Each cluster includes a primary server, and a group of participating servers:
The primary server contains the master copy of the configuration repository for all servers in the cluster. The primary server distributes (synchronizes) its configuration to the other servers in the cluster.
Participating, or nonprimary servers, must have unique names.
A cluster consists of at least two servers. Each instance of EAServer operates its own independent name service, and each cluster member knows the connection information for all the other members. See “How does the EAServer naming service work?”.
When you configure a cluster, you can define cluster partitions, which determine where remote clients using lookup can find components. To ensure high availability, a cluster partition must include at least two physical servers. For load balancing, a cluster partition may contain only one server.
Cluster support is tightly integrated with the EAServer naming service, so that all client services from a cluster are made available through the naming service. See Chapter 5, “Naming Services” for more information.
The SessionManager::Session::lookup operation implicitly consults the naming service, so you can write a client that does not explicitly use the naming service but still takes advantage of cluster services. For example, the PowerBuilder connection object uses SessionManager::Session::lookup.
Typically, each server in a cluster runs on a different host, so each server has its own copy of the entire repository and all the files required for component execution. Sybase recommends that you run each cluster member from its own installation directory.
Cluster members can be on different platforms
A cluster can include servers on different platforms, including
Windows, Solaris, AIX, HP, and Linux. Each server in the cluster
must be the same version of EAServer. If a cluster uses PowerBuilder
components, all the servers must use the same version of the PBVM.
