Re: [Zope] Non-responsive folders

[Dieter Maurer]

Garth B. wrote at 2005-8-16 19:21 -0400:
> ...
>When I hit an offending
>folder, I simply get no response and the browser just waits and waits.

Visit "Control_Panel --> Debug Information" and check whether
your request indeed does not finish.
You see the active requests at the bottom of this page.

If this page shows you, that the request was finished by Zope,
then you hit a wide spread browser bug:

  In some cases, Zope returns an 204 (no content) response.
  For unknown reasons (and against the HTTP spec),
  most browsers treat this as "continue to wait".

If the request is not finished, you can use
Forent's "DeadlockDebugger" to find out where your
request is spinning.

-- 
Dieter
___
Zope maillist  -  Zope@zope.org
http://mail.zope.org/mailman/listinfo/zope
**   No cross posts or HTML encoding!  **
(Related lists - 
 http://mail.zope.org/mailman/listinfo/zope-announce
 http://mail.zope.org/mailman/listinfo/zope-dev )

[Zope] Non-responsive folders

[Garth B.]

Hello everyone,

Python 2.3.4
Zope 2.7.3
Apache 2.0.46

I have a website that has been running flawlessly since late last year
sitting behind Apache using the usual rewrite rules.  I've recently
started experiencing some folders becoming non-responsive when trying
to view them (e.g. http://www.someserver.com/somefolder).  I can still
access most other folders without a problem.  When I hit an offending
folder, I simply get no response and the browser just waits and waits.
 I haven't been able to establish a pattern as to what's behind the
folders that might be causing the problem.  On a clean start the
problematic folders usually respond normally for awhile, then stop. 
So far, which folders stop responding are always the same.

Zope's event.log shows a number of the following in case it is relevant:

2005-08-16T13:50:35 ERROR(200) ZServer uncaptured python exception,
closing channel 
(socket.error:(104, 'Connection reset by peer')
[/var/tmp/python2.3-2.3.4-root/usr/lib/python2.3/asynchat.py|initiate_send|218]
[/usr/local/Zope/2.7.3/lib/python/ZServer/medusa/http_server.py|send|417]
[/var/tmp/python2.3-2.3.4-root/usr/lib/python2.3/asyncore.py|send|337])

The problem occurs sporadically enough that I haven't been able to
tell whether these lines from the event.log happen in conjunction with
a non-responsive folder.

I also see the occasional ZODB conflict error.

Interestingly enough I can directly view the offending folders without
a problem if I circumvent Apache (e.g.
http://zope.someserver.com:8080/website/somefolder).  As far as I can
tell nothing has changed with my Apache configuration.

At first I thought that maybe something had gotten screwed up with my
Zope instance or data.fs (~1200MB), but since I can directly view the
offending folders without a problem by skipping Apache, I'm starting
to think otherwise.  Restarting Zope *usually* temporarily clears the
problem.

Can anyone offer any suggestions for what might be going on or what
other evidence I can look for?  As a possible solution, I'm hastily
preparing to upgrade to a more recent version of Zope, but I'd sure
like to know what's suddenly gone wrong (especially since upgrading
may not solve the problem).

Thank you!
Garth
___
Zope maillist  -  Zope@zope.org
http://mail.zope.org/mailman/listinfo/zope
**   No cross posts or HTML encoding!  **
(Related lists -
 http://mail.zope.org/mailman/listinfo/zope-announce
 http://mail.zope.org/mailman/listinfo/zope-dev )