|
Graham Dumpleton
grahamd at dscpl.com.au
Wed Feb 15 17:05:02 EST 2006
On 16/02/2006, at 8:33 AM, Graham Dumpleton wrote: > On 16/02/2006, at 6:17 AM, Tomasz Wlodek wrote: > >> Hi, >> >> I noticed that when the mod_python handler fails several times, >> then I can >> fix it by restarting apache. It will then work, while producing the >> following message in apache error log: >> >> [Wed Feb 15 14:14:21 2006] [notice] cannot use a full URL in a 401 >> ErrorDocument directive --- ignoring! >> [Wed Feb 15 14:14:30 2006] [error] [client 66.249.72.211] [Wed Feb 15 >> 14:14:30 2006] rdiff: \t(in cleanup) Can't call method "store" on an >> undefined value at /usr/lib/perl5/site_perl/5.8.5/CGI/Session.pm >> line 192 >> during global destruction. >> >> Could there be some sort of leak of some resource in _mp_mod_ssl >> which >> causes the code to fail after a while? I haven't looked at how mod_perl implements server cleanup functions, but the way that Python does has been found to cause problems. As I haven't found any other way it can be done yet, can only assume that mod_perl does it the same way and thus is also having problems. If I am right, mod_perl could have these issues independent of whether mod_python is used. See: http://issues.apache.org/jira/browse/MODPYTHON-109 for the server cleanup problem as it pertains to mod_python. Graham
|