Error added: 2010-09-22T08:58:03Z
There is always the case of the interface IP address is not up. /sbin/ifconfig eth0 or eth1 should show your chosen interface IP address from /etc/cobbler/settings - the 'server' setting Confirm the IP address or hostname to IP address mapping and make sure you can access the HTTP page therein. lynx or links seem to be available for those with no Xorg X11 installed links http://133.133.133.133/ for instance lynx http://122.122.122.122/ is another example To debug the cobbler check command install strace strace -o /tmp/trouble_ticket cobbler check and wait for a while and hit Control-C and examine the log file /tmp/trouble_ticket.
If the httpd server is running when you install cobbler, you must restart the httpd server.
make sure the /etc/hosts in 127.0.0.1 has localhost at it. 127.0.0.1 localhost localhost.localdomain <ip> <hostname>
/etc/cobbler/settings has to have a \"server\" parameter that is equal to your actual IP...
This message obviously means it wasn't able to connect to the httpd server. list of things to check: 1) httpd is actually running and listening on all ports (does the test page come up in the browser?) 2) that the value used for the "server" parameter is either an IP address or a hostname that is currently resolving correctly (just running "dig" should work) 3) there are no iptables rules blocking the connection to the httpd (I would just "iptables -F && cobbler check" and if it worked reload the rules and see which one's blocking). 4) If SELinux is the issue (does "setenforce 0 && cobbler check" get passed that and give you something else?) 5) All else fails, check to see if /var/log/cobbler/cobbler.log has any sexiness.
setsebool -P httpd_can_network_connect=1
remove port number (:80) from the server's url in /etc/cobbler/settings i.e. it should be like following without a port number: server: 192.168.122.108 HOW I FOUND OUT: I opened cobbler's "cli.py" and searched for "httpd does" in it and add something like below to it's "except:" statement before "sys.exit": print >> stderr, self.url_cobbler_api Then after "sudo cobbler check", it printed for me httpd does not appear to be running and proxying cobbler http://192.168.122.108:80:80/cobbler_api Hope it helps :)
If you've checked your ports, your certs, daemons, and everything else seems to be fine, check the "client_use_https" setting in the cobbler server settings file. In addition to getting the httpd server and SELINUX error, I got “SSL23_GET_SERVER_HELLO:unknown protocol” when trying to access via web. This happened also when trying to use curl from the same machine while ignoring cert validation, trying all the different ssl protocols (v2/3, tls 1.0+), and none of it worked. Just remember to restart cobblerd. :)
Can also be caused by a name resolution failure ie dns does not match, for whatever reason, the server name. The nsswitch.conf file should have files dns specified for the hosts entry. Server has two IP's 10.0.0.1 - management 10.0.0.2 - live The satellite server has been set up on 10.0.0.1 The nsswitch.conf file hosts entry is set to dns files The dns entry for the host is 10.0.0.2 The httpd not running ................................... message will be displayed Solution: 1. Use IP addresses instead of host names 2. Set the hosts entry of nsswitch.conf to files dns and ahh 10.0.0.1 server_name to your hosts file. Hope this helps.
I would also add that doing a `service cobblerd restart && cobbler sync` is a good check as well
If you know the answer, please add your own solution below.
If you don't know, but find out later, please come back and share your answer - there will be other people
struggling with this too.
If you want to be notified via email when this is solved, enter your email address here: