Oracle - dg坏块修复(二)
一、概述
本文是坏块修复(一)的续篇，这篇文章将介绍如何在dg环境中模拟坏块，以及出现坏块该如何修复。实验分为以下几个步骤。

1. 主库表出现坏块
2. dg库表出现坏块

二、环境准备
本实验是在oracle 11G，主库 + ADG环境下进行

1. 准备相关表
   create tablespace tbs01 datafile '/u01/app/oracle/oradata/orcltest/tbs01.dbf' size 100m;
   create table scott.t01 tablespace tbs01 as select * from dba_objects where rownum<=100;

select object_id, rowid, dbms_rowid.rowid_relative_fno(rowid) file_id, dbms_rowid.rowid_block_number(rowid) block_id from scott.t01;

复制代码
OBJECT_ID ROWID FILE_ID BLOCK_ID

    20 AAAVpfAAGAAAACDAAA          6        131
    46 AAAVpfAAGAAAACDAAB          6        131
    28 AAAVpfAAGAAAACDAAC          6        131
    15 AAAVpfAAGAAAACDAAD          6        131
    ...
    99 AAAVphAAGAAAACEAAJ          6        132
   100 AAAVphAAGAAAACEAAK          6        132
   101 AAAVphAAGAAAACEAAL          6        132

复制代码

2. 全库备份
   RMAN> backup database; // 全库备份

RMAN> list backup; // 查看备份

复制代码

List of Backup Sets

BS Key Type LV Size Device Type Elapsed Time Completion Time

19 Full 1.08G DISK 00:01:59 12-MAR-20

    BP Key: 19   Status: AVAILABLE  Compressed: NO  Tag: TAG20200312T150629
    Piece Name: /home/oracle/backupdir/ORCLTEST_2750922031_40_1_20200312_1034867190.bkp

List of Datafiles in backup set 19
File LV Type Ckp SCN Ckp Time Name
---- -- ---- ---------- --------- ----
1 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/system01.dbf
2 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/sysaux01.dbf
3 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/undotbs01.dbf
4 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/users01.dbf
5 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/example01.dbf
6 Full 1148218 12-MAR-20 /u01/app/oracle/oradata/orcltest/tbs01.dbf
复制代码

三、主库表出现坏块

1. 模拟坏块
   RMAN> blockrecover datafile 6 block 131 clear; // 将131数据块清空，即相当于产生了坏块

SQL> select * from scott.t01; // 对表进行查询，正常查询
看过我上一篇文章的就会知道，照道理这里应该会报错，但是实际并没有

查看alert日志

2. 检测坏块
   RMAN> backup check logical validate datafile 6;

复制代码

List of Datafiles

File Status Marked Corrupt Empty Blocks Blocks Examined High SCN

6 OK 0 11187 12800 1255050
File Name: /u01/app/oracle/oradata/orcltest/tbs01.dbf
Block Type Blocks Failing Blocks Processed
---------- -------------- ----------------
Data 0 1248
Index 0 195
Other 0 170
复制代码
并没有发现任何坏块，结合前面alert日志，可以看到主库的坏块已经被自动修复了，这个其实是ADG的功能（自动修复主库的坏块）。

四、dg库表出现坏块

1. 模拟坏块
   RMAN> blockrecover datafile 6 block 131 clear; // 将131数据块清空，即相当于产生了坏块

RMAN> backup check logical validate datafile 6; // 使用rman检测坏块

复制代码

List of Datafiles

File Status Marked Corrupt Empty Blocks Blocks Examined High SCN

6 FAILED 0 11187 12800 1574361
File Name: /u01/app/oracle/oradata/orcltestdg/tbs01.dbf
Block Type Blocks Failing Blocks Processed
---------- -------------- ----------------
Data 1 1249
Index 0 194
Other 0 170

validate found one or more corrupt blocks
复制代码

SQL> select * from scott.t01; // 对表进行查询，直至报错，时间可能稍微有点久

select * from scott.t01

                *

ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 6, block # 131)
ORA-01110: data file 6: '/u01/app/oracle/oradata/orcltestdg/tbs01.dbf'

