[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Update on corrupted working copies

From: Kevin Pilch-Bisson <kevin_at_pilch-bisson.net>
Date: 2001-11-01 18:26:21 CET

Here's an updated list of what has happened. (for anyone who cares).

1) memtest (userland) found no problems.

2) upgraded to kernel 2.4.13: still problems

3) ran memtest86 in real mode: found no problems

4) switched nics with a different one, but same driver: still problems

5) upgraded all packages from debian/testing to debian/unstable in the hopes
that it was a glibc/kernel mismatch: still problems

6) tried running ethereal: First time failed, second time got a complete,
clean checkout.

7) tried running tcpdump > /dev/null: Got a checkout, but some files were
corrupted.

Conclusions:
1) Seems to be a problem in my network stack. tcpdump/ethereal seems to
help, but not completely solve the problem.

Suggestions?

On Wed, Oct 31, 2001 at 12:03:00PM -0500, Kevin Pilch-Bisson wrote:
> It would appear that my computer has developed a hardware or software problem
> unrelated to subversion.
>
> Here is what I have seen so far.
>
> 1) I often ssh into my box from remote locations. For the last week or
> so my connections have been dropped due to corrupted packets. At first
> I thought this was due to some new routers and switches installed on the
> network I connect from, but now I am not so sure.
>
> 2) Last night and tonight, I have had a large (about 20), number of
> 1 bit errors in files, and a couple of 2 and 3 bit errors.
>
> 3) These bit errors occur in both my text-base and in my working copy.
>
> 4) I just tried to do a checkout and it barfed because the reponse had
> non-matching tags. Why were they non-matching? Because the closing
> tag had a on bit error in the tag name.
>
> I am currently running the memtest program, trying to see if my ram is a problem
> but I am not sure it will work since it can only test 67 of 128MB of physical
> ram in my system.
>
> This has led me to beleive that either 1) I have bad ram, or 2) I have a bug
> in my Linux 2.4.7 networking stack. I am tempted to think that it is 2), since
> I haven't had any programs crash unexpectedly from jumping to the wrong address
> or anything, so I guess my next step is to build a 2.4.12 kernel, and hope
> my system becomes happier.
>
> Other suggestions welcome. :)
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Kevin Pilch-Bisson http://www.pilch-bisson.net
> "Historically speaking, the presences of wheels in Unix
> has never precluded their reinvention." - Larry Wall
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Kevin Pilch-Bisson                    http://www.pilch-bisson.net
     "Historically speaking, the presences of wheels in Unix
     has never precluded their reinvention." - Larry Wall
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  • application/pgp-signature attachment: stored
Received on Sat Oct 21 14:36:47 2006

This is an archived mail posted to the Subversion Dev mailing list.