should things go
awry.

If the Repair install doesn't work, we can go on to other things...
Anna

P. S.
May I suggest that in your future posts - should there be any - you include
sufficient background info re your problem and any pertinent responses
you've rec'd, either the postings themselves or a summary of them - just as
I've done in this post. Don't merely respond with your response/comment
without sufficient background info.

 

[DL]

I had a sys with sata in mirror config, a hd kept failing and the hd manu
tool reported the drive as failed. It was replaced under warranty. On the
third occassion I thought something else most be occuring. After
conversations with Adaptec (my sata controler manu) I replaced the psu,
which was also under warranty, from a recognised manu, problem went away.

 

[Malke]

Thanks for your help – but... you say ARE caused by hardware, rather than
“may be†etc. So you're certain of this? Is that because it’s intermittent?

Thanks

Of course I'm not certain. I can't see your computer and I can only rely
on the information you've provided. You say you've already tested the
hard drive. It is very commonly a power supply problem when you push the
power button and nothing happens until you restart again. Sometimes it
even takes a few times. What happens is that the psu doesn't have quite
enough "juice" to get the hard drive up and spinning the first time but
the drive's momentum is enough to get it going subsequently. In any
case, it's easy enough to test by simply detaching the old psu and
attaching a known-working one. You don't even have to remove the old one
from its case.

I see later in this rather lengthy thread that you gave further
information which leads Anna (a poster for whom I have the highest
respect) to conclude that there is OS corruption. There certainly could
be OS corruption - and if you have bad hardware and were continually
starting Windows and having it fail, that could corrupt the data. I
would still test the psu. You may wind up having to do a Repair Install
or even a Clean Install, but my first troubleshooting step would not be
that.


Malke

 

[Rock]

Wow… in all my years fixing PCs, I must have wiped or sometimes replaced
hard
drives in over 200 Compaqs which wouldn’t boot, with this same or similar
error. Every time, it fixed the problem… and now it’s the power supply! I
can’t get my head round that one at all – I wish somebody would tell me
why
all the symptoms point to the PSU, when everything I’ve ever done tells me
the symptoms point to the disk.

I grant you none of those other machines ever booted again until the disk
was wiped or replaced, and mine does, so I guess its intermittent nature
is
why you diagnose the PSU? I just find it so hard to see why a dodgy PSU
causes something that looks like a disk problem. Or maybe it’s because it’s
had time to warm up a little between first and second boot?

There are no certainties. And of course you have to make a decision which
way to go. In a situation like this where it could be either, and it's
relatively straight forward to test the hardware aspect, I personally would
eliminate the hardware issue first.

 

[Curt Christianson]

Hi Hogweed,

I agree with the others in the PSU as being a possible or even likely
culprit

As you already mentioned, disk-checking utilities yielded nothing. I don't
know if you've added any addition components to further tax your PSU, but
even if you may have:


1) An ailing power supply may not be able to delievery to current need to
get a disk up and running to speed within the time constraints is has
available.

2) A resonable "normal" PSU (still within specs) may still delivery what
it's designed to do, but a disk that is beginning to show it's age may be
drawing excessive current initially in an effort, to get ther HD platters
running to speed within time. The greatest demand on a PSU is at that
moment when the HD's are striving to get to proper speed within a certain
time constraint. Once the platers are up to speed, the current demands of
the drive, and hence the load on the PSU drop to within more acceptable
limits. This would certainly explain subsquent boots woorking ok.

A very rough analogy might be trying to start a car in sub-zero weather,
where the oil has gotten thick, and the car battery is at best marginal,
causing a much greater load on the battery, compared after the car is
already is "wrammed up", and the engine spinning smoovely. A warm, loose
engine (or HD) is much easier to gettup about to full spinning sped, and has
a far greater chance of starting. Same wiht a HD drivr.. They can become
"stiif
' after cooling off due to in-activity, but once the "blodd starts pumpng,
they loosen up, an one ruuning only require a fraction of what starting
them from a cold, dead stop.

It's to bad one can noy go into a HD drive, and "do a grease,, oil ,and
filter, and all will be haplly agin.

Unfortunately, there'is is no way, short of a clean room repair to retrofit
a HD back to new-- and if you coulf, the cost would likely to be
prohibitive.

Even though the HD *does* appear to boot upsubsequently, if no additional
hardware has been added to creatre additional strain on the harddrive, I t
may be time to start looking for a new drive, and/or PSU

Power supplies, and HD's are most often seen as the most immediate cuprits
insituations like this, but in great may of instances, it's often a
combination of both.

If it comes down to wanted to replace the PSU, you wwill never go wrong,
The 175 watt PSU of your would not stand a chance with the demands of todays
mobo's an their associated processores, graphics consume power they way a
growing teenager does (no end in sight), More and more peripherals are
added alsmot everday. Where a 275 Watt PS may have been over-kill 10 years
ago, a 500W (or better) is not way unreasonable nowadays. Pereipherals and
add-ons directively rise to meet or exceed machines capabily, jst as income
and expensives do in a somewhat similiar vein..

On final thought, it's the sum-total of all ad-ons (those after the
originally setup from the vendor, that have *likely cause this situation'.
Un fornunately, with the prices of HD's pummentind manufacturers are trying
to cut non-essential corners in an efort to save money, and maintain a
profit margin, margin they can live with, without selling junk )
The first think to skimp on is the power-and they do.

--
HTH,
Curt

Windows Support Center
http://aumha.org/

 

[mikeyhsd]

I had a similar situation recently, thought it was the PS, turn out to be a memory stick going bad.
replaced the 2 sticks and all is well.



(e-mail address removed)

Thanks for your help – but... you say ARE caused by hardware, rather than
“may be†etc. So you're certain of this? Is that because it’s intermittent?

Thanks

Of course I'm not certain. I can't see your computer and I can only rely
on the information you've provided. You say you've already tested the
hard drive. It is very commonly a power supply problem when you push the
power button and nothing happens until you restart again. Sometimes it
even takes a few times. What happens is that the psu doesn't have quite
enough "juice" to get the hard drive up and spinning the first time but
the drive's momentum is enough to get it going subsequently. In any
case, it's easy enough to test by simply detaching the old psu and
attaching a known-working one. You don't even have to remove the old one
from its case.

I see later in this rather lengthy thread that you gave further
information which leads Anna (a poster for whom I have the highest
respect) to conclude that there is OS corruption. There certainly could
be OS corruption - and if you have bad hardware and were continually
starting Windows and having it fail, that could corrupt the data. I
would still test the psu. You may wind up having to do a Repair Install
or even a Clean Install, but my first troubleshooting step would not be
that.


Malke

 

[w_tom]

Your last paragraph hits it. Of course if you want to check before
replacing it get a multimeter and do a few checks. There were several posts
here over the past six months with detailed info on how to do this check.

In another thread in 24hoursupport.helpdesk, Alice replaced a power
supply and still power was unstable. Was the second power supply also
defective? Was load too large for both supplies? After all that
swapping, still, nothing learned. Then Alice took readings with a 3.5
digit multimeter. Problem immediately traced to one part of the
motherboard.

That procedure was posted - with unique voltage numbers - in "When
your computer dies without warning....." starting 6 Feb 2007 in the
newsgroup alt.windows-xp at:
http://tinyurl.com/yvf9vh

Either power supply 'system' is defective, or move on to other
suspects. Swapping power supplies cannot provide that necessary and
definitive 'yes or no' answer.

Anything can appear defective if a power supply 'system' is
failing. A defective power supply that does not yet create failures
can also be identified using this same procedure.