2. 修复坏块，使用备份文件修复
   主库
   scp ORCLTEST_2750922031_40_1_20200312_1034867190.bkp 10.40.16.121:~ // 主库拷贝备份到dg库中

dg库
RMAN> catalog start with '/home/oracle/'; // 注册备份到dg库中
SQL> alter database recover managed standby database cancel; // 先关闭dg的同步
SQL> shutdown immediate // 停库
rm -rf /u01/app/oracle/oradata/orcltestdg/tbs01.dbf // 删除有坏块的数据文件
SQL> startup mount // 启库到mount状态
RMAN> restore datafile 6; // 还原数据文件
SQL> alter database recover managed standby database disconnect from session; // 打开mrp进程恢复数据库
SQL> alter database recover managed standby database cancel; // 恢复数据库一段时间后关闭mrp进程
SQL> alter database open; // 启库到open状态
SQL> alter database recover managed standby database using current logfile disconnect from session; // 打开mrp进程

SQL> select * from scott.t01; // 查询表正常

RMAN> backup check logical validate datafile 6; // 使用rman检测坏块，已经没了坏块

复制代码

List of Datafiles

File Status Marked Corrupt Empty Blocks Blocks Examined High SCN

6 OK 0 1 12801 1574361
File Name: /u01/app/oracle/oradata/orcltestdg/tbs01.dbf
Block Type Blocks Failing Blocks Processed
---------- -------------- ----------------
Data 0 1249
Index 0 194
Other 0 11356
复制代码

3. 补充说明
   最开始我想利用blockrecover的方式去修复坏块，结果不行，报错信息如下：
   RMAN> blockrecover corruption list;

复制代码
Starting recover at 16-MAR-20
using channel ORA_DISK_1

channel ORA_DISK_1: restoring block(s)
channel ORA_DISK_1: specifying block(s) to restore from backup set
restoring blocks of datafile 00006
channel ORA_DISK_1: reading from backup piece /home/oracle/ORCLTEST_2750922031_40_1_20200312_1034867190.bkp
channel ORA_DISK_1: piece handle=/home/oracle/ORCLTEST_2750922031_40_1_20200312_1034867190.bkp tag=TAG20200312T150629
channel ORA_DISK_1: restored block(s) from backup piece 1
channel ORA_DISK_1: block restore complete, elapsed time: 00:00:01

starting media recovery
media recovery failed
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 03/16/2020 14:23:00
ORA-00283: recovery session canceled due to errors
ORA-01122: database file 6 failed verification check
ORA-01110: data file 6: '/u01/app/oracle/oradata/orcltestdg/tbs01.dbf'
ORA-01207: file is more recent than control file - old control file
复制代码

五、总结

1. 主库的坏块，可以通过adg自动修复，不需要人工干预。
2. dg的坏块需要通过修复数据文件的方式去处理，不能使用blockrecover，如果大家有更好的方法，欢迎留言。
3. 逻辑坏块和物理坏块：
   a. 逻辑坏块一般是oracle系统bug造成，指的是块内的数据逻辑存在问题，比如说索引块的索引值没有按从小到大排列(非官方解释)。逻辑坏块一般会伴随着ora-600和ora-1578。检测逻辑坏块使用RMAN> backup check logical validate...
   b. 物理坏块一般是由于底层os/disk系统错误/损坏，导致数据块被修改。常见的物理坏块有块头和块尾信息不一致，checksum值无效，数据块信息全部为0等情况，并且可能伴随错误ORA-1578和ORA-1110
4. 数据库坏块的相关参数 db_block_checksum：默认是typical，不需要更改。该参数控制写入数据文件的时候是否将checksum值写入块头，当第二次读取该块时，重新计算checksum值，并与块头的checksum值进行比较，如果两次checksum值不一样，则标记该块为坏块。用于快速发现物理坏块。
   db_block_checking：默认是fasle，可以将其改为true。该参数控制块发生任何变化的时候进行逻辑上的完整性和正确性检查。用于阻止内存逻辑坏块和数据逻辑坏块。但会增加1%-10%的性能消耗。
5. 坏块优秀文章分享
   物理坏块 https://blogs.oracle.com/database4cn/oraclecorruption-，该文章介绍了如果没有备份如何通过dbms_repair这个包去将坏块跳过，并把表备份出来。

rman>blockrecover datafile 3 block 2,150,152;

RMAN> blockrecover corruption list;

Starting recover at 14-MAY-23
using channel ORA_DISK_1

starting media recovery
media recovery complete, elapsed time: 00:00:00

Finished recover at 14-MAY-23

RMAN>

variable 1 varchar2;
exec :1:='HGV2305200017879'
variable 2 varchar2;
exec :2:='5569093'
variable 3 varchar2;
exec :3:='FQC_3D'

select * from BYDMESAMF.MES2_BYD_SFC_KEY t where t.sfc = :1 and t.name = :2 and t.value = :3

select * from BYDMESAMF.MES2_BYD_SFC_KEY t where t.sfc = 'HGV2305200017879' and t.name = '5569093' and t.value = 'FQC_3D?'

