Recently, I've found that my Apache2 installation
  (2.0.53, i386, NetBSD 2.0.2) has stopped responding, and that people
  cannot access my website. This is not a frequent thing, but it seems
  to happen every now and then, and only returns to normal when I
  stop, and then restart apache. I haven't seen anything suspicious in
  the error logs or anything, so I was just wondering if anyone else
  has came across this before, and if so, what the solution was?

I have the same problem with 1.6.2ish and various versions of apache 2
up to very recent.  When the problem happens there are about a dozen
httpds in state Z.  /etc/rc.d/apache restart fixes things.

I can see several solutions, neither of which I've implemented, listed
in the order I'm likely to try them

  try 3.0 and see if that helps

  write a script that looks at ps output and does restarts and run it
  from cron.  Kind of like rebooting windows, but easy.

  debug the problem.  I suspect it has something to do with threads,
  but I have no real basis for this suspicion. 


-- 
        Greg Troxel <gdt@ir.bbn.com>