What are the possible options if application server is unable to handle usage load?
What is the correct way of defining these loads(nominal and peak volumes and their response times??)
Which scalability mechanisms are generally used (are they implemented by application servers?
or by doing "clustering"??) and what are their alternatives? Thank you.
With WebSphere you do clustering by adding more application servers and HTTP Servers (if you have an IP Sprayer like Network Dispatcher).
But there are some code and other considerations before you scale.
For a very good book on scalability goto www.redbooks.ibm.com it's free. "WebSpphere 4.0 Advanced Edition Workload Management and Scalability" ....very damn good.....there's also one for WAS3.x