DG主备同步状态异常。alert日志报错ORA-01274，MRP进程异常中断。
问题现象
具体现象
问题分析
问题原因
解决方案：
扩展
备库standby_file_management参数为auto
备库standby_file_management参数为manual
备库standby_file_management参数为auto
结论
问题现象
DG主备同步状态异常。alert日志报错ORA-01274，MRP进程异常中断。

具体现象
SQL> col client_pid for a10
SQL> SELECT inst_id, thread#, process, pid, status, client_process, client_pid,
2 sequence#, block#, active_agents, known_agents FROM gv$managed_standby ORDER BY thread#, pid;

INST_ID THREAD# PROCESS PID STATUS CLIENT_P CLIENT_PID SEQUENCE# BLOCK# ACTIVE_AGENTS KNOWN_AGENTS

   1        0 ARCH             2702 CONNECTED    ARCH     2702               0          0             0            0
   1        0 ARCH             2706 CONNECTED    ARCH     2706               0          0             0            0
   1        0 ARCH             2708 CONNECTED    ARCH     2708               0          0             0            0
   1        0 RFS              2840 IDLE         UNKNOWN  2830               0          0             0            0
   1        0 RFS              2858 IDLE         UNKNOWN  2824               0          0             0            0
   1        0 RFS              2869 IDLE         ARCH     2828               0          0             0            0
   1        1 ARCH             2704 CLOSING      ARCH     2704              62          1             0            0
   1        1 RFS              2860 IDLE         LGWR     2832              63       1439             0            0

8 rows selected.

日志应用状态：

SQL> 2 FROM (select thread# thrd, MAX(sequence#) almax
3 FROM v$archived_log
4 WHERE resetlogs_change#=(SELECT resetlogs_change# FROM v$database) GROUP BY thread#) al,
5 (SELECT thread# thrd, MAX(sequence#) lhmax
6 FROM v$log_history
7 WHERE resetlogs_change#=(SELECT resetlogs_change# FROM v$database) GROUP BY thread#) lh
8 WHERE al.thrd = lh.thrd;

Thread Last Seq Received Last Seq Applied

     1                65               62

1 row selected.

查看同步状态

SELECT * FROM v$dataguard_stats WHERE name LIKE '%lag%';SQL>

NAME VALUE UNIT TIME_COMPUTED DATUM_TIME

transport lag +00 00:00:00 day(2) to second(0) interval 05/09/2020 16:15:43 05/09/2020 16:15:43
apply lag +00 00:05:49 day(2) to second(0) interval 05/09/2020 16:15:43 05/09/2020 16:15:43

2 rows selected.

已经出现将近6min的lag。所以现在dg的同步状态是异常的。

查看主库alert日志：

Sat May 09 16:09:54 2020
alter tablespace zhuo add datafile size 20M
Completed: alter tablespace zhuo add datafile size 20M

查看备库的alert日志：

Sat May 09 16:09:54 2020
File #6 added to control file as 'UNNAMED00006' because
the parameter STANDBY_FILE_MANAGEMENT is set to MANUAL
The file should be manually created to continue.
MRP0: Background Media Recovery terminated with error 1274
Errors in file /u01/app/oracle/diag/rdbms/zhuodg/zhuodg/trace/zhuodg_mrp0_2906.trc:
ORA-01274: cannot add datafile '/u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_hcdsblpt_.dbf' - file could not be created
Managed Standby Recovery not using Real Time Apply
Recovery interrupted!
Recovered data files to a consistent state at change 789594
MRP0: Background Media Recovery process shutdown (zhuodg)

发现主库在添加数据文件成功后，备库报错ORA-01274。并且MRP0进程异常关闭。

先再备库手动重启下MRP进程：

SQL> recover managed standby database using current logfile disconnect;
Media recovery complete.

