[Mailman-Users] cgi-script "low level failure"

[Don Burns]

Hi Mark,

Thanks for the quick reply.  I've tried your suggestions and came up 
empty.  db_check --all did not report errors, list_lists output was clean 
and everything expected, A couple of debug lines in listinfo.py never 
showed up (I assumed they would print in logs/error).  There are no 
foreign files or directories in lists. 

I do have a new piece to the puzzle.  The apache2 log file reports this 
when accessing a cgi-script:

[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] 
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] [----- Mailman 
Version: 2.1.4 -----]
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] [----- Traceback 
------]
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] Traceback (most 
recent call last):
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8]   File 
"/usr/lib/mailman/scripts/driver", line 241, in ?
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8]     run_main()
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8]   File 
"/usr/lib/mailman/scripts/driver", line 69, in run_main
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8]     import 
xml.sax.saxutils
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] ImportError: No 
module named xml.sax.saxutils
[Sat Jan 22 13:12:46 2005] [error] [client 63.204.157.8] [Mailman: low 
level unrecoverable exception]

Upon investigation, it looks as if the error at line 69 (import of 
xml.sax.saxutils) is one relating to an error message.  Perhaps this is 
why the traceback on the error (useful debug info) is failing?

My apologies, I am a python novice.

-don


On Sat, 22 Jan 2005, Mark Sapiro wrote:

> Don Burns wrote:
> >
> >The problem:
> >
> >  http://www.openscenegraph.net/listinfo
> >
> >Produces:
> >/////////////////////////////////////////////////////////////////////////
> >Bug in Mailman version 2.1.4
> >
> >We're sorry, we hit a bug!
> >
> >If you would like to help us identify the problem, please email a copy of
> >this page to the webmaster for this site with a description of what
> >happened.  Thanks!
> >
> >Traceback:
> >
> >Content-type: text/html
> >
> >We're sorry, we hit a bug!
> >
> >Mailman experienced a very low level failure and could not even generate a
> >useful traceback for you.  Please report this to the Mailman administrator
> >at this site.
> >/////////////////////////////////////////////////////////////////////////
> >
> >I have 
> >  o checked my installation and configuration of mailman and all seems 
> >ok.  
> >  o run check_perms and fixed some minor problems, also fine.
> >  o locks are legitimate (qrunner is the only one and it is running)  
> >  o plenty of disk space and no disk problems
> >  o postfix and apache are working fine.  
> >
> >I wrote a small test script and tried it in the mailman/cgi-bin and it
> >worked fine.  I've also restarted postfix, apache, mailman.  The problem 
> >persists.
> >
> >The problem started occuring suddenly this morning after having run for 
> >several months with no issues.  The last entry in the error log is from 
> >Jan 18 (4 days ago).
> >
> >For reference, the server is running Suse 9.1, kernel 2.6.5-7.111.5-smp.  
> >Mailman is version 2.1.4 (as reported in the error message).
> >
> >I'm stuck!  Please help.
> 
> Does "bin/check_db --all" report any errors?
> 
> Does "bin/list_lists" work? If not, try adding "        print n" at the
> top of the "    for n in names:" loop to see which list has the
> problem.
> 
> If "bin/list_lists" works, you could try adding "syslog('error',
> <replace this with something to print...>)" in Mailman/Cgi/listinfo.py
> to try to pinpoint where it goes wrong.
> 
> Also, is there anything in lists/ that doesn't belong? I.e. a directory
> that isn't a list or a top level file?
> 
> --
> Mark Sapiro <msapiro at value.net>       The highway is for gamblers,
> San Francisco Bay Area, California    better use your sense - B. Dylan
> 
>