PMON failed to acquirelatch, see PMON dump

PMON failed to acquirelatch, see PMON dump these two days suddenly found a set of running Oracle 10.2.0.1 for RHEL 5.8x86 _ x64 database shut down very slowly, up to 4 minutes, shutdown immediate after alert. log reports a large number of "PMON failed to acquire latch, see PMON dump", and there are not many connections in this database. The current Process is 60, and no exception is reported except this error, there were no connection errors, but the number of processes remained high. I checked it online and said it was a bug. I found a solution on Metalink. Because the current database is temporarily suspended for upgrading and fixing, therefore, by modifying SQLNET. the ora parameter is tested for a period of time. **************************************** * ************************** [Plain] To implement the solution, execute the followingsteps: 1. in $ ORACLE_HOME/network/admin/listener. ora, of the database where the MetadataRepository is to be loaded, add the line: INBOUND_CONNECT_TIMEOUT_LISTENER = 0 2. in the OracleServer 10g sqlnet. ora, add the line: SQLNET. INBOUND_CONNECT_TIMEOUT = 0 3. stop andrestart the database and the listener for the changes to take effect This is on metalink: Applies to: Oracle Server-Enterprise Edition-Version: 10.2.0.1.0 to 10.2.0.3.0 This problem can occur on any platform. symptomsDatabase Instance hangs and connections to database using 'sqlplus 'are alsonot possible. checkingalert. log we see following messages PMON failed toacquire latch, see PMON dumpFri Oct 5 10:33:00 2007 PMON failed to acquire latch, see PMON dumpFri Oct 5 10:34:05 2007 PMON failed to acquire latch, see PMON dumpErrors in file/dwrac/BDUMP/dwhp_pmon_1912834.trc: This will alsodump a systemstate dump and the location will be mentioned in alert. log Also at OSlevel, we see that MMAN is consuming lot of CPU. causeCurrently this issue is being worked upon bydevelopment in Bug 6488694-databse hung with pmon failed to acquire latch message SolutionAs of now only workaround is to disable Automatic IC Shared Memory Management (ASMM) I. e Setting SGA_TARGET = 0 Also as per bug, you can set following event and restart the instance EVENT = "10235 trace name context forever, level 2" Development is suspecting memory resume uption in thiscase. so with the above event, database might hit ORA-600 before spin. thetrace file of ORA-600 wocould help investigate the issue. these trace files needto be sent to Oracle support for investigation. note:-Event10235 with level 2 or higher can impact latch contention. though may not bequite critical. in case you see latch contention then unset the event ReferencesBug 6488694-databse hung with pmon failed to acquire latch message ****************** **************************************** ********