alert日志：

Sat May 09 16:11:42 2020
ALTER DATABASE RECOVER managed standby database using current logfile disconnect
Attempt to start background Managed Standby Recovery process (zhuodg)
Sat May 09 16:11:42 2020
MRP0 started with pid=28, OS id=3140
MRP0: Background Managed Standby Recovery process started (zhuodg)
Serial Media Recovery started
Managed Standby Recovery starting Real Time Apply
Sat May 09 16:11:47 2020
Errors in file /u01/app/oracle/diag/rdbms/zhuodg/zhuodg/trace/zhuodg_dbw0_2668.trc:
ORA-01186: file 6 failed verification tests
ORA-01157: cannot identify/lock data file 6 - see DBWR trace file
ORA-01111: name for data file 6 is unknown - rename to correct file
ORA-01110: data file 6: '/u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006'
File 6 not verified due to error ORA-01157
MRP0: Background Media Recovery terminated with error 1111
Errors in file /u01/app/oracle/diag/rdbms/zhuodg/zhuodg/trace/zhuodg_mrp0_3140.trc:
ORA-01111: name for data file 6 is unknown - rename to correct file
ORA-01110: data file 6: '/u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006'
ORA-01157: cannot identify/lock data file 6 - see DBWR trace file
ORA-01111: name for data file 6 is unknown - rename to correct file
ORA-01110: data file 6: '/u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006'
Managed Standby Recovery not using Real Time Apply
MRP0: Background Media Recovery process shutdown (zhuodg)
Completed: ALTER DATABASE RECOVER managed standby database using current logfile disconnect

也是有一大堆报错，都是备库不能锁定这个文件。
在主备库查询这个数据文件的状态：
主库：

 FILE# NAME                                                                                                 STATUS

     1 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_system_gxd20h14_.dbf                                     SYSTEM
     2 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_sysaux_gxd20k1y_.dbf                                     ONLINE
     3 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_undotbs1_gxd20lnp_.dbf                                   ONLINE
     4 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_users_gxd20pxk_.dbf                                      ONLINE
     5 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_gxdcfr5s_.dbf                                       ONLINE
     6 /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_hcdsblpt_.dbf                                       ONLINE  ----<<<新添加的数据文件

6 rows selected.
备库

SQL> select file#,name,status from v$datafile;

 FILE# NAME                                                                                                 STATUS

     1 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_system_05uvp9bj_.dbf                            SYSTEM
     2 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_sysaux_06uvp9bj_.dbf                            ONLINE
     3 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_undotbs1_07uvp9bj_.dbf                          ONLINE
     4 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_users_08uvp9c4_.dbf                             ONLINE
     5 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_04uvp9bj_.dbf                              ONLINE
     6 /u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006                                             RECOVER ----<<<同步过来的数据文件命名和状态都不对

问题分析
根据mos：Background Media Recovery terminated with ORA-1274 after adding a Datafile (Doc ID 739618.1)
1)查看参数

SQL> show parameter standby;

NAME TYPE VALUE

standby_archive_dest string ?/dbs/arch
standby_file_management string MANUAL

2）重命名数据文件

SQL> alter database create datafile '/u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006' as new;

Database altered.

SQL> select file#,name,status from v$datafile;

 FILE# NAME                                                                                                 STATUS

     1 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_system_05uvp9bj_.dbf                            SYSTEM
     2 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_sysaux_06uvp9bj_.dbf                            ONLINE
     3 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_undotbs1_07uvp9bj_.dbf                          ONLINE
     4 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_users_08uvp9c4_.dbf                             ONLINE
     5 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_04uvp9bj_.dbf                              ONLINE
     6 /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_hcdw395t_.dbf                              RECOVER

6 rows selected.

3）设置参数
SQL> alter system set standby_file_management=auto;

System altered.
4)重启mrp进程
SQL> recover managed standby database using current logfile disconnect;
Media recovery complete.

