Re: core archive corrupted

[Tyson Whitehead]

Gregory Casamento wrote:

> We've update ftp server on July 27. I've checked the proftpd site but found
> nothing about a probleme like that. We've
> installed the last version.

This coincides very well with when I started to see the problem.   Could we go
back to the previous version to see if that alleviates the problem?

Maybe upgrading my client might help as well.

I finally located the RFC for the FTP protocol.  It's number 542. 
To quote:

ERROR RECOVERY AND RESTART

   There is no provision for detecting bits lost or scrambled in data
   transfer.  This issue is perhaps handled best at the NCP level where
   it benefits most users.

As best as I can tell the NCP (network control protocol?) in our case would be TCP.  So it looks like the 16 bit checksum in the TCP packet is about the only form of error protection a guy gets.  FTP sounds like kind of a lame protocol in my books *grin*.

Anyway, if it was me and all the hops to the host checked out, I would have to wonder about my TCP stack not checking checksums.  If it was the server I would expect others to be having problems as well (of course, maybe there is something specific in the way your client talks to the server that brings out a bug -- try a different ftp program?).

Any chance you updated to a kernel with a flaky TCP stack lately?

-T

-- 
 Tyson Whitehead  (-twhitehe_nospam@nospam_uwo.ca -- WSC 140-)
 Computer Engineer                          Dept. of Applied Mathematics,
 Graduate Student- Applied Mathematics      University of Western Ontario,
 GnuPG Key ID# 0x8A2AB5D8                   London, Ontario, Canada