Vista Memory Diagnostics failing at 21%

[Guest]

I have 4 1gB sticks of Kingston memory in a Gigabyte GA-965p-ds3 ver 3.3,
with bios at ver F10, running at normal speeds. Due to a concern about faulty
data after doing an Easy Transfer from another system, I ran Vista Memory
Diagnostics in "Extended" mode (Standard mode tests ok). It fails at 21%
whether I run it on all four modules, remove two of the modules, replace them
with the other two modules, or run just one module.
Memtest86 runs ok.
Any ideas? I have even changed the BIOS settings to run the memory at 533mhz
(it is 667mhz memory).
Is there a way to view the MemDiag.bin file that is created in the
\Windows\system32\logfiles folder? All I get from Memory Diagnostics is a
message that my memory failed the test and I should go to my supplier!

 

[DevilsPGD]

In message <[email protected]> Kim

I have 4 1gB sticks of Kingston memory in a Gigabyte GA-965p-ds3 ver 3.3,
with bios at ver F10, running at normal speeds. Due to a concern about faulty
data after doing an Easy Transfer from another system, I ran Vista Memory
Diagnostics in "Extended" mode (Standard mode tests ok). It fails at 21%
whether I run it on all four modules, remove two of the modules, replace them
with the other two modules, or run just one module.
Memtest86 runs ok.
Any ideas? I have even changed the BIOS settings to run the memory at 533mhz
(it is 667mhz memory).
Is there a way to view the MemDiag.bin file that is created in the
\Windows\system32\logfiles folder? All I get from Memory Diagnostics is a
message that my memory failed the test and I should go to my supplier!

Try memtest86+ instead, it gives a lot more details about what and where
the failure occurred.

Have you tried placing in only a single stick, but in each of the
different slots?

Failing at 21% is odd, that should change when the total amount of
memory is reduced.

 

[Guest]

In message <[email protected]> Kim


Try memtest86+ instead, it gives a lot more details about what and where
the failure occurred.

Have you tried placing in only a single stick, but in each of the
different slots?

Failing at 21% is odd, that should change when the total amount of
memory is reduced.
--

I know, very odd. I have tried using various combinations, but not putting a
single stick in different slots. I am not sure it is always failing, here is
some more info:
First, I tried running the Diagnostics on the older computer that I am
upgrading from. Pentium 4, 2.53ghz, 1GB memory. It goes through the whole
test fine.
Second, I have been stopping the test when it gets to 21% because it sits
there for a long time (greater than 1/2 hr even with just 1GB in), whereas
the old system ran through each % pretty much proportionately. However, if I
let it go on, it does not fail - sometimes. Not very comforting.
Three, I emailed Kingston (the memory is their valueram) and they said to
RMA it,
Four, I have run Memtest86 on it and it passed, but I have not let it run
overnight.
Five, I am returning the memory to Newegg and instead of replacing it with
the same memory I decided to go to a higher performance memory.
Do you know how to interpret the memdiag.bin that is created by WMD?

I will probably run Memtest86 overnight just out of curiousity, before I
send it back.
Thanks for your help.

 

[Rock]

I have 4 1gB sticks of Kingston memory in a Gigabyte GA-965p-ds3 ver 3.3,
with bios at ver F10, running at normal speeds. Due to a concern about
faulty
data after doing an Easy Transfer from another system, I ran Vista Memory
Diagnostics in "Extended" mode (Standard mode tests ok). It fails at 21%
whether I run it on all four modules, remove two of the modules, replace
them
with the other two modules, or run just one module.
Memtest86 runs ok.
Any ideas? I have even changed the BIOS settings to run the memory at
533mhz
(it is 667mhz memory).
Is there a way to view the MemDiag.bin file that is created in the
\Windows\system32\logfiles folder? All I get from Memory Diagnostics is a
message that my memory failed the test and I should go to my supplier!

You've multiposted this to several newsgroups. In the future if you feel it
necessary to post to more than one newsgroup do it by crossposting, sending
one message at the same time time all groups. Multiposting which is sending
separate messages to each group, is counterproductive, leads to duplication
and folks in one group don't see the replies made in the other groups.

 

[Kim]

Sorry, Rock. I wasn't sure which group the orginal message should go in.
I'll just proceed in this group and not do it again.

