Strange software failure messages in exim_mainlog

[jackie46]

I'm seeing some strange error messages in exim_mainlog

2006-05-15 12:00:07 1 accept() failure: Software caused connection abort
2006-05-15 12:00:34 1 accept() failure: Software caused connection abort
2006-05-15 12:00:34 getsockname() failed: Connection reset by peer
2006-05-15 12:00:34 daemon: fclose(smtp_out) failed: Broken pipe
2006-05-15 12:01:01 H=(adsl-51-211-192-81.adsl2.iam.net.ma) [81.192.211.51] F=<jweqg@chaletont.com> rejected RCPT <gfgfgfgfgfgfggf@64g.com>: No Such User Here
2006-05-15 12:01:44 1 accept() failure: Software caused connection abort
2006-05-15 12:02:08 1 accept() failure: Software caused connection abort
2006-05-15 12:02:18 1 accept() failure: Software caused connection abort

I'm not sure where its coming from. It could be MailScanner that we installed a couple of days ago but i cant be sure. I'v never seen these error messages on any server running MS. Anyone seen this before?

[webignition]

Could be errors generated from failed message pipes i.e. when a user tries to pipe a message to a PHP or Perl script and the piping fails due to an error with the script to which the message is being piped.

[AndyReed]

I'm seeing some strange error messages in exim_mainlog

2006-05-15 12:00:07 1 accept() failure: Software caused connection abort
2006-05-15 12:00:34 1 accept() failure: Software caused connection abort

Run 'netstat' to see which process is already listening on port 25.

Or lsof ...

The other possible meaning for this error is that a connection abort was caused internal to the host, as there's no space in the socket's queue and it can't receive further connections.
It is very likely that a connection is made by a remote machine or locally - but suddenly drops.

Possible solutions:
- Increase the timeouts variable in exim. If that doesn't work,
- increase smtp_connect_backlog.