对PDB执行按时间点恢复类似于执行数据库按时间点恢复。当对一个或多个PDB恢复到指定时间点时，CDB中的其它PDB不受影响。在恢复之后，PDB原来的保留的旧备份仍然有效可以在出现介质恢复时使用,不需要创建新的备份。当对使用共享UNDO的CDB中的一个或多个PDB执行数据库按时间点恢复时,对于包含被恢复PDB的CDB的root与CDB seed(PDB$SEES)需要有备份。从Oracle 12.2开始，如果compatible参数被设置为12.2，那么可以跨PDB闪回操作或PDB按时间点恢复来对CDB执行闪回数据库操作。在DG环境中，对于备库将跟随主库PDB会被恢复到指定的时间点，你可以闪回整个备库，恢复PDB或对PDB执行闪回。
对PDB执行按时间点恢复的操作步骤如下:
1.登录数据库记录当前SCN号，然后将表t1中的数据删除。
SQL> conn jy/jy@jypdb
Connected.
SQL> SELECT CURRENT_SCN FROM V$DATABASE;
CURRENT_SCN
-----------
6255735
SQL> alter session set nls_date_format='yyyy-mm-dd hh24:mi:ss';
Session altered.
SQL> select sysdate from dual;
SYSDATE
-------------------
2017-12-20 16:52:31
SQL> select count(*) from t1;
COUNT(*)
----------
39
SQL> truncate table t1;
Table truncated.
SQL> select count(*) from t1;
COUNT(*)
----------
0
2.如果使用时间表达式来代替目标SCN，那么在调用RMAN之前设置时间格式环境变量
[oracle@jytest1 ~]$ export NLS_DATE_FORMAT='yyyy-mm-dd hh24:mi:ss'
3.使用RMAN连接到root容器
[oracle@jytest1 ~]$ rman target/ catalog rco/xxzx7817600@jypdb_173 Recovery Manager: Release 12.2.0.1.0 - Production on Wed Dec 20 16:53:26 2017 Copyright (c) 1982, 2017, Oracle and/or its affiliates. All rights reserved. connected to target database: JY (DBID=979425723) connected to recovery catalog database
4.将要执行恢复的PDB关闭，其它的PDB与CDB仍然处于open状态
RMAN> alter pluggable database jypdb close immediate; starting full resync of recovery catalog full resync complete Statement processed starting full resync of recovery catalog full resync complete
5.使用RUN块来执行以下操作
a.对于数据库按时间点鶋，使用set until来指定恢复的目标时间，scn或日志序列号，或者使用set to来指定还原点。如果指定时间那么使用环境变量nls_lang与nls_date_format中所指定的日期格式。
b.如果RMAN没有配置自动通道，那么需要手动分配磁盘与磁带通道。
c.还原与恢复CDB
下面的命令将PDB(jypdb)恢复到SCN=6255735所在的状态
RMAN> run
2> {
3> set until scn 6255735;
4> restore pluggable database jypdb;
5> recover pluggable database jypdb;
6> }
executing command: SET until clause
Starting restore at 2017-12-20 17:00:38
using channel ORA_DISK_1
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_DISK_1: restoring datafile 00010 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/system.271.962209649
channel ORA_DISK_1: restoring datafile 00011 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/sysaux.316.962209649
channel ORA_DISK_1: restoring datafile 00012 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/undotbs1.264.962209649
channel ORA_DISK_1: restoring datafile 00013 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/undo_2.268.962209649
channel ORA_DISK_1: restoring datafile 00014 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/users.278.962209649
channel ORA_DISK_1: restoring datafile 00015 to +DATA/JY/5F9AC6865E87549FE053AB828A0ADE94/DATAFILE/test.275.962210609
channel ORA_DISK_1: reading from backup piece +TEST/rman_backup/jy_979425723_962563516_11slv3ds_1_1
channel ORA_DISK_1: piece handle=+TEST/rman_backup/jy_979425723_962563516_11slv3ds_1_1 tag=TAG20171212T184328
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:35
Finished restore at 2017-12-20 17:01:15
Starting recover at 2017-12-20 17:01:16
current log archived
using channel ORA_DISK_1
starting media recovery
archived log for thread 1 with sequence 38 is already on disk as file +TEST/arch/1_38_961976319.dbf
archived log for thread 1 with sequence 39 is already on disk as file +TEST/arch/1_39_961976319.dbf
archived log for thread 1 with sequence 40 is already on disk as file +TEST/arch/1_40_961976319.dbf
archived log for thread 1 with sequence 41 is already on disk as file +TEST/arch/1_41_961976319.dbf
archived log for thread 1 with sequence 42 is already on disk as file +TEST/arch/1_42_961976319.dbf
archived log for thread 1 with sequence 43 is already on disk as file +TEST/arch/1_43_961976319.dbf
archived log for thread 1 with sequence 44 is already on disk as file +TEST/arch/1_44_961976319.dbf
archived log for thread 1 with sequence 45 is already on disk as file +TEST/arch/1_45_961976319.dbf
