[gentoo-user] possible defective memory

[gentoo-user] possible defective memory

[bruce harding]

I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is it
possible that this if I ran memtest86+ for 2 days straight and found no
error that the memory could still be defective?

I ask because I can't get a complete compile of glibc.  I have to
restart the process at lease 3 times before the compile will complete.


Let me know what you think.
--
bruce
-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[Hemmann, Volker Armin]

On Sunday 09 October 2005 03:08, bruce harding wrote:
> I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is it
> possible that this if I ran memtest86+ for 2 days straight and found no
> error that the memory could still be defective?
>
> I ask because I can't get a complete compile of glibc.  I have to
> restart the process at lease 3 times before the compile will complete.
>
>
> Let me know what you think.
> --

yes it is completly possible.

But it is also possible, that your PSU is not powerfull enough. A big compile 
needs a lot of processing power and stresses the ram, so a lot of current is 
needed - and some PSUs aren't able to cope with such a load -exspecially if 
they are cheap and/or a little bit older.
Try another PSU, do you still have problems, RMA the ram.

Memtest86(+) is known not to find all errors.
-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[Ted Ozolins]

Hemmann, Volker Armin wrote:

>>
>>I ask because I can't get a complete compile of glibc.  I have to
>>restart the process at lease 3 times before the compile will complete.
>>
>>
>>Let me know what you think.
>>--
>>    
>>
>
>yes it is completly possible.
>
>But it is also possible, that your PSU is not powerfull enough. A big compile 
>needs a lot of processing power and stresses the ram, so a lot of current is 
>needed - and some PSUs aren't able to cope with such a load -exspecially if 
>they are cheap and/or a little bit older.
>Try another PSU, do you still have problems, RMA the ram.
>
>Memtest86(+) is known not to find all errors.
>  
>
I had the same problem here, Changed RAM and still the same prob.
Changed PSU, glibc compiled just fine. This was on an MSI MB with an AMD
750 Duron.

Cheers.

-- 
Ted Ozolins(VE7TVO)
Westbank, B. C

-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[Hemmann, Volker Armin]

On Sunday 09 October 2005 03:52, Ted Ozolins wrote:
> Hemmann, Volker Armin wrote:
> >>I ask because I can't get a complete compile of glibc.  I have to
> >>restart the process at lease 3 times before the compile will complete.
> >>
> >>
> >>Let me know what you think.
> >>--
> >
> >yes it is completly possible.
> >
> >But it is also possible, that your PSU is not powerfull enough. A big
> > compile needs a lot of processing power and stresses the ram, so a lot of
> > current is needed - and some PSUs aren't able to cope with such a load
> > -exspecially if they are cheap and/or a little bit older.
> >Try another PSU, do you still have problems, RMA the ram.
> >
> >Memtest86(+) is known not to find all errors.
>
> I had the same problem here, Changed RAM and still the same prob.
> Changed PSU, glibc compiled just fine. This was on an MSI MB with an AMD
> 750 Duron.
>

I had 4 PSUs so far going weak - sometimes the box only booted after the nth 
try, sometimes it suddenly crashed, segfaults and so on. Almost everything 
that you could also blame on faulty memory or bad cooling.
That's why I always blame the PSU first, if it is not a quality brand one ;)

I have an enermax for more than a year now (should be almost two, if I 
remember correctly) since then, all this problems vanished.

I friend brought a PSU which had a defect - I started it - boom it exploded.. 
for the second time, we learnt, after disecting it ;)
Another one killed his PSU, when he connected a big old scsi-drive - burned 
capacitors and transistors don't smell nice ....

-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[bruce harding]

On Sun, 9 Oct 2005 03:16:52 +0200
"Hemmann, Volker Armin" <volker.armin.hemmann@tu-clausthal.de> wrote:

> On Sunday 09 October 2005 03:08, bruce harding wrote:
> > I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is
> > it possible that this if I ran memtest86+ for 2 days straight and
> > found no error that the memory could still be defective?
> >
> > I ask because I can't get a complete compile of glibc.  I have to
> > restart the process at lease 3 times before the compile will
> > complete.
> >
> >
> > Let me know what you think.
> > --
> 
> yes it is completly possible.
> 
> But it is also possible, that your PSU is not powerfull enough. A big
> compile needs a lot of processing power and stresses the ram, so a
> lot of current is needed - and some PSUs aren't able to cope with
> such a load -exspecially if they are cheap and/or a little bit older.
> Try another PSU, do you still have problems, RMA the ram.
> 
> Memtest86(+) is known not to find all errors.

Actually, I'm on my third PSU I now own a dual rail, 650watt
SilverStone.

