WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK!

環境： AIX5.3+hacmp5.3+oracle 10.2.0.2 RAC

故障現象：
- instance1 hang (sqlplus 都不能進去）
- alert log 裡出現：WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK

解決過程：

[@more@]

環境： AIX5.3+hacmp5.3+oracle 10.2.0.2 RAC

故障現象：
- instance1 hang (sqlplus 都不能進去）
- alert log 裡出現：WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK

解決過程：

DATA COLLECTED
===================
Tue Oct 30 08:19:18 2007
GES: Potential blocker (pid=917874) on resource TX-00320007-00028042;
enqueue info in file /oracle/app/oracle/admin/prdtest/udump/prdtest1_ora_966744.trc and DIAG trace file
Tue Oct 30 08:53:01 2007
GES: Potential blocker (pid=540996) on resource CI-0000001E-00000002;
enqueue info in file /oracle/app/oracle/admin/prdtest/bdump/prdtest1_lmd0_991268.trc and DIAG trace file
Tue Oct 30 09:02:34 2007
GES: Potential blocker (pid=856078) on resource DR-00000000-00000000;
enqueue info in file /oracle/app/oracle/admin/prdtest/bdump/prdtest1_lmd0_991268.trc and DIAG trace file
Tue Oct 30 09:04:30 2007
MMNL absent for 1989 secs; Foregrounds taking over
Tue Oct 30 09:22:34 2007
>>> WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=107
System State dumped to trace file /oracle/app/oracle/admin/prdtest/udump/prdtest1_ora_8809
90.trc
Tue Oct 30 09:24:59 2007
>>> WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=83
System State dumped to trace file /oracle/app/oracle/admin/prdtest/udump/prdtest1_ora_987404.trc
Tue Oct 30 09:28:11 2007
>>> WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=80
System State dumped to trace file /oracle/app/oracle/admin/prdtest/bdump/prdtest1_q004_733658.trc
Tue Oct 30 09:36:21 2007
>>> WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=16
System State dumped to trace file /oracle/app/oracle/admin/prdtest/bdump/prdtest1_reco_856078.trc
Tue Oct 30 09:36:23 2007
MMNL absent for 3843 secs; Foregrounds taking over

解決方案：

From the systemstate and AWR report, I see that there is heavey contention for and internal sequence SYS.AUDSES$ .
Please carry out the following :
SQL> conn / as sysdba
SQL> alter sequence SYS.AUDSES$ cache 10000 ;

Also, the please increase the shared pool to atleast 3.5 GB.

Let me know if the issue occurs after this or not.
If it occurs again, the upload similar set of files and AWR report.