HP Serviceguard Enterprise Cluster Master Toolkit User Guide (5900-2145, April 2013)

The following are the two methods can be used to configure Apache Web server:
• Local configuration.
• Shared configuration.
Local configuration
In a local configuration, you must place the configuration and other web-site files on a single node,
and then replicate the files to all other nodes in the cluster. Also, in a typical local configuration,
nothing is shared between the nodes. Identical copies of the Apache server configuration file and
web documents reside in exactly the same locations on each node. You must maintain identical
copies of the apache components on the different nodes. This is useful when the information on
the web-page is static.
If you are using local configuration, you must ensure that the data is propagated to all nodes, and
is consistently maintained. A disadvantage of storing the configuration on local disks is that this
can increase the chance of the configuration for an Apache instance becoming inconsistent if
changes are made, but not distributed to all nodes that can run that Apache instance.
Shared configuration
In a typical shared configuration, the document root directories are all on the shared file system.
(Placing the SERVER_ROOT directory in the shared file system is optional.) The web documents
(along with the apache configuration file) are on shared storage and are accessible to all nodes
in the cluster, therefore, there is no need to maintain local identical copies of the files on each
node. The mount point of the shared file system must be identical across all the Apache package
nodes. Therefore, this is the recommended Apache package configuration.
An IP addresses (or domain addresses that maps to particular IP addresses) is assigned to each
website through the configuration file. These relocatable IP addresses are created for each Apache
package and added to its Package Control Script in case of legacy packages or the Package
ASCII file in case of modular packages. When the Apache package is switched over from one
node to another, this particular instance is stopped and IP addresses are then removed on the
primary node. The IP addresses are reallocated and the instance is started on the adoptive node.
After this process is complete, clients are automatically connected through these IP addresses to
the web site on the adoptive node.
Multiple Apache Instances configuration
Apache Web Server is a multi-instance application, which means more than one instance of the
Apache Web Server can run on a single node at the same time. For example, if two nodes are
running an instance of Apache and one node fails, the Apache instance on the failed node can
be successfully failed over to the healthy node. In addition, the healthy node can continue to run
its own instance along with the failover instance. Multiple Apache instance configuration can either
be done as a local configuration or shared configuration or a combination of both.
Configuring the Apache Web Server with Serviceguard
To manage an Apache Web Server by Serviceguard, you must modify the default Apache
configuration. Before creating and configuring Serviceguard packages, make sure that you complete
the following configurations for the Apache Web Server application for all cluster nodes:
• When the Apache Server is installed, the default instance is automatically configured to start
during system startup via the runlevel (rc) script "hpws22_apacheconf" in the /etc/rc.config.d/
directory by setting "HPWS22_APACHE_START=1". Ensure that you disable this by setting
"HPWS22_APACHE_START=0".
• The httpd.conf file associated with the default Apache instance has a Listen directive "Listen
80", which is equivalent to listening for all IP addresses at port 80. If you need to configure
100 Using an Apache Toolkit in a HP Serviceguard Cluster