And I don't get any errors.   The compile just appears to stop, but if
I do "top" the thread for the compile is running at 90%.  I got off the
phone with Kingston and they are going to replace the ram.  I hope that
solves my problem.

==
bruce
-- 
gentoo-user@gentoo.org mailing list

RE: [gentoo-user] possible defective memory

[Bob Young]

FWIW, I had the same problem, (compile would hang, no error, no message,
nothing), while installing a dual AMD64 system last week. I got past it by
using kernel switches to force single CPU (nosmp), and disabling the apic
(noapic), during the install. Don't know for sure which switch resolved it,
or if they were both needed. May not help you, but thought I'd mention it
just in case.

Regards,
Bob Young


-----Original Message-----
From: bruce harding [mailto:bshlists@rogers.com]
Sent: Tuesday, October 11, 2005 3:13 PM
To: gentoo-user@lists.gentoo.org
Subject: Re: [gentoo-user] possible defective memory

On Sun, 9 Oct 2005 03:16:52 +0200
"Hemmann, Volker Armin" <volker.armin.hemmann@tu-clausthal.de> wrote:

> On Sunday 09 October 2005 03:08, bruce harding wrote:
> > I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is
> > it possible that this if I ran memtest86+ for 2 days straight and
> > found no error that the memory could still be defective?
> >
> > I ask because I can't get a complete compile of glibc.  I have to
> > restart the process at lease 3 times before the compile will
> > complete.
> >
> >
> > Let me know what you think.
> > --
>
> yes it is completly possible.
>
> But it is also possible, that your PSU is not powerfull enough. A big
> compile needs a lot of processing power and stresses the ram, so a
> lot of current is needed - and some PSUs aren't able to cope with
> such a load -exspecially if they are cheap and/or a little bit older.
> Try another PSU, do you still have problems, RMA the ram.
>
> Memtest86(+) is known not to find all errors.

Actually, I'm on my third PSU I now own a dual rail, 650watt
SilverStone.

And I don't get any errors.   The compile just appears to stop, but if
I do "top" the thread for the compile is running at 90%.  I got off the
phone with Kingston and they are going to replace the ram.  I hope that
solves my problem.

==
bruce
--
gentoo-user@gentoo.org mailing list


-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[Michael Crute]

[-- Attachment #1: Type: text/plain, Size: 574 bytes --]

On 10/8/05, bruce harding <bshlists@rogers.com> wrote:
>
> I ask because I can't get a complete compile of glibc. I have to
> restart the process at lease 3 times before the compile will complete.
>
>
I would doubt that you have memory issues if memtest doesn't say anything.
What are the other symtoms? Error messages?

-Mike

--
________________________________
Michael E. Crute
Software Developer
SoftGroup Development Corporation

Linux, because reboots are for installing hardware.
"In a world without walls and fences, who needs windows and gates?"

[-- Attachment #2: Type: text/html, Size: 899 bytes --]

Re: [gentoo-user] possible defective memory

[Joseph]

On Sat, 2005-10-08 at 21:08 -0400, bruce harding wrote:
> I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is it
> possible that this if I ran memtest86+ for 2 days straight and found no
> error that the memory could still be defective?
> 
> I ask because I can't get a complete compile of glibc.  I have to
> restart the process at lease 3 times before the compile will complete.
> 
> 
> Let me know what you think.
> --
> bruce

That was my issue.  I had exactly the same problem when I assembling
together AMD64.  It took me one week to investigate.  Memorytest86 pass
all the test and the system was still crashing when I was compiling for
a longer time.  So in other words Memorytest86 it is not worth the disk
space it occupies.

Look on Google for another test called: memtest.sh from RedHat and run
it.
In my case it was memory issue.  I got a new pair of stick and the
problem went away.

-- 
#Joseph
-- 
gentoo-user@gentoo.org mailing list

Re: [gentoo-user] possible defective memory

[Spider (D.m.D. Lj.)]

[-- Attachment #1: Type: text/plain, Size: 683 bytes --]

On Sat, 2005-10-08 at 21:08 -0400, bruce harding wrote:
> I've got 2 sticks of Kingston HyperX 3200 DDR Registered & ECC.  Is it
> possible that this if I ran memtest86+ for 2 days straight and found no
> error that the memory could still be defective?
> 
> I ask because I can't get a complete compile of glibc.  I have to
> restart the process at lease 3 times before the compile will complete.
> 
> 
> Let me know what you think.

Well, here's a good list to go through.:

http://people.redhat.com/davej/hardware-problems.txt



//Spider

-- 
begin  .signature
Tortured users / Laughing in pain
See Microsoft KB Article Q265230 for more information.
end


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]