[Zope] Long delay when error occurs in External Method

[nwingfield at che-llp.com]

OK, how about this one then (I like to ask the hard questions) ...
On one of my Zope servers (this one is running on Windows - ick), if a
python script or page template throws a Zope error, and I debug my code so
that it should no longer throw the error, it will take 5 minutes before
Zope acknowledges my change.  Until then, the page will continue to throw
the same error again and again.  Scripts and page templates with no bugs
immediately respond to code changes, as they should.  I can reproduce this
behavior as easily as creating a new python script containing the code
'return x'.  When I run it, I get an error (x does not exist).  If I
rewrite it to 'x = 1; return x' I still get the error for 5 minutes, at
which point Zope seems to become aware of my change.
Nathaniel



                                                                                                                         
                      Dieter Maurer                                                                                      
                      <dieter at handshake        To:       nwingfield at che-llp.com                                          
                      .de>                     cc:       zope at zope.org                                                   
                                               Subject:  Re: [Zope] Long delay when error occurs in External Method      
                      10/02/2003 06:17                                                                                   
                      PM                                                                                                 
                                                                                                                         




nwingfield at che-llp.com wrote at 2003-10-1 12:27 -0400:
 > I find that if my External Method throws an error, Zope will spin for
 > several minutes before it finally presents the error.  If there is no
 > error, my script will return immediately.  Obviously it's taking me a
long
 > time to debug.  Does anyone know what the problem might be?

No, no idea.

And we do not see this problem...


Dieter