ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist

一：系统环境

操作系统：

 uname -a
AIX egapdb2 1 6 00F81BFA4C00

数据库：

11.2.0.3.5

 

二：错误信息

    在alert.log发现如下错误信息：

DBMS_STATS: GATHER_STATS_JOB encountered errors. Check the trace file. 
Errors in file /apps/oracle/diag/rdbms/egapdb/egapdb2/trace/egapdb2_j003_13369700.trc: 
ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist 

trace文件中信息如下：

Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production 
With the Partitioning, Real Application Clusters, OLAP, Data Mining 
and Real Application Testing options 
ORACLE_HOME = /apps/oracle/product/11.2.0.3/db_1 
System name: AIX 
Node name: egapdb2 
Release: 1 
Version: 6 
Machine: 00F81BFA4C00 
Instance name: egapdb2 
Redo thread mounted by this instance: 2 
Oracle process number: 780 
Unix process pid: 13369700, image: oracle@egapdb2 (J003) 


*** 2014-07-07 22:00:17.682 
*** SESSION ID:(1514.32367) 2014-07-07 22:00:17.682 
*** CLIENT ID:() 2014-07-07 22:00:17.682 
*** SERVICE NAME:(SYS$USERS) 2014-07-07 22:00:17.682 
*** MODULE NAME:(DBMS_SCHEDULER) 2014-07-07 22:00:17.682 
*** ACTION NAME:(ORA$AT_OS_OPT_SY_16083) 2014-07-07 22:00:17.682 

ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist 

*** 2014-07-07 22:00:17.682 
DBMS_STATS: GATHER_STATS_JOB: GATHER_TABLE_STATS('"EGCMS"','"ET$0B72AAD30001"','""', ...) 
DBMS_STATS: ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist 

*** 2014-07-07 22:00:17.708 
DBMS_STATS: GATHER_STATS_JOB: GATHER_TABLE_STATS('"EGCMS"','"ET$044324A10001"','""', ...) 
DBMS_STATS: ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist 

*** 2014-07-07 22:00:17.719 
DBMS_STATS: GATHER_STATS_JOB: GATHER_TABLE_STATS('"EGCMS"','"ET$070CA05B0001"','""', ...) 
DBMS_STATS: ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not exist 

*** 2014-07-07 22:00:17.729 
DBMS_STATS: GATHER_STATS_JOB: GATHER_TABLE_STATS('"EGCMS"','"ET$0D4D92ED0001"','""', ...) 
DBMS_STATS: ORA-20011: Approximate NDV failed: ORA-06564: object EXPDDDP does not 

##"ET$XXXX"是系统自动命名的外部表表名（datapump的使用可能会产生系统命名的外部表）

关于DataPump的external_table模式可以参见Maclean Liu的博客：

http://www.oracledatabase12g.com/archives/%E5%85%B3%E4%BA%8Edatapump%E7%9A%84external_table%E6%A8%A1%E5%BC%8F.html

三：问题分析

1.进行如下信息查询

conn / as sysdba 
set pages 100 
set num 20 
alter session set nls_date_format='yyyy-mm-dd hh24:mi:ss'; 
set mark html on 
spool info.html 
select * from dba_external_tables order by 1,2; 
select * from dba_directories; 
select * from dba_objects where object_name like 'ET$%' or object_name='EXPDDDP'; 
select * from dba_datapump_jobs where job_name not like 'BIN$%' order by 1, 2; 
select o.status, o.object_id, o.object_type, o.owner || '.' || object_name "OWNER.OBJECT" from dba_objects o, dba_datapump_jobs j where o.owner = j.owner_name and o.object_name = j.job_name and j.job_name not like 'BIN$%' order by 4, 2; 
spool off 
exit 

   上面脚本查得的信息请见附件info.html（下载链接如下）

http://download.csdn.net/detail/shaochenshuo/7641089

(CSDN博客不能上传附件，只能放在资源里，有点不方便啊！如果需要看的朋友就去资源里下载吧，没办法)

   通过dba_external_tables查询信息我们可以判断"ET$XXXX"表是由datapump任务产生的外部表，外部表的路径为EXPDDDP

   通过dba_directories视图查询发现系统中已经不存在EXPDDDP路径（可能是任务后被删除），所以alert日志会报ORA-06564: object EXPDDDP does not exist 
。但是这并不是此错误的根本原因（根本原因是“ETXXXX”外部表没有被清除）

   通过dba_datapump_jobs视图发现有两个not running的import任务，从dba_objects视图可以查看到两张同datapump任务名相同的表SYS_IMPORT_TABLE_02，SYS_IMPORT_TABLE_01（master 表）

##由上面的综合分析我们可以得出问题的原因，即datapump任务产生的临时表不知是什么原因，没有被清除。导致GATHER_STATS_JOB在收集该表统计信息时报错

SQL> select OWNER,TABLE_NAME,NUM_ROWS,LAST_ANALYZED,STALE_STATS from dba_tab_statistics where table_name like 'ET%';

OWNER      TABLE_NAME             NUM_ROWS LAST_ANALYZE STA
---------- -------------------- ---------- ------------ ---
EGCMS      ET$044324A10001
EGCMS      ET$070CA05B0001
EGCMS      ET$072384130001
EGCMS      ET$0B72AAD30001
EGCMS      ET$0D4D92ED0001

    从dba_tab_statistics表中可以得出这些外部表确实需要收集统计信息。

##但是具体是什么原因导致datapump任务产生的外部表没有被正常清除目前还不清楚，提了SR求助oracle原厂工程师，他们也没给出具体原因，只说可能是datapump任务被异常终止时会在dba_external_tables视图中出现“ETXXXX”表
##如果后面发现了具体原因会及时发布

四，解决方案

    1.删除报错中涉及到的相关“ETXXXX”外部表

drop table EGCMS.ET$044324A10001 purge;
drop table EGCMS.ET$070CA05B0001 purge;
drop table EGCMS.ET$072384130001 purge;
drop table EGCMS.ET$0B72AAD30001 purge;
drop table EGCMS.ET$0D4D92ED0001 purge;

    2.删除datapump任务的master表

select o.status, o.object_id, o.object_type, o.owner || '.' || object_name "OWNER.OBJECT" from dba_objects o, dba_datapump_jobs j where o.owner = j.owner_name and o.object_name = j.job_name and j.job_name not like 'BIN$%' order by 4, 2;

##注意做此操作时要确认被删除master表的不是暂停的，而是失败的不会再重新开始的任务

可以参考我的另外一篇博客确认并删除：

http://blog.csdn.net/shaochenshuo/article/details/37599303