alert日志：

Sat May 09 17:00:11 2020
ALTER DATABASE RECOVER managed standby database using current logfile disconnect
Attempt to start background Managed Standby Recovery process (zhuodg)
Sat May 09 17:00:11 2020
MRP0 started with pid=19, OS id=3386
MRP0: Background Managed Standby Recovery process started (zhuodg)
Serial Media Recovery started
Managed Standby Recovery starting Real Time Apply
Waiting for all non-current ORLs to be archived...
All non-current ORLs have been archived.
Media Recovery Log /u01/app/oracle/oradata/arch/zhuodg/ZHUODG/archivelog/2020_05_09/o1_mf_1_63_hcdsmnt1_.arc
Media Recovery Log /u01/app/oracle/oradata/arch/zhuodg/ZHUODG/archivelog/2020_05_09/o1_mf_1_64_hcdsmx23_.arc
Media Recovery Log /u01/app/oracle/oradata/arch/zhuodg/ZHUODG/archivelog/2020_05_09/o1_mf_1_65_hcdsn2tw_.arc
Media Recovery Waiting for thread 1 sequence 66 (in transit)
Recovery of Online Redo Log: Thread 1 Group 4 Seq 66 Reading mem 0
Mem# 0: /u01/app/oracle/oradata/zhuodg/ZHUODG/onlinelog/o1_mf_4_hcdrt3nx_.log
Completed: ALTER DATABASE RECOVER managed standby database using current logfile disconnect

—日志传输已经正常，以前没应用的日志已经应用。

5）查看同步状态

SQL> SELECT * FROM v$dataguard_stats WHERE name LIKE '%lag%';
NAME VALUE UNIT TIME_COMPUTED DATUM_TIME

transport lag +00 00:00:00 day(2) to second(0) interval 05/09/2020 17:01:52 05/09/2020 17:01:51
apply lag +00 00:00:00 day(2) to second(0) interval 05/09/2020 17:01:52 05/09/2020 17:01:51

2 rows selected.

SQL> col client_pid for a10
SQL> SELECT inst_id, thread#, process, pid, status, client_process, client_pid,
2 sequence#, block#, active_agents, known_agents FROM gv$managed_standby ORDER BY thread#, pid;

INST_ID THREAD# PROCESS PID STATUS CLIENT_P CLIENT_PID SEQUENCE# BLOCK# ACTIVE_AGENTS KNOWN_AGENTS

   1        0 ARCH             2706 CONNECTED    ARCH     2706               0          0             0            0
   1        0 RFS              2840 IDLE         UNKNOWN  2830               0          0             0            0
   1        0 RFS              2858 IDLE         UNKNOWN  2824               0          0             0            0
   1        0 RFS              2869 IDLE         ARCH     2828               0          0             0            0
   1        1 ARCH             2702 CLOSING      ARCH     2702              64          1             0            0
   1        1 ARCH             2704 CLOSING      ARCH     2704              65          1             0            0
   1        1 ARCH             2708 CLOSING      ARCH     2708              63          1             0            0
   1        1 RFS              2860 IDLE         LGWR     2832              66      10444             0            0
   1        1 MRP0             3386 APPLYING_LOG N/A      N/A               66      10444             0            0

9 rows selected.

同步状态已经正常

问题原因
物理standby端的standby_file_management参数设置错误，导致在primary端添加数据文件或者创建表空间的时候，不能正常传输到备库。

解决方案：
alter database create datafile '/u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00006' as new;
alter system set standby_file_management=auto;
recover managed standby database using current logfile disconnect;

扩展
多数情况下，primary数据库的修改会随着redo数据传播到物理standby数据库端并被应用，不需要在物理standby端做额外的操作，不过根据实际配置的不同，也会有例外，
有些操作不是没有被传播到standby端，而是传播过去了，但不能正确执行，其中最常见的就是对表空间和日志文件的管理操作。

创建表空间或数据文件
初始化参数standby_file_management用来控制是否自动将primary数据库增加表空间或数据文件的改动，传播到物理standby数据库。该参数有两个值：
1）auto：如果该参数值设置为AUTO，则primary数据库执行的表空间创建操作也会传播到物理standby数据库上执行。
2）MANUAL:如果设置为MANUAL或者未设置任何值（默认值是MANUAL），需要手工复制新创建的数据文件到物理standby服务器。

standby_file_management参数特指primary数据库端的表空间或数据文件创建，如果数据文件是从其他数据库复制而来（比如通过TTS传输表空间），则不管standby_file_management参数值如何设置。
都必须手工复制到standby数据库，并重建物理standby数据库的控制文件。

备库standby_file_management参数为auto
SQL> show parameter standby;

NAME TYPE VALUE

standby_archive_dest string ?/dbs/arch
standby_file_management string AUTO
SQL>

主库创建表空间
SQL> create tablespace test datafile size 20M;

