Scalix11 has broken Apache

[tubs1]

Can anyone help?

I have finally completed an install of Scalix 11 on a Fedora Care 5 server and as soon as the installation completed all the virual hosts I had setup on Apache were lost, no matter which URL I enter I get routed to the default document root of the Apache server /var/www/html

Backing out the Scalix installation removes this problem without any changed needed to my httpd.conf file

Any ideas what might have cuased this as I actually rent out hosting and the customers would not be happy if people entered their URL and were directed to my home page??

Cheers

[dougp23]

Apache and Tomcat make strange bedfellows. What happened to you happens to everyone who tries something with Tomcat. Basically, Tomcat sort of takes over handling requests from Apache. (Tomcat can serve static HTML, but we usually defer to Apache, Apache is much faster at it).

Here's a HOWTO on how to get back what you think you lost:

http://tomcat.apache.org/tomcat-3.2-doc ... howto.html

Don't be intimidated by the length of it. Just go step by step. It's not too bad.

[tubs1]

Gulp

It looks very scary. I think I will be spending my bank holiday making sure I understand the instructions before I start hacking around. Hopefully, from the brief look I just took, it might be down to the later part of the instructions where the crucial changes are needed and I am thinking perhaps one tomcat instance for all the virtual hosts and see how performance is hit?

Thank you very much for the pointer, much appreciated

[dkelly]

Please take a look at this link http://www.scalix.com/wiki/index.php?title=HowTos/ApacheConfiguration

which is in the FAQs that I posted at the top of each forum.

Cheers

Dave

[tubs1]

I recovered my virtual hosts by moving parts of the instance-*.conf to the httpd.conf file where I defined my other virtual hosts.

I now have the following error appear in /var/log/httpd/error_log :

[Wed Apr 11 13:30:04 2007] [error] (111)Connection refused: proxy: AJP: attempt to connect to 127.0.0.1:8009 (localhost) failed
[Wed Apr 11 13:30:04 2007] [error] ap_proxy_connect_backend disabling worker for (localhost)
[Wed Apr 11 13:30:04 2007] [error] proxy: AJP: failed to make connection to backend: localhost
[Wed Apr 11 13:30:10 2007] [error] (111)Connection refused: proxy: AJP: attempt to connect to 127.0.0.1:8009 (localhost) failed
[Wed Apr 11 13:30:10 2007] [error] ap_proxy_connect_backend disabling worker for (localhost)
[Wed Apr 11 13:30:10 2007] [error] proxy: AJP: failed to make connection to backend: localhost
[Wed Apr 11 13:41:01 2007] [error] (111)Connection refused: proxy: AJP: attempt to connect to 127.0.0.1:8009 (localhost) failed
[Wed Apr 11 13:41:01 2007] [error] ap_proxy_connect_backend disabling worker for (localhost)
[Wed Apr 11 13:41:01 2007] [error] proxy: AJP: failed to make connection to backend: localhost
[Wed Apr 11 13:41:08 2007] [error] proxy: AJP: disabled connection for (localhost)
[Wed Apr 11 13:41:09 2007] [error] proxy: AJP: disabled connection for (localhost)
[Wed Apr 11 13:41:10 2007] [error] proxy: AJP: disabled connection for (localhost)
[Wed Apr 11 13:41:11 2007] [error] proxy: AJP: disabled connection for (localhost)

catalina.out contains :

Apr 11, 2007 11:18:35 AM org.apache.jk.common.ChannelSocket init
INFO: JK: ajp13 listening on marigold/87.237.57.217:8009
Apr 11, 2007 11:18:35 AM org.apache.jk.server.JkMain start
INFO: Jk running ID=0 time=0/88 config=null


lsof | grep 8009 gives :

java 6960 root 31u IPv4 7450963 TCP marigold .mindsheet.com:8009 (LISTEN)



As far as I can see Tomcat is running , it is listening on 8009 for AJP13 messages , Apache is aware to forward requests for /webmail to Tomcat but the connection fails . Any idea why???

Cheers in advance for any help.

[dougp23]

Hmm...why are u using localhost?
I believe AJP wants either the FQDN or IP address. try that.

[tubs1]

I have just found that there was another config file in the httpd area that redirected the relevant request to Localhost instead of the machines proper IP. Once that file was removed this is all working nicely.

Our basic problem was that the Scalix standard install seems to over ride the virtual hosts we had defined before. In solving this by putting the calls to the Scalix Connector Config files within the relevent virtual host definition we got the system working to a point. However a call to /SAC generates a call to the machines actual name so the Scalix Connector Config files have to be attached to that DNS name as well.