oracle active dataguard 环境下 primary 创建 pdb 导致 standby 无法正确创建并取消应用

os: centos 7.4
db: oracle 12.1.0.2

rac + adg + rac

primary 上创建 pdb

如下:

CREATE PLUGGABLE DATABASE WALTPDB1 FROM PAYPDB1
STORAGE UNLIMITED
; 

standby 上报错

如下:

Recovery created pluggable database WALTPDB1
Recovery copied files for tablespace SYSTEM
Recovery successfully copied file +DG_DATA01/ORCLS1/88FC523A843D42F5E0538847650A98D0/DATAFILE/system.283.1008426641 from +DG_DATA01/ORCLS1/884C18C477EC5A90E0538847650A8293/DATAFILE/system.277.1007757591
MRP0: Background Media Recovery terminated with error 1274
Thu May 16 14:30:41 2019
Errors in file /u01/app/oracle/diag/rdbms/orcls1/orcls11/trace/orcls11_pr00_29395.trc:
ORA-01274: cannot add data file that was originally created as '+DG_DATA01/ORCLP/88FC523A843D42F5E0538847650A98D0/DATAFILE/system.306.1008426623'
ORA-19729: File 33 is not the initial version of the plugged in datafile
Thu May 16 14:30:41 2019
Managed Standby Recovery not using Real Time Apply
Thu May 16 14:30:41 2019
Recovery interrupted!

$ tail -n 200 /u01/app/oracle/diag/rdbms/orcls1/orcls11/trace/orcls11_pr00_29395.trc

*** 2019-05-16 14:30:18.057
MRP: Prodding archiver at standby for thread 2 seq 1360

*** 2019-05-16 14:30:18.200
Media Recovery of Online Log [Thread=1, Seq=1433]

*** 2019-05-16 14:30:18.201
Recovery of Online Redo Log: Thread 1 Group 10 Seq 1433 Reading mem 0

*** 2019-05-16 14:30:19.006
Media Recovery of Online Log [Thread=2, Seq=1361]

*** 2019-05-16 14:30:19.007
Recovery of Online Redo Log: Thread 2 Group 15 Seq 1361 Reading mem 0

*** 2019-05-16 14:30:41.570
kcf_record_fn: control file transaction invalid
DDE rules only execution for: ORA 1274
----- START Event Driven Actions Dump ----
---- END Event Driven Actions Dump ----
----- START DDE Actions Dump -----
Executing SYNC actions
----- START DDE Action: 'DB_STRUCTURE_INTEGRITY_CHECK' (Async) -----
Successfully dispatched
----- END DDE Action: 'DB_STRUCTURE_INTEGRITY_CHECK' (SUCCESS, 0 csec) -----
Executing ASYNC actions
----- END DDE Actions Dump (total 0 csec) -----
*** 2019-05-16 14:30:41.602436 4929 krsh.c
MRP0: Background Media Recovery terminated with error 1274
ORA-01274: cannot add data file that was originally created as '+DG_DATA01/ORCLP/88FC523A843D42F5E0538847650A98D0/DATAFILE/system.306.1008426623'
ORA-19729: File 33 is not the initial version of the plugged in datafile
MRP: Prodding archiver at standby for thread 1 seq 1433
==== Redo read statistics for thread 1 ====
Total physical reads (from disk and memory): 1532779Kb
-- Redo read_disk statistics --
Read rate (ASYNC): 1532779Kb in 189257.80s => 0.01 Mb/sec
Total redo bytes: 1532779Kb Longest record: 54Kb, moves: 538/1282719 moved: 3Mb (0%)
Longest LWN: 1147Kb, reads: 231099 
Last redo scn: 0x0000.00771c43 (7806019)
Change vector header moves = 209488/2577731 (8%)
----------------------------------------------

*** 2019-05-16 14:30:41.619
Media Recovery drop redo thread 1
MRP: Prodding archiver at standby for thread 2 seq 1361
==== Redo read statistics for thread 2 ====
Total physical reads (from disk and memory): 1554017Kb
-- Redo read_disk statistics --
Read rate (ASYNC): 1554017Kb in 189257.82s => 0.01 Mb/sec
Total redo bytes: 1554017Kb Longest record: 58Kb, moves: 533/1394993 moved: 3Mb (0%)
Longest LWN: 1199Kb, reads: 224256 
Last redo scn: 0x0000.00771c39 (7806009)
Change vector header moves = 232100/2825619 (8%)
----------------------------------------------

*** 2019-05-16 14:30:41.628
Media Recovery drop redo thread 2
KCBR: Redo cache copies/changes = 11246/11246
Wait to push change maps to slaves = 1149s

*** 2019-05-16 14:30:41.785
Completed Media Recovery
Unloading kcnibr
Checking to start in-flux buffer recovery from SCN 0.7805802 to SCN  (non-inclusive) 0.7806019
Influx recovery found in-flux buffers

*** 2019-05-16 14:30:41.834
Influx Media Recovery add redo thread 1 seq 1433

*** 2019-05-16 14:30:41.834
Influx Media Recovery add redo thread 2 seq 1360
Managed Standby Recovery: Standby online log for thr 2  seq 1360 not found. Looking whether archived..
Looking for archived log thr 2 seq 1360 recovery branch id 1007668940 that contains SCN 7805802
Checking whether scn 7805802 in los 7803861 and nxs 7805805
Managed Standby Recovery: Opening archived log +DG_FRA/ORCLS1/ARCHIVELOG/2019_05_16/thread_2_seq_1360.1289.1008426619 during invocation of recoverable recovery
MRP: Prodding archiver at standby for thread 2 seq 1361
MRP: Prodding archiver at standby for thread 1 seq 1433
Managed Recovery: Not Active posted.
Recovered data files to a consistent state at change 7806019

*** 2019-05-16 14:30:42.077
* kju_rora: called with opts x2, roram 1, requester 32767, adg enabled? 1 
*           adg roram 32767, adg requester 32767 
* readable standby - end of redo apply indicated 
* kju_rora TRACEADG: rora_requester = myinst = 1 

*** 2019-05-16 14:30:43.085
ORA-01274: cannot add data file that was originally created as '+DG_DATA01/ORCLP/88FC523A843D42F5E0538847650A98D0/DATAFILE/system.306.1008426623'
ORA-19729: File 33 is not the initial version of the plugged in datafile

在 mos 上搜索,越看越感觉是个bug.临时的处理方式是铲掉 standby 重做,又不麻烦.
参考:
<<MRP Fails with ORA-1274 while Creating PDB on Standby (文档 ID 2423108.1)>>
https://support.oracle.com/epmos/faces/SearchDocDisplay?_adf.ctrl-state=10yjbgefsv_9&_afrLoop=440705193547261

<<Bug 21466090 - MRP fails with ORA-01274 when PRIMARY uses ASM and STANDBY does not after adding a PDB (文档 ID 21466090.8)>>
https://support.oracle.com/epmos/faces/SearchDocDisplay?_adf.ctrl-state=10yjbgefsv_9&_afrLoop=440705193547261

<<Bug 28174951 - CDB: ORA-19729 Datafile Not Created Automatically On ADG (文档 ID 28174951.8)>>
https://support.oracle.com/epmos/faces/SearchDocDisplay?_adf.ctrl-state=10yjbgefsv_9&_afrLoop=440705193547261