 

[mikeyhsd]

you should just write it off as a virtual spanking by our resident net nanny.and get on with life.
no one is allowed to post anything unless he approves of it in advance.
no one else is allowed to present help or advise unless he approves of it in advance.
he even feels it is necessary to post contradicting or correcting other mvps. excusing it as they did not present sufficient info. when it just getting his name in lights again.

if you do not post using PERFECT english you are wrong.
if you do not post using PERFECT grammar you are wrong.
if you do not post using perfect PUNCTUATION you are wrong.
and worse heaven forbid if you fail to post in the proper news group you are denied an answer but told to buzz off and go to the proper news group.

our little net nanny seems to forget there are many people who post here whose first language is not english. and have a lacking of understanding on how things should work. more tolerance should be given with the prime purpose being to HELP PEOPLE and not shame them or tell them where to go.



(e-mail address removed)



Sorry, Rock. I wasn't sure which group the orginal message should go in.
I'll just proceed in this group and not do it again.

 

[DevilsPGD]

In message <[email protected]> Kim

I know, very odd. I have tried using various combinations, but not putting a
single stick in different slots.

The theory being "Test the slot" (as well as testing the RAM, and the
motherboard's ability to handle different configurations)

I am not sure it is always failing, here is
some more info:
First, I tried running the Diagnostics on the older computer that I am
upgrading from. Pentium 4, 2.53ghz, 1GB memory. It goes through the whole
test fine.
Second, I have been stopping the test when it gets to 21% because it sits
there for a long time (greater than 1/2 hr even with just 1GB in), whereas
the old system ran through each % pretty much proportionately. However, if I
let it go on, it does not fail - sometimes. Not very comforting.

Yeah. Unreliable hardware is the worst to try and track down I've
got some RAM here that is temperamental in one of my systems, but works
fine in others, but it usually passes diagnostic tests, but often fails
out a Windows install half way through. Grr.

Three, I emailed Kingston (the memory is their valueram) and they said to
RMA it,

Kingston is pretty good that way.

Four, I have run Memtest86 on it and it passed, but I have not let it run
overnight.

That's usually good news...

Five, I am returning the memory to Newegg and instead of replacing it with
the same memory I decided to go to a higher performance memory.

Sounds reasonable -- Start off leaving your system's configuration as-is
and see if it works before you start tweaking for performance, if the
problem happens again when you're under-specing the RAM, chances are
it's the motherboard (or CPU, if your CPU has an onboard memory
controller -- I don't believe the P4 chip does, but many newer processor
lines have a memory controller on-chip...)

Do you know how to interpret the memdiag.bin that is created by WMD?

That one, I do not...

I will probably run Memtest86 overnight just out of curiousity, before I
send it back.

Certainly can't hurt -- I've yet to see an error that doesn't crop up on
the first full pass, but I still put all new motherboards/RAM through at
least 24 hours of memtest86+ just in case...

 

[DevilsPGD]

In message <[email protected]> "mikeyhsd"

more tolerance should be given with the prime purpose being to HELP PEOPLE and not shame them or tell them where to go.

Indeed. In fact, look how you've attempted to help.

Oh wait, you didn't. Your post was just as useless.

There is a practical reason to crosspost rather then multipost. There
is a limited amount of time and brainpower available to help posters.
Anything which a poster does that reduces that pool of brainpower
decreases the overall amount of "help" available.

By crossposting instead of multiposting, everyone can see the same post,
avoid typing replies which are duplicates / offering duplicate
suggestions, and instead spend that time helping someone else or making
different suggestions rather then everyone retying the same first-step
suggestions.

*shrugs*

Only you can choose your participation level. You can help people, you
can spend your time whining, OR you can whine about the whining about
whining, your call.

 

[Rock]

In message <[email protected]> "mikeyhsd"


Indeed. In fact, look how you've attempted to help.

Oh wait, you didn't. Your post was just as useless.

There is a practical reason to crosspost rather then multipost. There
is a limited amount of time and brainpower available to help posters.
Anything which a poster does that reduces that pool of brainpower
decreases the overall amount of "help" available.

By crossposting instead of multiposting, everyone can see the same post,
avoid typing replies which are duplicates / offering duplicate
suggestions, and instead spend that time helping someone else or making
different suggestions rather then everyone retying the same first-step
suggestions.

*shrugs*

Only you can choose your participation level. You can help people, you
can spend your time whining, OR you can whine about the whining about
whining, your call.

Lol you just have to laugh at mickeyhsd. I replied to Kim in their thread
on the other newsgroup with several suggestions. They are aware of the
issue and will continue in this thread. This as opposed to mikeyhsd, a
quite pathetic character, who is generally of no help.

