awr 无法生成处理

客户数据库遇到AWR快照无法生成的问题，下面记录了下处理的过程。

OS：Linux

DB：Oracle 10.2.0.5 RAC

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758

CURTIME Inst:Sid ID1 ID2 LMODE TYPE CTIME-------------------------- --------------------- ---------- ---------- ---------- ---- ----------20130927 16:45:19 Holder: 2:811 0 0 4 WF 243628820130927 16:45:19 Holder: 1:945 0 0 4 WF 243628720130927 16:45:19 Waiter: 2:709 0 0 0 WF 120130927 16:45:19 Holder: 1:945 9155 0 6 TM 243628720130927 16:45:19 Waiter: 2:811 9155 0 0 TM 2436287 /*WF AWR Flush MMON (Manageability Monitor) is attempting the autoflush function which means that we are trying to flush the ASH (Active Session History) data into the AWR (Active Workload Repository) for the historical repository.We take out the WF enqueue for doing this and hence in your wait sessions you may see WF enqueue contention.*/ SQL> select * from v$locked_object 2 XIDUSN XIDSLOT XIDSQN OBJECT_ID SESSION_ID ORACLE_USERNAME OS_USER_NAME PROCESS LOCKED_MODE---------- ---------- ---------- ---------- ---------- ------------------------------ ------------------------------ ------------ ----------- 10 11 5977841 9155 945 oracle 10185 6 10 11 5977841 9228 945 oracle 10185 3 SQL> SQL> select owner,object_name,object_type from dba_objects where object_id in ('9155','9228'); OWNER OBJECT_NAME OBJECT_TYPE------------------------------ -------------------------------------------------------------------------------- -------------------SYS WRH$_ACTIVE_SESSION_HISTORY TABLESYS WRM$_DATABASE_INSTANCE TABLE SQL> SQL> SQL> SELECT s.event event, 2 SUBSTR(s.program, 1, 39) program, 3 s.sid || ':' || s.SERIAL# || ':' || 4 DECODE(sql_hash_value, 0, prev_hash_value, sql_hash_value) || ':' || 5 decode(sql_id, null, prev_sql_id, sql_id) "SESS:SERIAL#:SQL_HASH:SQL_ID", 6 s.username, 7 SUBSTR(s.osuser || '@' || s.machine || '@' || s.process || '@' || 8 p.spid, 9 1, 10 60) "OSUSER@MACHINE@PROCEE@SPID", 11 BLOCKING_SESSION_STATUS, 12 blocking_session, 13 TO_CHAR(s.logon_time, 'mm-dd hh24:mi') logon_time 14 FROM v$session s, v$process p 15 where S.PADDR = P.ADDR 16 and s.wait_class <> 'Idle' 17 order by logon_time; EVENT PROGRAM SESS:SERIAL#:SQL_HASH:SQL_ID USERNAME OSUSER@MACHINE@PROCEE@SPID BLOCKING_SESSION_STATUS BLOCKING_SESSION LOGON_TIME-------------------------- ----------------------------------- -------------------------------------------------- ------------------------------ --------------------------------- ----------------------- ---------------- ---------------------------------------------------------------------------gc current request oracle@dbserver1 (m000) 945:12572:0: oracle@dbserver1@10185@10185 UNKNOWN 08-30 12:00

查看mmon文件，发现以下内容：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

*** 2013-09-27 16:37:39.385 *** KEWRAFC: Flush slave failed, AWR Enqueue Timeout *** 2013-09-27 16:38:39.431 *** KEWRAFC: Flush slave failed, AWR Enqueue Timeout *** 2013-09-27 16:39:39.471 *** KEWRAFC: Flush slave failed, AWR Enqueue Timeout *** 2013-09-27 16:40:41.216 *** KEWRAFC: Flush slave failed, AWR Enqueue Timeout -----kill 掉m000进程 [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ ps -ef|grep m0 oracle 10185 1 99 Aug30 ? 28-05:04:19 ora_m000_orcl1 oracle 21027 30461 0 17:10 pts/2 00:00:00 grep m0 [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ kill -9 10185 [oracle@dbserver1 osw]$ ps -ef|grep m0 oracle 21118 30461 0 17:10 pts/2 00:00:00 grep m0 [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ ps- ef|grep pmon -bash: ps-: command not found [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ ps -ef|grep pmon oracle 15120 1 0 Aug25 ? 00:16:00 ora_pmon_orcl1 oracle 21221 30461 0 17:10 pts/2 00:00:00 grep pmon [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ [oracle@dbserver1 osw]$ ---awr生成快照正常 SQL> SQL> select snap_id,instance_number, 2 to_char(begin_interval_time, 'yyyy-mm-dd hh24:mi:ss') begin_interval_time, 3 to_char(end_interval_time, 'yyyy-mm-dd hh24:mi:ss') end_interval_time 4 from dba_hist_snapshot 5 order by snap_id,instance_number; SNAP_ID INSTANCE_NUMBER BEGIN_INTERVAL_TIME END_INTERVAL_TIME ---------- --------------- --------------------------------------------------------------------------- --------------------------------------------------------------------------- 6519 2 2013-08-30 11:00:28 2013-08-30 12:00:31 6520 1 2013-08-30 12:00:31 2013-09-27 17:10:41 6520 2 2013-08-30 12:00:31 2013-09-27 17:10:41 6521 1 2013-09-27 17:10:41 2013-09-27 17:13:47 6521 2 2013-09-27 17:10:41 2013-09-27 17:13:47 SQL> SQL>

在之前尝试了MOS文档中的处理办法：

1 2 3 4 5

MOS： WHEN AWR SNAPSHOTS ARE NOT GENERATED AUTOMATICALLY, HOW To DIAGNOSE. (文档 ID 787409.1)中的方法 1. Take backup of table SYS.WRI$_OPTSTAT_HISTHEAD_HISTORY , Do create table as select of the table. 2. Start the database in restricted mode alter table sys.WRI$_OPTSTAT_HISTHEAD_HISTORY move; 3. Monitor the trace files

尝试了truncate wri$_optstat_histhread_history问题依旧，不知道是不是没有把数据库启动到restricted模式下进行操作的原因

文章转载于http://www.cddba.com/2013/09/27/awr%e6%97%a0%e6%b3%95%e7%94%9f%e6%88%90%e5%bf%ab%e7%85%a7%e5%a4%84%e7%90%86/