Re: [expert] Suspected FS corruption bug

2000-01-26 Thread Sheldon Lee Wen

> > In practically ever case of problems similar to what you described I have
> > found the cause to be bad ram. About 1 of every 10 PC100 SDRAMS we have
> > received from our vendor have been bad. They would pass the short POST at
> > power-on but hey raised hell after words in a number of subtle ways.

OK. I ran the ram check and your right. The ram has gone bad. Sucks.

Thanks,
Sheldon.

-- 
==
Sheldon Lee Wenhttp://members.xoom.com/Lycadican 
"Superstition is a word the ignorant use to describe their ignorance."
  -- Sifu.
==



Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Sheldon Lee Wen

> does it happen with the 'kernel-linus' ?

I recompiled my kernel last night from stock source code
I downloaded last night. I'll let you know how it goes.
So far no problems except wierd errors with VMWare's NT install.
But that's most likely due to the fact that I transferred the entire
VMWare disk to this machine over the internal network. I've fixed it
now if I notice anything more I'll let you know.

The suggestion about the memory is also possible I guess. Does linux
have problems with ECC PC100 SDRAM? I've got a "ram exam" program I'll
run on it to see if it turn's anything up.

Cheers,
Sheldon.

-- 
==
Sheldon Lee Wenhttp://members.xoom.com/Lycadican 
"Superstition is a word the ignorant use to describe their ignorance."
  -- Sifu.
==



Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Sheldon Lee Wen

Palmer C Byrne wrote:
> 
> In practically ever case of problems similar to what you described I have
> found the cause to be bad ram. About 1 of every 10 PC100 SDRAMS we have
> received from our vendor have been bad. They would pass the short POST at
> power-on but hey raised hell after words in a number of subtle ways.

Hmm. OK I'll check the ram. However this machine was running fine for
about 4 months until I put mandrake 7.0 on it. Never had any problems
with this machine with mandrake 6.0 on it.

Cheers,
Sheldon.
-- 
==
Sheldon Lee Wenhttp://members.xoom.com/Lycadican 
"Superstition is a word the ignorant use to describe their ignorance."
  -- Sifu.
==



Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Palmer C Byrne

At 09:07 AM 1/22/00 -0700, you wrote:
>On Fri, 21 Jan 2000, Sheldon Lee Wen wrote:
>
> > Hi,
> >
> > I'm posting this again b/c I've been having email difficulties
> > and I'm not sure it got through the first time. If it did I apologize.
> >
>


In practically ever case of problems similar to what you described I have 
found the cause to be bad ram. About 1 of every 10 PC100 SDRAMS we have 
received from our vendor have been bad. They would pass the short POST at 
power-on but hey raised hell after words in a number of subtle ways.

regards palmer





Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Axalon Bloodstone

On Fri, 21 Jan 2000, Sheldon Lee Wen wrote:

> Hi,
> 
> I'm posting this again b/c I've been having email difficulties
> and I'm not sure it got through the first time. If it did I apologize.
> 
> I'm beginning to suspect that there may be a bug in the stock
> mandrake kernel that causes file system corruption. (I installed 
> mandrake7.0 almost a week ago now.)
> 
> Eg. I downloaded star office to my home directory and then su'd 
> and copy'd the file to the directory I wanted to install it in. 
> I untarred it and installed it; had some problems with kde
> integration so figured I'd just reinstall it. ( I was tired ok! )
> However the second time I untarred it I got crc errors. So I copied it
> back from my home dir and untarred it and it worked. However half 
> way through the install I got crc errors while extracting files 
> (I had previously removed the old install and setup directories 
> so there were no preexisting files there). 
> 
> So I downloaded it again and it worked fine for about two hours then
> some lib's started getting corrupted and it refused to load, it 
> would immediately segfault. So I completely deleted it and the 
> install directory and untarred it again and ran it and got crc 
> errors. Anyways I ended up downloading it again and
> reinstalling it.
> 
> Then I was tired so I shutdown the computer normally with "shutdown -h
> now" I waited until it was done and went to bed. When I fired up the
> computer today I got alot of errors on the root drive and a nasty fsck. 
> 
> If this is fs corruption as a result of a bug in the kernel or something
> else is there a way that I can specifically test this?
> 
> Sheldon.
 
does it happen with the 'kernel-linus' ?  

-- 
MandrakeSoft  http://www.mandrakesoft.com/
--Axalon



Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Civileme

Civileme wrote:

> 
>
> That does not sound like anything but either a) memory or b) HDD.  Most likely b.
>
> I just recovered a filesystem where a processor had done a HCF (Halt and catch
> fire) and the disk repaired itself beautifully.
>
> Civileme

To test, put the HDD into another computer running linux from another HDD and use
fdsik to print a partition table then fsck -c every partition using alternate blocks
as necessary (except, of course the swap partition)

Then move it to the boot position and see if it starts OK

If it does, you most likely have a memory problem or an IDE interface problem
(unlikely) wtih the computer you took it out of.

