Re: Reboot several times a day

On Thu, 26 Jan 2006 21:01:02 -0800, Leland Sheppard

>I don't know why that first drive has no temp. reading. I noticed that too.
>It and the one listed after it are both 120gb WD drives which I purchased
>the same day.

Wow, that is strange! Do they have the same board part numbers and
model numbers? Are they connected to the same sort of controller? By
120G, I'd expect all HDs to monitor temperature, it's usually 20G, 10G
and smaller that come up as --C.

>I buy drives in pairs because I use one to back up the other so I always get
>2 at the same time.

OK - it's RAID- and board-swap-friendly too ;-)

>Regarding memory - no. It's not ECC. Don't know how XP detects corruption...

Me neither. Even when there's parity error checking, it's nearly
always the BIOS that steps in with a Black Screen of Death, rather
than a polite Windows error dialog and log entry. That's why I'm
thinking that "memory error" is spurious, or possibly an "out of
memory" error that is almost equally likely to be spurious.

>Interestingly, I went through the device manager list and did an 'update
>driver' on EVERY device or pseudo device on the system. One device showed a
>driver problem - my HP PSC750XI - and I let XP update the driver from MS.

Ew... I wouldn't do that (as discussed a while back in another
thread). MS is the wrong place to get drivers, unless desperate.

>Since that driver was installed, I have not had a crash (knocking on the
>nearest piece of wood). I have gone as much as a week or 10 days without a
>crash before only to have them return so I'm not rejoicing yet. Cautiously
>optimistic is the phrase that comes to mind... It's been 6 or 7 days now, I
>think...

It could be, either by genuinely updating some driver error, orsimply
by replacing damaged code files for the same (or even later) driver
version. This is particularly likely if the HP device was added at a
time you had bad RAM; the original driver code files could have been
corrupted as they were being created. Bad RAM = baaad news.

>I will go get the memory test you mentioned. I'm not really too concerned
>about the memory. I got the error with the original memory and swapped it
>for a new batch and still got the error. The MS diagnostic says it is
>usually hardware; not always. I'm guessing this thing has been software.

Secondary damage.

Primary effects of bad RAM are transient, random, poorly
reproduceable, and typically cover all scopes and contexts.

But when this causes secondary corruption of the hard drive contents,
this secondary damage will often be limited to certain contexts, and
reproduceable within those contexts, e.g. mileage such as "whenever I
do abc, the system does xyz".

These effects will persist after the bad RAM is replaced, and then be
even more "static" (limited to specific scopes and contexts,
reproduceable) as the primary effects of bad RAM are removed.

Secondary damage can affect anything, because any disk read can be
bit-flipped to a disk write, and any disk address to be written to can
be bit-flipped to write to an arbitrary raw disk address instead.
However, it will most likely affect code created (installed) since the
RAM went bad - making "just re-install Windows" a disaster.

Doing a defrag can have the same effect.

In the case of a recently added hardware item, the drivers for that
item may be corrupted. A "repair install" done after the RAM is fixed
won't replace these code files, and if these code files are always in
effect, the result can be ongoing, broadly-scoped hell.

So you may well have "whacked the rat" - well done, if so!



>---------- ----- ---- --- -- - - - -
Don't pay malware vendors - boycott Sony
>---------- ----- ---- --- -- - - - -
.

Relevant Pages

  • =?windows-1252?Q?www=2Eelectronicsbestseller=2Ecom____HP_Pavilion_DV9650U?= =?windows-1252?Q
    ... GeForce 8600M graphics (up to 1023 MB of available memory) ... RAM: 2000 MB ... external DVD or music buttons to launch HP QuickPlay (which bypasses ... drive space from two 160 GB drives. ...
    (de.comp.sys.notebooks)
  • Segfaults in mkdir under high load. Software or hardware?
    ... I have tried to eliminate the RAM by letting memtest86plus run overnight ... If we don't unpack more data than memory can hold ... usbcore: registered new driver usbfs ... usbcore: registered new driver hub ...
    (Linux-Kernel)
  • Re: M2N32-SLI Premium Vista wont post
    ... Understand that with no memory, I don't get the 1 long and 2 short beeps which say there is no memory. ... I can verify that this ram does work well with this board, I use it all the time and have never had a compatibility issue, on over 50 machines so far and counting. ... 2 lite on optical drives; dvd and ...
    (alt.comp.periphs.mainboard.asus)
  • Re: Reboot several times a day
    ... I never have had bad ram on this machine either. ... had memory available for another PC I'm building and just used that. ... >>driver' on EVERY device or pseudo device on the system. ... > time you had bad RAM; the original driver code files could have been ...
    (microsoft.public.windowsxp.general)
  • Re: determining speed using window xp
    ... I said nothing about the memory speed. ... RAM is random access memory. ... Drives ... Microsoft MVP Windows - Shell/User ...
    (microsoft.public.windowsxp.basics)