 

[mikeyhsd]

for sure I do not have a holier than thou attitude where I spend my time reading news messages and refuse to try and help those that do not meet my perceptions of grammar, punctuation, proper english or correct news group.
intentionally denying someone a answer for a problem because of this is WRONG.

you need to re-read the mvp code of conduct especially where it talks about being courteous.

(e-mail address removed)

In message <[email protected]> "mikeyhsd"


Indeed. In fact, look how you've attempted to help.

Oh wait, you didn't. Your post was just as useless.

There is a practical reason to crosspost rather then multipost. There
is a limited amount of time and brainpower available to help posters.
Anything which a poster does that reduces that pool of brainpower
decreases the overall amount of "help" available.

By crossposting instead of multiposting, everyone can see the same post,
avoid typing replies which are duplicates / offering duplicate
suggestions, and instead spend that time helping someone else or making
different suggestions rather then everyone retying the same first-step
suggestions.

*shrugs*

Only you can choose your participation level. You can help people, you
can spend your time whining, OR you can whine about the whining about
whining, your call.

Lol you just have to laugh at mickeyhsd. I replied to Kim in their thread
on the other newsgroup with several suggestions. They are aware of the
issue and will continue in this thread. This as opposed to mikeyhsd, a
quite pathetic character, who is generally of no help.

 

[Guest]

Rock, More info on the problem:

I have been running VMD with no failures. The room I work in has no
air-conditioning (we live at 9000 ft and really don't need it) so the temp
does vary from day to day and with the cold front that went through here
yesterday the room is about 5-10 degrees below two days ago. If the problem
is marginal, which is what it appears to be, this could explain it passing.
On the other hand, it still hangs for disproportionately long times at 21%.
For example, I have all 4GB back in (I have not sent them back, yet, since
the new ones won't arrive til later today) and it goes from 1% to 2%, 2% to
3%, etc at about 25-30 sec per %. When it gets to 21% it sits there for
approximately 2 1/2 hours before moving on to 22%. This normally wouldn't
bother me, since the programs warns that it may look inactive at times,
except that on my older Vista computer it sits on 21% for about the same time
it sits on the other percentages.
Regarding memtest86, although I have a lot of respect for the program, I
have been in the PC design and engineering business for 25 yrs and I get a
little paranoid about marginal problems. After seeing some of my data
corrupted after a "Easy Transfer" I want to make sure this memory is solid.
So, two questions: First, does Microsoft offer ANY support for VMD (Vista
Memory Diagnostics), so I can find out more about why it is hesitating at 21%?
Second, does Easy Transfer do any error checking while doing its job? Or,
maybe it does error checking and the problem was that during the read of the
data on the old computer it read the data wrong and created a CRC code based
on the wrong data so that when it put the data on the new computer the CRC
code checked ok? I.e., How could I transfer data using Easy Transfer from one
computer to another and end up with corrupted data, without it knowing it.
This seems, to me, crazy! (This is related, because it is what got me looking
into a possible memory problem to begin with).
Sorry for the long winded post, but this is a bit frustrating and would be
less so, if there was more tech info/support for these two products.
Kim

 

[mikeyhsd]

have you considered letting the mem diagnostic run over night.



(e-mail address removed)



Rock, More info on the problem:

I have been running VMD with no failures. The room I work in has no
air-conditioning (we live at 9000 ft and really don't need it) so the temp
does vary from day to day and with the cold front that went through here
yesterday the room is about 5-10 degrees below two days ago. If the problem
is marginal, which is what it appears to be, this could explain it passing.
On the other hand, it still hangs for disproportionately long times at 21%.
For example, I have all 4GB back in (I have not sent them back, yet, since
the new ones won't arrive til later today) and it goes from 1% to 2%, 2% to
3%, etc at about 25-30 sec per %. When it gets to 21% it sits there for
approximately 2 1/2 hours before moving on to 22%. This normally wouldn't
bother me, since the programs warns that it may look inactive at times,
except that on my older Vista computer it sits on 21% for about the same time
it sits on the other percentages.
Regarding memtest86, although I have a lot of respect for the program, I
have been in the PC design and engineering business for 25 yrs and I get a
little paranoid about marginal problems. After seeing some of my data
corrupted after a "Easy Transfer" I want to make sure this memory is solid.
So, two questions: First, does Microsoft offer ANY support for VMD (Vista
Memory Diagnostics), so I can find out more about why it is hesitating at 21%?
Second, does Easy Transfer do any error checking while doing its job? Or,
maybe it does error checking and the problem was that during the read of the
data on the old computer it read the data wrong and created a CRC code based
on the wrong data so that when it put the data on the new computer the CRC
code checked ok? I.e., How could I transfer data using Easy Transfer from one
computer to another and end up with corrupted data, without it knowing it.
This seems, to me, crazy! (This is related, because it is what got me looking
into a possible memory problem to begin with).
Sorry for the long winded post, but this is a bit frustrating and would be
less so, if there was more tech info/support for these two products.
Kim