If it doesn't repair or errors out of the fscks no matter what, your Hdd is due for
that last ride

If it repairs but immediately shows errors when you boot it or after two or three
boots then perhaps there is a bug everyone wants to know about.  However this is the
most unlikely of the three since the routines that handle the filesystem are the
nearly the oldest and certainly most stable of all and only the drivers are updated.

I hit the send key too soon before.  Apologies.

Civileme


>
>
> Sheldon Lee Wen wrote:
>
> > Hi,
> >
> > I'm posting this again b/c I've been having email difficulties
> > and I'm not sure it got through the first time. If it did I apologize.
> >
> > I'm beginning to suspect that there may be a bug in the stock
> > mandrake kernel that causes file system corruption. (I installed
> > mandrake7.0 almost a week ago now.)
> >
> > Eg. I downloaded star office to my home directory and then su'd
> > and copy'd the file to the directory I wanted to install it in.
> > I untarred it and installed it; had some problems with kde
> > integration so figured I'd just reinstall it. ( I was tired ok! )
> > However the second time I untarred it I got crc errors. So I copied it
> > back from my home dir and untarred it and it worked. However half
> > way through the install I got crc errors while extracting files
> > (I had previously removed the old install and setup directories
> > so there were no preexisting files there).
> >
> > So I downloaded it again and it worked fine for about two hours then
> > some lib's started getting corrupted and it refused to load, it
> > would immediately segfault. So I completely deleted it and the
> > install directory and untarred it again and ran it and got crc
> > errors. Anyways I ended up downloading it again and
> > reinstalling it.
> >
> > Then I was tired so I shutdown the computer normally with "shutdown -h
> > now" I waited until it was done and went to bed. When I fired up the
> > computer today I got alot of errors on the root drive and a nasty fsck.
> >
> > If this is fs corruption as a result of a bug in the kernel or something
> > else is there a way that I can specifically test this?
> >
> > Sheldon.
> >
> > --
> > ==
> > Sheldon Lee Wen  http://members.xoom.com/Lycadican
> > "Superstition is a word the ignorant use to describe their ignorance."
> >   -- Sifu.
> > ==
>
> --
> experimentation involving more than 500 trials with an
> ordinary slice of bread and a tablespoon of peanut butter
> has determined that the probability a random toss will
> land sticky side down (SSD) is approximately .98

--
experimentation involving more than 500 trials with an
ordinary slice of bread and a tablespoon of peanut butter
has determined that the probability a random toss will
land sticky side down (SSD) is approximately .98





Re: [expert] Suspected FS corruption bug

2000-01-22 Thread Civileme



That does not sound like anything but either a) memory or b) HDD.  Most likely b.

I just recovered a filesystem where a processor had done a HCF (Halt and catch
fire) and the disk repaired itself beautifully.

Civileme


Sheldon Lee Wen wrote:

> Hi,
>
> I'm posting this again b/c I've been having email difficulties
> and I'm not sure it got through the first time. If it did I apologize.
>
> I'm beginning to suspect that there may be a bug in the stock
> mandrake kernel that causes file system corruption. (I installed
> mandrake7.0 almost a week ago now.)
>
> Eg. I downloaded star office to my home directory and then su'd
> and copy'd the file to the directory I wanted to install it in.
> I untarred it and installed it; had some problems with kde
> integration so figured I'd just reinstall it. ( I was tired ok! )
> However the second time I untarred it I got crc errors. So I copied it
> back from my home dir and untarred it and it worked. However half
> way through the install I got crc errors while extracting files
> (I had previously removed the old install and setup directories
> so there were no preexisting files there).
>
> So I downloaded it again and it worked fine for about two hours then
> some lib's started getting corrupted and it refused to load, it
> would immediately segfault. So I completely deleted it and the
> install directory and untarred it again and ran it and got crc
> errors. Anyways I ended up downloading it again and
> reinstalling it.
>
> Then I was tired so I shutdown the computer normally with "shutdown -h
> now" I waited until it was done and went to bed. When I fired up the
> computer today I got alot of errors on the root drive and a nasty fsck.
>
> If this is fs corruption as a result of a bug in the kernel or something
> else is there a way that I can specifically test this?
>
> Sheldon.
>
> --
> ==
> Sheldon Lee Wen   http://members.xoom.com/Lycadican
> "Superstition is a word the ignorant use to describe their ignorance."
>   -- Sifu.
> ==

--
experimentation involving more than 500 trials with an
ordinary slice of bread and a tablespoon of peanut butter
has determined that the probability a random toss will
land sticky side down (SSD) is approximately .98