[Zope] Fresh install fails in minutes

[Scott Wunsch]

I've just set up a fresh new Red Hat 9 installation, and dropped in a fresh
install of Zope 2.6.1 (from the binary tarball on the Zope site, which I
believe includes its own Python interpreter).  I'm currently testing it out
by starting it like this:

	su - zope -s /bin/sh -c "/opt/zope/start"

Yes, the /opt/zope/var directory is properly owned by the zope user.

It works fine for a few minutes, but after I poke around a bit (especially
once I hit the management interface), it suddenly stops responding to
requests.  The process is still running, and it's still accepting
connections, but it completely ignores those connections.  Nothing is
logged (not even the requests I'm making) after it stops responding.  I'm
running it through Apache via FastCGI, but I've also tried running only the
HTTP server on port 8080, and the same thing happens there.

Any ideas what I should be looking for?

-- 
Take care,
Scott \\'unsch

... Oh give me a phone, with a modem on loan...