From: roobaron Newsgroups: comp.databases.oracle.server Subject: Re: Library Caches Reports Date: Tue, 10 Sep 2002 00:42:16 +0000 Organization: dBforums Lines: 31 Sender: roobaron Message-ID: <1795219.1031618536@dbforums.com> References: <3d7ce17f.1688808@news.hccnet.nl> NNTP-Posting-Host: p-737.newsdawg.com User-Agent: dBforums X-Original-NNTP-Posting-Host: ns.spikeball.com X-Original-Trace: 10 Sep 2002 01:22:36 GMT, ns.spikeball.com hmmm ratio tuning :) Stuff could be getting aged out of the library cache due to a high turnover of different SQL over a period of time. i.e.period 1: Same SQL run over and over period 2: Different SQL run, requires reload, then run over and over. period 3: SQL from period 1 is required again etc. This might be what is happening, high get% and also high reload, the library cache is churning, but the database is finding the SQL required once it is loaded. You might want to check the size of the library cache (maybe publish the size here, along with version) Is the database waiting on Library cache latches? or is the main wait something else? You will get more mileage tuning the chronic waits first. Ratio tuning is hit and miss. have a look at v$session_wait, join by the sid from v$session. Have Fun -- Posted via http://dbforums.com