Tablespace created.
查看数据文件路径
SQL> col tbs for a10
SQL> col files for a100
SQL> select a.ts#,a.name tbs,b.name files from ts$ a,v$datafile b where a.ts#=b.ts#;

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_system_gxd20h14_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_sysaux_gxd20k1y_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_undotbs1_gxd20lnp_.dbf
 4 USERS      /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_users_gxd20pxk_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_gxdcfr5s_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_hcdsblpt_.dbf
 6 TEST       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_test_hcf6r72x_.dbf    ---《《《新创建的

7 rows selected.
备库查看数据文件路径：
SQL> col tbs for a10
SQL> col files for a100
SQL> set lines 200

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_system_05uvp9bj_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_sysaux_06uvp9bj_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_undotbs1_07uvp9bj_.dbf
 4 USERS      /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_users_08uvp9c4_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_04uvp9bj_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_hcdw395t_.dbf
 6 TEST       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_test_hcf6r765_.dbf  ---《《显示正常

7 rows selected.

SQL>

备库standby_file_management参数为manual
备库修改参数为manual
SQL> alter system set standby_file_management=manual;

System altered.
主库添加数据文件并查看数据文件路径：

SQL> alter tablespace test add datafile size 10m;

Tablespace altered.

SQL> select a.ts#,a.name tbs,b.name files from ts$ a,v$datafile b where a.ts#=b.ts#;

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_system_gxd20h14_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_sysaux_gxd20k1y_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_undotbs1_gxd20lnp_.dbf
 4 USERS      /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_users_gxd20pxk_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_gxdcfr5s_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_hcdsblpt_.dbf
 6 TEST       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_test_hcf6r72x_.dbf
 6 TEST       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_test_hcf76186_.dbf

8 rows selected.

备库查看：
SQL> select a.ts#,a.name tbs,b.name files from ts$ a,v$datafile b where a.ts#=b.ts#;

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_system_05uvp9bj_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_sysaux_06uvp9bj_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_undotbs1_07uvp9bj_.dbf
 4 USERS      /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_users_08uvp9c4_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_04uvp9bj_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_hcdw395t_.dbf
 6 TEST       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_test_hcf6r765_.dbf
 6 TEST       /u01/app/oracle/product/11.2.0/dbhome_1/dbs/UNNAMED00008         ----《《《文件命名异常，有unname字样，路径在$ORACLE_HOME/dbs存放

8 rows selected.
alert日志：

Sat May 09 19:59:03 2020
Successfully added datafile 7 to media recovery
Datafile #7: ‘/u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_test_hcf6r765_.dbf’
Sat May 09 20:03:32 2020
WARNING: Heavy swapping observed on system in last 5 mins.
pct of memory swapped in [56.13%] pct of memory swapped out [37.44%].
Please make sure there is no memory pressure and the SGA and PGA
are configured correctly. Look at DBRM trace file for more details.
Sat May 09 20:06:09 2020
ALTER SYSTEM SET standby_file_management=‘MANUAL’ SCOPE=BOTH;
Sat May 09 20:06:25 2020
File #8 added to control file as ‘UNNAMED00008’ because
the parameter STANDBY_FILE_MANAGEMENT is set to MANUAL
The file should be manually created to continue.
MRP0: Background Media Recovery terminated with error 1274
Errors in file /u01/app/oracle/diag/rdbms/zhuodg/zhuodg/trace/zhuodg_mrp0_3386.trc:
ORA-01274: cannot add datafile ‘/u01/app/oracle/oradata/ZHUO/datafile/o1_mf_test_hcf76186_.dbf’ - file could not be created
Managed Standby Recovery not using Real Time Apply
Recovery interrupted!
Recovered data files to a consistent state at change 808919
MRP0: Background Media Recovery process shutdown (zhuodg) ----《《《MRP进程异常停止

发现和上述案例的现象一致，就是因为参数为manual的原因。
处理方式就是上述方法，不在叙述。

备库standby_file_management参数为auto
主库删除表空间
SQL> drop tablespace test including contents and datafiles;

Tablespace dropped.

