[XML-SIG] xmlrpclib.py exception while trying to raise exception

[sste...@gmail.com]

Hi!

Not totally sure this is the right list for this but...

This just popped up in a program that has successfully (in this run) made thousands of multicall()s.

Then, out of the blue, this came up.

I couldn't find anything like this in the bug tracker.

Anyone have any clues about why the exception itself wouldn't be defined or should I just file a bug?

Thanks,

S
aka/ssteinerX
aka/Steve Steiner

mc_result = multicall()
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1042,
in __call__
return MultiCallIterator(self.__server.system.multicall(marshalled_list))
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1199, in __call__
return self.__send(self.__name, args)
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1489, in __request
verbose=self.__verbose
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1253, in request
return self._parse_response(h.getfile(), sock)
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1392, in _parse_response
return u.close()
File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 838, in close
raise Fault(**self._stack[0])
Fault: <Fault 1: "<type 'exceptions.NameError'>:global name 'Fault' is not defined">

_______________________________________________
XML-SIG maillist - XML...@python.org
http://mail.python.org/mailman/listinfo/xml-sig

["Martin v. Löwis"]

> Anyone have any clues about why the exception itself wouldn't be defined or should I just file a bug?

This is probably during Python shutdown. Python is in the process of
clearing all module dictionaries, and then some destructor still wants
to find the Fault class.

You should make sure that all connections get closed properly before
shutting down Python.

Regards,
Martin

[sste...@gmail.com]

On Jan 16, 2010, at 1:23 PM, Martin v. Löwis wrote:

>> Anyone have any clues about why the exception itself wouldn't be defined or should I just file a bug?
>
> This is probably during Python shutdown. Python is in the process of
> clearing all module dictionaries, and then some destructor still wants
> to find the Fault class.
>
> You should make sure that all connections get closed properly before
> shutting down Python.

Ok, this seems to have happened at the end of a script and it looks like an exception when it was trying to shut down the xmlrpc connection.

S

[Dieter Maurer]

sste...@gmail.com wrote at 2010-1-15 21:47 -0500:
>This just popped up in a program that has successfully (in this run) made thousands of multicall()s.

> ...

>Anyone have any clues about why the exception itself wouldn't be defined or should I just file a bug?

> ...

>File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1042,
>in __call__
> return MultiCallIterator(self.__server.system.multicall(marshalled_list))
> File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1199, in __call__
> return self.__send(self.__name, args)
> File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1489, in __request
> verbose=self.__verbose
> File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1253, in request
> return self._parse_response(h.getfile(), sock)
> File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 1392, in _parse_response
> return u.close()
> File "/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/xmlrpclib.py", line 838, in close
> raise Fault(**self._stack[0])
>Fault: <Fault 1: "<type 'exceptions.NameError'>:global name 'Fault' is not defined">

Almost surely, the "NameError" comes from your XMLRPC server, not
the XMLRPC client.

If you are lucky, your server logs sufficient information for exceptions
(e.g. the traceback) to allow you to find out where the exception
occurred.

--
Dieter

[sste...@gmail.com]

Thanks. Unfortunately I don't have access to the remote -- it's a service run by another company.

This seems to have occurred during shutdown when the connection was trying to close itself but the Fault object had already been removed.

Thanks,

S