ogg重新初始化

参考该文档：http://blog.mchz.com.cn/?p=9927

在target端因为磁盘空间满replicat ABANDING，清理磁盘空间后，再次start rep1 ，报错：找不到上一个trace文件，此时，“杀招”出现了。

alter replicat rep1 extseqno  0 extrba 0;

然后在检查点检查中一直：

Current Checkpoint Detail:

Read Checkpoint #1

  GGS Log Trail

  Startup Checkpoint (starting position in the data source):
    Sequence #: 0
    RBA: 0
    Timestamp: Not Available
    Extract Trail: /oracle/oggrep/dirdat/lt

  Current Checkpoint (position of last record read in the data source):
    Sequence #: 0
    RBA: 0
    Timestamp: Not Available
    Extract Trail: /oracle/oggrep/dirdat/lt

GGSCI (xxxdb1) 34> info rep1,detail


REPLICAT   REP1      Last Started 2015-01-21 10:49   Status RUNNING
Checkpoint Lag       00:00:00 (updated 00:00:08 ago)
Log Read Checkpoint  File /oracle/oggrep/dirdat/lt000000
                     First Record  RBA 0


  Extract Source                          Begin             End             


  /oracle/oggrep/dirdat/lt000000          * Initialized *   First Record    
  /oracle/oggrep/dirdat/lt000000          * Initialized *   First Record    
  /oracle/oggrep/dirdat/lt000000          * Initialized *   First Record    
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51
  /oracle/oggrep/dirdat/lt001385          2015-01-11 23:51  2015-01-11 23:51

主要是trail文件没有了，不然可以再次alter replicat rep1 extseqno  XXX  extrba XXX; 试试的！

没办法，重新初始化了

ogg重新初始化
一、在源端导出数据
SQL>select to_char(current_scn) from v$database; 查询当前scn
2719112586
导出数据
[oracle@yyghdb odc]$expdp odc/odc directory=odc_dir schemas=BOOKPLAT dumpfile=BOOKPLAT2.dmp flashback_scn=2719112586 logfile=BOOKPL[oracle@yyghdb odc]$ expdp odc/odc directory=odc_dir
 
二、目标端导入数据
drop user BOOKPLAT cascade
删除原有的同步用户
impdp odc/odc@zjhorcl1 directory=ODC_DIR dumpfile=bookplat2.dmp schemas=BOOKPLAT logfile=bookplat2.log
重新导入同步用户
Select trigger_name from dba_triggers where owner='BOOKPLAT' and status='ENABLED';
TRIG_DUP
查询同步用户触发器
alter trigger BOOKPLAT.TRIG_DUP disable;
禁用该触发器
移除JOB：
select job,log_user,schema_user from dba_jobs;
exec dbms_ijob.remove(<that job id>);
select * from dba_scheduler_jobs;
禁用：
DBMS_SCHEDULER.DISABLE('job1, job2, job3, sys.jobclass1, sys.jobclass2');
 
三、设置复制进程位置，并启动
查看dirdat下trail文件日期，找出数据导出时间前的trail文件
ls -rtl
total 3743578
-rw-rw-rw-   1 oracle     oinstall   99999967 Apr 28 03:01 ws000686
-rw-rw-rw-   1 oracle     oinstall   99999725 Apr 28 04:37 ws000687
-rw-rw-rw-   1 oracle     oinstall   99999993 Apr 28 15:01 ws000688
设置复制进程位置，让复制进程从000687开始复制 
/*在该步骤时要尽量选取靠近SCN时间的trail文件*/


alter ws_repc extseqno  000687 extrba 0
alter ws_rep4 extseqno  000687 extrba 0
/*===extseqno  指的就是trail文件的文件=====*/


用atcsn启动复制进程
start ws_rep4 atcsn 2719112586
start ws_repc atcsn 2719112586
 
查看进程状态，同步完成
GGSCI (ehrdb1) 1> info all
 
Program     Status      Group       Lag at Chkpt  Time Since Chkpt
 
MANAGER   RUNNING                                              
REPLICAT    RUNNING     WS_REP4     00:00:00      30:33:17   
REPLICAT    RUNNING     WS_REPC     00:00:00      03:15:02    


在插入的时候会报以下错误：重复插入
2015-01-21 13:51:15  WARNING OGG-00869  Oracle GoldenGate Delivery for Oracle, rep1.prm:  OCI Error ORA-00001: unique constraint (AAAA.PK_WS_BUSINESS_LOG) violated (status = 1). INSERT INTO "AAAA"."TBL_WS_BUSINESS_LOG" ("PRIMARY_KEY","TRANS_ID","TRANS_DATE","TRANS_TIME","ORDER_STATE","EXTERNAL_KEY","BATCH_NO","SEQ_NO","BUSINESS_TYPE","PROV_CODE","CHANNEL","NODE_NO","NODE_NAME","OPE_NO","OPE_NAME","CARD_TYPE","CARD_NO","CARD_BIN","TXN_SCRAPE_PIN","TRANS_PIN","CVV","CVV2","SETTLE_DT","PAY_TYPE","PAYABLE_AMOUNT","REALITY_AMOUNT","AMT","OIL_CARD_NO","OIL_TER_SEQ_NO","COMMODITY_TP","COMMODITY_SMALL_TP","COMMODITY_ID","RESERVE","PRODUCTID","TOTALQUANTITY","REALITYAMOUNT","KEYID","MESSAGEKEY","PINKEY","MESSAGEKEYCHECKVALUE","PINKEYCHECKVALUE","REMARKS","RESP_CODE","RE_CEN_SEQ_NO","BALANCE","VAILDDATE","CREATE_TIME","MODIFY_TIME","STATUS","TRANS_SRC") VALUES (:a0,:a1,:a2,:a3,:a4,:a5,:a6,:a7,:a8,:a9,:a10,:a11,:a12,:a13,:a14,:a15,:a16,:a17,:a18,:a19,:a20,:a21,:a22,:a23,:a24,:a25,:a26,:a27,:a28,:a29,:a30,:a31,:a32,:a33,:a34,:a35,:a36,:a37,:a38,:a39,:a40,:a41,:a42,:a43,:a44,:a45,:a46,:a47,:a48,:a49).


解决方法：


1 . 尝试加入 handlecollisions 参数解决该问题


2.  尝试手动修复 target 上的数据


3. 尝试加入REPERROR  参数 ignore这个错误


来自Maclean Liu 刘相兵

我直接使用handlecollisions，在追踪到最新的trail文件之前停止ogg，然后删除该参数。重启ogg

其中对checkpointtable的理解存在误区，导致自动purge时失败，手工删除，没文化太可怕！！！