I have a setup where a certain link on my Web Server is redirected to a FCGID module.
When I disable SELinux with 'setenforce 0' the site works. When I enable it won't work, it gives me a "Error 500 - Internal Server error".
I'm used to fixing these SELinux erros, but when I looked in the audit.log there's no new entry related to apache or my fcgid program. In the past, when I had a SELinux error, everytime I refreshed the page it gave me a new entry on this file.
My apache error_log adds these entries after refreshing the site:
[Thu Jul 07 14:34:04.240551 2016] [fcgid:warn] [pid 18054] (104)Connection reset by peer: [client XXX.XXX.XXX.XXX:34622] mod_fcgid: error reading data from FastCGI server
[Thu Jul 07 14:34:04.240629 2016] [core:error] [pid 18054] [client XXX.XXX.XXX.XXX:34622] End of script output before headers: rt-server.fcgi
This is my server information:
OS: CentOS Linux release 7.2.1511 (Core)
Apache: 2.4.6
FCGID program: Request Tracker 4.2 (rt-server.fcgid)
How can I be certain which policies I have to enable on SELinux for my module to work when I have no log on the audit.log file? Please don't tell me I have to read the code to know this. It's huge and written in Perl. Everytime someone opens a Perl script a puppy dies somewhere.
0 Answers