SQL> select a.ts#,a.name tbs,b.name files from ts$ a,v$datafile b where a.ts#=b.ts#;

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_system_gxd20h14_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_sysaux_gxd20k1y_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_undotbs1_gxd20lnp_.dbf
 4 USERS      /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_users_gxd20pxk_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_gxdcfr5s_.dbf
 5 ZHUO       /u01/app/oracle/oradata/ZHUO/datafile/o1_mf_zhuo_hcdsblpt_.dbf

6 rows selected.
备库查看
SQL> select a.ts#,a.name tbs,b.name files from ts$ a,v$datafile b where a.ts#=b.ts#;

TS# TBS FILES
1

 0 SYSTEM     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_system_05uvp9bj_.dbf
 1 SYSAUX     /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_sysaux_06uvp9bj_.dbf
 2 UNDOTBS1   /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_undotbs1_07uvp9bj_.dbf
 4 USERS      /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_users_08uvp9c4_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_04uvp9bj_.dbf
 5 ZHUO       /u01/app/oracle/oradata/zhuodg/ZHUODG/datafile/o1_mf_zhuo_hcdw395t_.dbf

6 rows selected.

正常删除，同步到了备库

结论
综上，初始化参数standby_file_management为manual的时候，对于表空间和数据文件的增加和删除操作必须手工处理。
为auto的时候，对于表空间和数据文件的操作无需dba手工干预，物理standby能自动进行处理。

Plan hash value: 2490995645
 
-------------------------------------------------------------------------------------------
| Id  | Operation           | Name                | Rows  | Bytes | Cost (%CPU)| Time     |
-------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT    |                     |       |       |   465K(100)|          |
|   1 |  SORT AGGREGATE     |                     |     1 |    38 |            |          |
|   2 |   NESTED LOOPS OUTER|                     | 51621 |  1915K|   465K  (1)| 01:33:05 |
|*  3 |    TABLE ACCESS FULL| BYDD9WMS_PRINT_DATA | 51621 |  1058K|   465K  (1)| 01:33:02 |
|*  4 |    INDEX UNIQUE SCAN| UK_PO_DATA          |     1 |    17 |     1   (0)| 00:00:01 |
-------------------------------------------------------------------------------------------
 
Predicate Information (identified by operation id):
---------------------------------------------------
 
   3 - filter(("AA"."CODE_DATE" IS NULL AND "AA"."DATECODE" IS NOT NULL AND 
              NVL("AA"."UPDATE_NAME",'X')<>'UPCD2303210001'))
   4 - access("PD"."PO"="AA"."PO" AND "PD"."LINE"="AA"."LINE")
 
Note
-----
   - cardinality feedback used for this statement

查看执行计划发现：cardinality feedback used for this statement
11g之后的新特性， 如果cbo发现e_row和a_row的差距过多， 认为执行计划不准确，则重新生成执行计划。
检查发现对应的表的统计信息，上次分析是3年前，说明上次导入之后，连带统计信息也被导入，到目前为止，更新的次数不超过原来的10%，所以一直未更新。
所以可以在业务空闲时间重新收集统计信息，但是目前等待事件为全表扫描的io，cpu资源空闲较多，所以重新收集统计信息也无用，需要sql优化，减少全表扫描次数。才会有明显改善，检查业务逻辑发现，该动作是对一张800万的表进行全表扫描，可以预见，未来随着业务的增长，该语句的性能会越来越低，最终将无法执行下去，可能的措施是归档历史数据，或者改变表结构为分区表。减少查询扫描的范围。

declare   
    ---define Cursor 
    cur_policy_sql  VARCHAR2(32767) := 'select * from  (select object_id,object_name 
     from t1111 order by object_id )  
     where rownum < 10' ;
    Cursor cur_policy is  
     select * from  (select object_id,object_name 
     from t1111 order by object_id )  
     where rownum < 10 ;  
    curPolicyInfo cur_policy%rowtype;---定义游标变量  
Begin  
   Dbms_Output.put_line(cur_policy_sql);
   open cur_policy; ---open cursor  
   Loop   
     --deal with extraction data from DB  
     Fetch cur_policy into curPolicyInfo;  
     Exit when cur_policy%notfound;  
        
     Dbms_Output.put_line(curPolicyInfo.object_id); 
     Dbms_Output.put_line(cur_policy%ROWCOUNT); 
   end loop;  
   
   Exception   
     when others then  
         close cur_policy;  
         Dbms_Output.put_line(Sqlerrm);  
           
   if cur_policy%isopen then    
    --close cursor   
      close cur_policy;  
   end if;  
end;