Apache Web Server is called an Active/Active application with LifeKeeper. This means that more than one instance of Apache can be running on a server at any time. For example, if two servers are running an instance of Apache and one server fails, the Apache instance on this server can fail over to the other server and it can continue to run its own instance as well. Some applications simply don’t support this, so you would have to keep a server available for each instance of the application. These are called Active/Standby applications. Some applications can be configured either way.
There may be circumstances when you might want to operate Apache in an “active/standby” mode, particularly if only one of your servers is used primarily for running Apache. In this particular case, you should disable the automatic startup of the standard Apache default installation so that nothing is running on the backup server(s).
By manually bringing the Apache instances In Service on one or more particular servers, you can distribute the workload as you like. And by adjusting the server priorities for each of your instances, you can configure the Apache instances to fail over to a particular server only as a last resort, or to fail over to different servers to distribute the workload when a failure occurs.
If you disable automatic startup of Apache on all servers in the cluster, it is possible to use the default server root directory “/etc/http” for a single LifeKeeper Apache resource hierarchy by simply configuring this instance to use LifeKeeper IP addresses – and possibly using a shared file system for the document root directories. Note, however, that this would be an Active/Standby configuration (as described above), so you could no longer start up the default instance in the usual way. Of course, the default server root directory cannot be used for more than one hierarchy, since the server root must be unique.
Post your comment on this topic.