alert日志出现pmon failed to acquire latch该如何收集信息

SRDC - How to Collect Standard Information for Issues Where ‘PMON failed to acquire latch, see PMON dump’ Warnings are Seen in the Alert Log (Doc ID 1951971.1)
1.Alert log covering the time that the warning occurred
收集alert日志
select value from v$diag_info where name =‘Diag Trace’;
2.Collect the PMON trace file
收集pmon trace file
select value from v$diag_info where name =‘Diag Trace’;
The PMON trace is named SID_pmon_PROCESSID.trc.
In an example where the SID is “v120101”, PROCESSID = 3841 and the directory the alert log resided in was “/dbadmin/diag/rdbms/v12101/trace”, the full path of the PMON trace would be:
/dbadmin/diag/rdbms/v12101/trace/v12101_pmon_3841.trc

Required diagnostics if the database hangs at the time of the warning

When the ‘PMON failed to acquire latch, see PMON dump’ warnings are seen in the alert log,
it is likely that the database will hang or appear to hang.
It is very important to collect diagnostic traces immediately when you encounter the hang AND
while it is occurring. If you capture data long after you first encounter a hang,
it may not provide enough information as to what originally caused the hang.
The quicker you can capture hang diagnostics the easier they are to interpret since there are
likely to be less blocked processes. For this reason capture hanganalyze and systemstate
dumps prior to collecting other information such as: AWR/ASH/STATSPACK, etc.

Specifically, Hanganalyze and Systems State dumps must be taken while the hang is occurring otherwise they are of little to no use for diagnosis.

3.Collect HangAnalyze traces
收集hanganalyze trace。
要及时获取，在awr等信息之前获取。 必须在系统挂起的时候去收集，否则几乎没有作用。
Generate 2 HangAnalyze Level 3 traces 1 minute apart, connected as the sys user:

sqlplus / as sysdba

SQL> oradebug setmypid

SQL> oradebug unlimit
SQL> oradebug tracefile_name
SQL> oradebug hanganalyze 3
The location of the hanganalyze trace is written to the screen by the oradebug tracefile_name command and also after the hanganalyze command has been run. It can be found in the Oracle trace directory.
Wait 1 minute and repeat.

4.Collect Systemstate dump
Generate 1 Systemstate Level 258 , connected as the sys user:

sqlplus / as sysdba

SQL> oradebug setmypid
SQL> oradebug unlimit
SQL> oradebug tracefile_name
SQL> oradebug dump systemstate 258
The location of the systemstate trace is written to the screen by the oradebug tracefile_name command. It can be found in the Oracle trace directory.

5.Collect OSWatcher (osw) data
If you have OSWatcher installed and it was collecting data over the period of the problem, then “tar up” the OSWatcher archive data and upload the entire archive directory to the SR. The location of OSWatcher data can be found by looking in /tmp/osw.hb. If not then omit this step. See Document 301137.1 for details about OSWatcher.

6.Collect any other Oracle trace files that got generated during the time of the hang
Check the Oracle trace directory for any other files whose timestamp is within the timestamp window of the hang.

7.生成awr、addm等报告。
8.其他有关的文章id
Resolving Issues Where ‘PMON failed to acquire latch, see PMON dump’ Warnings are Seen in the Alert Log (Doc ID 976714.1)
Document 468740.1"Pmon Failed To Acquire Latch" Messages in Alert Log -Database Hung
Document 4632780.8Bug 4632780 - PMON “failed to acquire latch” during shutdown
Document 8502963.8Bug 8502963 - PMON fails to acquire latch and publish connection load balancing