Just to be clear, Terracotta, while it does cache and promise scalability by offloading from the primary database (and is pretty smart about it), some of the real magic in Terracotta is its ability to cluster POJOs (Plain Old Java Objects). So, you can take running java code and objects that are not worth storing in a database for performance reasons, or because they're transient, can be clustered to another machine. This can make Java app server failover much more elegant. I'm not qualified to weigh in on the right/wrong decision choice for performance optimization by using Terracotta to cache, but there are other use cases for it as well. Matt ________________________________ From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Sandra Becker Sent: Thursday, January 21, 2010 4:42 PM To: oracle-l Subject: Re: Understanding Terracotta caching Just an FYI for everyone.. They are already in contract negotiations for Terracotta. I was never consulted before the decision was made. It seems those who made the decision assumed that by caching a table that does a lot of disk reads, they will realize a performance gain when adding more users. Sandy Transzap, Inc.