 

[Guest]

I have been. It takes all night to do 2 runs.
Ok, so, new info:
I got the new memory, Crucial Ballistix, pc2-8000, and the same type of
results, except that it seems to also run very slow at 90% done.
It has to be one of two things, something going on in the motherboard (read:
memory controller) or VMD is doing things differently with DDR2 memory than
it does with DDR memory, such as doing severe fading testing with DDR2 where
it writes data and waits longer to check it. Just guessing since there is not
technical info from Microsoft. I am still a bit concerned, but since it
passes VMD (eventually, like in 3+ hours!!), and Memtest86, I am not sure
what to do. Have to stop testing for the next week, but at least the computer
seems to work well.
When I get back I'll try some other types of testing.
If you all have any other ideas, I will have my laptop with me, but won't be
able to try them until I get back.
Thanks,
Kim

 

[Rock]

Rock, More info on the problem:

I have been running VMD with no failures. The room I work in has no
air-conditioning (we live at 9000 ft and really don't need it) so the temp
does vary from day to day and with the cold front that went through here
yesterday the room is about 5-10 degrees below two days ago. If the
problem
is marginal, which is what it appears to be, this could explain it
passing.
On the other hand, it still hangs for disproportionately long times at
21%.
For example, I have all 4GB back in (I have not sent them back, yet, since
the new ones won't arrive til later today) and it goes from 1% to 2%, 2%
to
3%, etc at about 25-30 sec per %. When it gets to 21% it sits there for
approximately 2 1/2 hours before moving on to 22%. This normally wouldn't
bother me, since the programs warns that it may look inactive at times,
except that on my older Vista computer it sits on 21% for about the same
time
it sits on the other percentages.
Regarding memtest86, although I have a lot of respect for the program, I
have been in the PC design and engineering business for 25 yrs and I get a
little paranoid about marginal problems. After seeing some of my data
corrupted after a "Easy Transfer" I want to make sure this memory is
solid.
So, two questions: First, does Microsoft offer ANY support for VMD (Vista
Memory Diagnostics), so I can find out more about why it is hesitating at
21%?
Second, does Easy Transfer do any error checking while doing its job? Or,
maybe it does error checking and the problem was that during the read of
the
data on the old computer it read the data wrong and created a CRC code
based
on the wrong data so that when it put the data on the new computer the CRC
code checked ok? I.e., How could I transfer data using Easy Transfer from
one
computer to another and end up with corrupted data, without it knowing it.
This seems, to me, crazy! (This is related, because it is what got me
looking
into a possible memory problem to begin with).
Sorry for the long winded post, but this is a bit frustrating and would be
less so, if there was more tech info/support for these two products.
Kim

Kim, sorry but I don't know about the inner workings of either WET or VMD.
They are both new. If you have any free support calls left you might want
to contact MS Product support and see if it can be escalated.

 

[mikeyhsd]

do you have a temp monitor running, just wondering if maybe with all that testing running is the system may be having problems because of heat buildup.



(e-mail address removed)



I have been. It takes all night to do 2 runs.
Ok, so, new info:
I got the new memory, Crucial Ballistix, pc2-8000, and the same type of
results, except that it seems to also run very slow at 90% done.
It has to be one of two things, something going on in the motherboard (read:
memory controller) or VMD is doing things differently with DDR2 memory than
it does with DDR memory, such as doing severe fading testing with DDR2 where
it writes data and waits longer to check it. Just guessing since there is not
technical info from Microsoft. I am still a bit concerned, but since it
passes VMD (eventually, like in 3+ hours!!), and Memtest86, I am not sure
what to do. Have to stop testing for the next week, but at least the computer
seems to work well.
When I get back I'll try some other types of testing.
If you all have any other ideas, I will have my laptop with me, but won't be
able to try them until I get back.
Thanks,
Kim

 

[Cal Bear '66]

I also had the same pause at 21%.
It seemed to get stuck there for about 45 to 60 minutes, but
then resumed and completed successfully.

