In this chapter, we talked about the importance of knowing the terminology related to clustering, including load balancing, high availability, failover, and replication. While each feature has its benefits, there may also be drawbacks, especially in terms of maintenance and complexity. The two foundations of clustering in WildFly are as follows:
The cluster node communication by JGroups
Data distribution and caching with Infinispan
Large parts of the Java EE stack and its components can be clustered and tuned. In WildFly, these components include the following:
The different EJBs
Web applications
Load balancing of HTTP traffic using mod_cluster
The persistence layer
HornetQ
The golden rule for whenever replication is involved is to make sure that the dataset is as small as possible.
Now you can tune WildFly. Get out there and test its wings!