archived log for thread 1 with sequence 46 is already on disk as file +TEST/arch/1_46_961976319.dbf
archived log for thread 1 with sequence 47 is already on disk as file +TEST/arch/1_47_961976319.dbf
archived log for thread 1 with sequence 48 is already on disk as file +TEST/arch/1_48_961976319.dbf
archived log for thread 1 with sequence 49 is already on disk as file +TEST/arch/1_49_961976319.dbf
archived log for thread 1 with sequence 50 is already on disk as file +TEST/arch/1_50_961976319.dbf
archived log for thread 1 with sequence 51 is already on disk as file +TEST/arch/1_51_961976319.dbf
archived log for thread 1 with sequence 52 is already on disk as file +TEST/arch/1_52_961976319.dbf
archived log for thread 1 with sequence 53 is already on disk as file +TEST/arch/1_53_961976319.dbf
archived log for thread 1 with sequence 54 is already on disk as file +TEST/arch/1_54_961976319.dbf
archived log for thread 1 with sequence 55 is already on disk as file +TEST/arch/1_55_961976319.dbf
archived log for thread 1 with sequence 56 is already on disk as file +TEST/arch/1_56_961976319.dbf
archived log for thread 1 with sequence 57 is already on disk as file +TEST/arch/1_57_961976319.dbf
archived log for thread 2 with sequence 32 is already on disk as file +TEST/arch/2_32_961976319.dbf
archived log for thread 2 with sequence 33 is already on disk as file +TEST/arch/2_33_961976319.dbf
archived log for thread 2 with sequence 34 is already on disk as file +TEST/arch/2_34_961976319.dbf
archived log for thread 2 with sequence 35 is already on disk as file +TEST/arch/2_35_961976319.dbf
archived log for thread 2 with sequence 36 is already on disk as file +TEST/arch/2_36_961976319.dbf
archived log for thread 2 with sequence 37 is already on disk as file +TEST/arch/2_37_961976319.dbf
archived log for thread 2 with sequence 38 is already on disk as file +TEST/arch/2_38_961976319.dbf
archived log for thread 2 with sequence 39 is already on disk as file +TEST/arch/2_39_961976319.dbf
archived log for thread 2 with sequence 40 is already on disk as file +TEST/arch/2_40_961976319.dbf
archived log for thread 2 with sequence 41 is already on disk as file +TEST/arch/2_41_961976319.dbf
archived log for thread 2 with sequence 42 is already on disk as file +TEST/arch/2_42_961976319.dbf
archived log for thread 2 with sequence 43 is already on disk as file +TEST/arch/2_43_961976319.dbf
archived log for thread 2 with sequence 44 is already on disk as file +TEST/arch/2_44_961976319.dbf
archived log for thread 2 with sequence 45 is already on disk as file +TEST/arch/2_45_961976319.dbf
archived log for thread 2 with sequence 46 is already on disk as file +TEST/arch/2_46_961976319.dbf
archived log for thread 2 with sequence 47 is already on disk as file +TEST/arch/2_47_961976319.dbf
archived log for thread 2 with sequence 48 is already on disk as file +TEST/arch/2_48_961976319.dbf
archived log for thread 2 with sequence 49 is already on disk as file +TEST/arch/2_49_961976319.dbf
archived log for thread 2 with sequence 50 is already on disk as file +TEST/arch/2_50_961976319.dbf
archived log for thread 2 with sequence 51 is already on disk as file +TEST/arch/2_51_961976319.dbf
archived log for thread 2 with sequence 52 is already on disk as file +DATA/JY/ONLINELOG/group_4.262.961976705
archived log for thread 2 with sequence 53 is already on disk as file +DATA/JY/ONLINELOG/group_3.263.961976697
media recovery complete, elapsed time: 00:04:03
Finished recover at 2017-12-20 17:05:30
starting full resync of recovery catalog
full resync complete
6. 以读写方式打开PDB，放弃目标SCN之后的所有改变,执行以下命令
RMAN> alter pluggable database jypdb open resetlogs;
Statement processed
starting full resync of recovery catalog
full resync complete
SQL> conn jy/jy@jypdb
Connected.
SQL> select count(*) from t1;
COUNT(*)
----------
39