 

[Cal Bear '66]

I don't understand.
Spamming?
Previous posts?
Please explain.

 

[Guest]

You post looked ok here.
Anyway, sounds like I do not have a problem. Thanks for taking the time to
do the tests! It sure wasted a lot of mine worrying about the long delay.
Still don't understand why it was not like this in my old system running
Vista. 21-22% was low in that system but not as disproportionately slow ( by
a LONG shot) as the new system. When I get back to my system next week I'll
do some more testing and, if I come up with something new, I'll post back
here.

 

[cquirke (MVP Windows shell/user)]

On Thu, 12 Apr 2007 07:36:00 -0700, Kim

I have been running VMD with no failures. The room I work in has no
air-conditioning (we live at 9000 ft and really don't need it) so the temp
does vary from day to day and with the cold front that went through here
yesterday the room is about 5-10 degrees below two days ago. If the problem
is marginal, which is what it appears to be, this could explain it passing.
On the other hand, it still hangs for disproportionately long times at 21%.

RAM generally operates at full speed, whether it is working properly
or not - unlike a failing HD, the system does not detect errors and
retry operations until success. Even systems that do detect RAM
errors do not retry on error; they either correct on the fly (if
there's enough data redundancy to do so) or they halt the system.

Regarding memtest86, although I have a lot of respect for the program, I
have been in the PC design and engineering business for 25 yrs and I get a
little paranoid about marginal problems.

Me too, but I'd trust MemTest86 further than I'd trust MS's RAM
tester, the design of which is simply brain dead - rather like the
"idiot's parachute" (opens automatically on impact).

OTOH, if one gives errors and the other does not, then I'd believe
whichever one showed errors. When I use either test, I do so for
typically 24 hours; MemTest86 will accumulate results, whereas MS
tester shows only current pass (see what I mean about "brain dead"?)

After seeing some of my data corrupted after a "Easy Transfer" I want
to make sure this memory is solid.

That, too, is a test. Do two successive bulk copies of material that
does not change, and FC (File Compare) their contents. Any mismatch
indicates some sort of problem that should be fixed (i.e. such that
repeat testing is problem-free) before using the system again.

So, two questions: First, does Microsoft offer ANY support for VMD (Vista
Memory Diagnostics), so I can find out more about why it is hesitating at 21%?
Second, does Easy Transfer do any error checking while doing its job? Or,
maybe it does error checking and the problem was that during the read of the
data on the old computer it read the data wrong and created a CRC code based
on the wrong data so that when it put the data on the new computer the CRC
code checked ok? I.e., How could I transfer data using Easy Transfer from one
computer to another and end up with corrupted data, without it knowing it.

BTW: I've seen exactly this sort of error when one of the HDs was not
chassis-grounded, on doing 500M transfers followed by FC checking. FC
would show anything from 1 to 5 mismatches that were always 4 bytes
(32 bits) long, where the contents were completely different (i.e. not
as clean as flipping a single bit).

 

[cquirke (MVP Windows shell/user)]

On Fri, 13 Apr 2007 04:38:27 +0530, goatmonger

I've been having a very similar experience with the magical 21% mark.
First to clarify: the MDT (memory diagnostics tool) runs fine in
standard mode... whenever I try Extended it seems to hang at 21%.

There are 2 possibilities:

1) at the 21% stage, it is testing an operation that just happens to
trigger errors in both your and my configurations.

2) it's not really an error, and it just takes much longer to go from
21-22% than for any other interval, and I was impatient in cancelling
the test after waiting 15+ minutes for it to move on.

Or (3); it detects an error, and retries to see if it's "consistent",
and it is the tester's retries that keep you waiting.

Hardware doesn't retry RAM errors, even if it detects them, but MS's
RAM tester might. The design is so whacko...
- reachable only after booting a lot of GUI code
- reachable only after the HD is "read" to find a Vista installation
- writes tester to HD before rebooting to test
- shows only the current pass
- writes results to the HD
- requires you to boot the Vista installation to read the results
....that anything's possible.

Generally, bad RAM is a contra-indication to running complex OSs that
write to the file system all the time, because errors can cause writes
to be corrupted, be written to the wrong place on disk, and can also
flip what are supposed to be "read" operations to "writes". In
effect, it "puts quotes" around everytthing the system claims to do,
because what actually happens may differ.

That's why I wouldn't trust the MS RAM tester; the way it is designed,
shows little or no appreciation for the above issues. Scary.

--------------- ----- ---- --- -- - - -

Error Messages Are Your Friends