🚀 go-pugleaf

I'm running sendmail 5.61 under AU/X 1.1.  Periodically I find a few
sendmail processes have hung reading a message.  Typically I see in
the ps listing the original sendmail and a few others (usually 1,
sometimes more) with CPU times in the tens (sometimes hundreds) of
minutes.  mailq shows an entry for these messages, but usually there's
no control file yet.  netstat shows there's a connection but it's
usually marked as CLOSE_WAIT or something like that.

I've just been killing these processes and rm'ing the queue files, but
I'd really prefer to fix rather than work around the problem.  I
thought I saw something posted about this, but now I can't find it.
Does anyone know the solution?

Kevin Crowston