Summary

Reliability, availability, and scalability (RAS) are three critical attributes of managing the Quality of Services in a Java EE .NET interoperable solution. Managing RAS is a complex subject area that requires many micro-tuning exercises and iterative, trial-and-error changes. This chapter has identified some major design factors for RAS and related them to different integration strategies. The text has also reviewed the current states of existing RAS technologies and explored what they can do today.

The reliability requirements for an interoperable solution basically assumes that the service requests (or messages) and the business data should be reliable and accurate. Asynchronous and bridge integration strategies are more flexible and appropriate in supporting messaging reliability.

The availability requirements for an interoperable solution are usually implemented by clustering in the container level. Clustering supports session failover. Shared session data is another strategy that can support availability requirements.

The scalability requirements for an interoperable solution are supported by either upgrading the infrastructure or by micro-tuning each technical component. Sometimes changing the underlying integration strategy may be more appropriate to enhance the end-to-end system performance.