[Spambayes] RE: [spambayes-bugs] uncaptured python exception ... object has noattribute 'isClosed'

Tony Meyer tameyer at ihug.co.nz
Sat Feb 7 21:23:38 EST 2004


[redirected from the inappropriate spambayes-bugs list]

Please note that the spambayes-bugs list isn't a list for humans to post to,
so if you don't want to see the machine-generated email that appears on that
list, don't subscribe to it at all.  Descriptions of the 5 available mailing
lists are here:

    http://spambayes.sourceforge.net/contact.html

It's best to enter bug reports on the SpamBayes bug tracker (also at
SourceForge), but if that's too hard then mailing to the spambayes list has
a decent chance of getting noticed.

[...]
> This is working very well, except about once every 24 hours 
> spambayes quits with this message:
> 
> error: uncaptured python exception, closing channel 
> <__main__.BayesProxyListener listening :1110 at 0x83620d4> 
> (exceptions.IOError:[Errno 5] Input/output error 

This is probably some problem with reaching the pop server.

> error: uncaptured python exception, closing channel 
> <__main__.BayesProxy connected 127.0.0.1:42864 at 0x841acac> 
> (exceptions.AttributeError:'_socket.socket' object has no 
> attribute 'isClosed' [/usr/lib/python2.2/asyncore.py|poll|113]

I assume this is a consequence of the previous error, because it's very odd
(the object from which a 'isClosed' attribute is being requested is not of
type _socket.socket).

> Apart from the once-a-day-or-so failure rate, I have not seen 
> any pattern.  I have wondered whether the failures might be 
> related to one of the following reasons:
> 
>  1) The remote pop3 server is periodically unreachable, often for
>     several minutes.  Could this cause spambayes to fail?

Yes, I suspect that this could be the problem.  There is an open bug report
about SpamBayes failing to recover after a timeout, which sounds (I think)
fairly similar:

<https://sourceforge.net/tracker/?func=detail&atid=498103&aid=824589&group_i
d=61702>

If so, once this problem is solved, yours should be as well.  I'm not sure
what to suggest until then :(

>  2) It is not clear to me what happens if fetchmail has a polling
>     interval which is too short.  Could it be that fetchmail is
>     kicking off a new session *before* the previous session has
>     completed?

That shouldn't matter, AFAIK.

=Tony Meyer

---
Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes. This
way, you get everyone's help, and avoid a lack of replies when I'm busy.




More information about the Spambayes mailing list