Looks like a bug indeed ... I've once troubleshooted something similar... Check this: Bug 10082277 - Excessive allocation in PCUR heap of "kkscsAddChildNo" (ORA-4031) (Doc ID 10082277.8) https://supporthtml.oracle.com/ep/faces/secure/km/DocumentDisplay.jspx?id=10082277.8&h=Y Note that this bug description mentions the detailed permanent memory allocation breakdown (event 10235 level 65536), but think twice before enabling it (even if support recommends to) as there are a few hang/spin bugs related to setting it... -- Tanel Poder http://blog.tanelpoder.com On Thu, Jun 16, 2011 at 6:42 PM, Eagle Fan <eagle.f@xxxxxxxxx> wrote: > Hi: > > Thanks for that information. > > I run that script and found that the biggest one was the parent cursor: > > SQL> @curheaps 2038009379 65535 > old 20: KGLNAHSH in (&1) > new 20: KGLNAHSH in (2038009379) > old 21: and KGLOBT09 like ('&2') > new 21: and KGLOBT09 like ('65535') > > KGLNAHSH KGLHDPAR CHILD# KGLHDADR > KGLOBHD0 SIZE0 SIZE1 SIZE2 SIZE3 > ---------- ---------------- ---------- ---------------- ---------------- > ---------------------- -------- -------- -------- > KGLOBHD4 SIZE4 SIZE5 KGLOBHD6 SIZE6 SIZE7 > STATUS > ---------------- -------- -------- ---------------- -------- -------- > ---------- > 2038009379 0000000F3BC53E78 65535 0000000F3BC53E78 > 0000000F5BF1E648 *1883443712 *0 0 0 > 00 0 0 00 0 > 0 1 > > >