Jeffrey Ollie wrote:
The ping and traceroute tools in the Zabbix dashboard do not work
because the web front end does not know how to connect to the Zabbix
server to initiate the commands. The following patch adds that
configuration as well as tweaks a selinux boolean. The Zabbix web
front end runs on the app servers so this is affected by the change
freeze. The impact should be minor, reverting should be relatively
easy in case of problems.
diff --git a/manifests/servergroups/appRhel.pp
b/manifests/servergroups/appRhel.pp
index e43b275..a68b396 100644
--- a/manifests/servergroups/appRhel.pp
+++ b/manifests/servergroups/appRhel.pp
@@ -58,6 +58,7 @@ class appRhel {
}
## SElinux
+ selinux_bool { 'httpd_can_network_connect': bool => 'on' }
selinux_bool { 'httpd_can_network_connect_db': bool => 'on' }
selinux_bool { 'httpd_can_network_relay': bool => 'on' }
selinux_bool { 'httpd_enable_homedirs': bool => 'on' }
diff --git a/modules/zabbix/templates/zabbix.conf.php.erb
b/modules/zabbix/templates/zabbix.conf.php.erb
index c2cae5b..8486cf4 100644
--- a/modules/zabbix/templates/zabbix.conf.php.erb
+++ b/modules/zabbix/templates/zabbix.conf.php.erb
@@ -28,4 +28,8 @@ $DB_USER = "zabbix";
$DB_PASSWORD = "<%= zabbixPassword %>";
$IMAGE_FORMAT_DEFAULT = IMAGE_FORMAT_PNG;
+
+$ZBX_SERVER = "noc1.fedora.phx.redhat.com";
+$ZBX_SERVER_PORT = "10051";
+
?>
+1
-Toshio