Having found what you think is the reasonable balance for the memory you actually have, set them and turn ASMM off. Then Oracle will toss out old rarely used sql instead of slowly growing. Periodically (or review ADDM if you use it) your allocation and waits for either i/o or shared pool space to see if you're very far out of whack. I think of ASMM as a discovery process, not a long term solution for production. (By the way, growing 900MB in 3 months does not strick me as Fast.) Regards, mwf -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Purav Chovatia Sent: Monday, July 09, 2012 10:39 AM To: oracle-l@xxxxxxxxxxxxx Subject: Fast Growing Shared Pool Hi, We have a 10205 database on Solaris SPARC with ASMM enabled and sga_target & sga_max_size = 4G. What we observe is that shared_pool has grown from 700MB to 1.6GB in last 3 months whereas buffer cache has shrunk from 3.3G to 2.5G. This inspite of the database having to do physical reads i.e some of the hot objects not fitting in the buffer cache. What could be the reason? How do I find what is the breakup of the so big shared pool? (There are no bind variable issues) Thanks. -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l