I fully agree with you that tuning from this perspective would not be a good idea. I am just trying to figure out if this is related to a library cache mutex x wait which I believe maybe related to a vpd policy. I noticed these high miss percentages. I cannot find SQL AREA STATS or SQL AREA BUILD documented and it makes want to know the answer. The library cache mutex x wait was fixed by recreating the VPD policy which make me suspicious. To explain further I had about 6 query's running in parallel 8 on a Oracle RAC environment. The query's were running with parallel local so each query was not running across the cluster. These query's were selecting against 2 very large range hash partitioned tables. Thanks, Matt -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Mark Bobak Sent: Wednesday, October 23, 2013 4:32 PM To: mvshelton@xxxxxxxxxxxxx; oracle-l@xxxxxxxxxxxxx Subject: Re: Oracle Library Cache Hi Matt, My first thought would be, why were you compelled to look at that section of the AWR report? Do you have a performance problem? If so, what performance indicators pointed you in the direction of library cache hit ratios? In general, looking at hit ratios in isolation is probably not a valid tuning strategy. And no, I must admit, I don¹t know what ŒSQL AREA STATS¹ or ŒSQL AREA BUILD¹ means, in this context. -Mark On 10/23/13, 2:32 PM, "mvshelton@xxxxxxxxxxxxx" <mvshelton@xxxxxxxxxxxxx> wrote: >SQL> select namespace,gets,gethits,round(gethitratio,2) gethitratio > 2 from V$LIBRARYCACHE > 3 where namespace like 'SQL%'; > >NAMESPACE >GETS >---------------------------------------------------------------- >---------- > GETHITS GETHITRATIO >---------- ----------- >SQL AREA >13927433 > 13762659 .99 > >SQL AREA STATS >201188 > 27167 .14 > >SQL AREA BUILD >265640 > 109679 .41 > >I noticed in my AWR report that a database had high pct misses for SQL >AREA STATS of 87.56% and SQL AREA BUILD of 37% but I could not find any >information that describes these library cache waits. I can query from >v$library cache and get the hit ratio but I am not sure what this stat >is. Any help would be appreciated. > >Thanks, Matt > > >-- >//www.freelists.org/webpage/oracle-l > > > > -- //www.freelists.org/webpage/oracle-l ----- No virus found in this message. Checked by AVG - www.avg.com Version: 2013.0.3426 / Virus Database: 3222/6775 - Release Date: 10/23/13 -- //www.freelists.org/webpage/oracle-l