对丢弃已久的goldengate环境重新配置，使其重新开始跑起来
环境是一个主机上的两个库，都是单机，所以也就没配pump进程了，trail file都是在一个文件夹下的，extract写trail file到这，然后replicate在这读取trail file。
先来看下大致的环境情况：VDBSL3--->VDBCB3
源库：-----------VDBSL3（dom schema ）以下几张表要去用到ogg，
Table dom.YFS_ITEM;
Table dom.YFS_ITEM_ALIAS;
Table dom.YFS_INVENTORY_ITEM;
Table dom.YFS_INVENTORY_ALERTS;
Table dom.YFS_INVENTORY_SUPPLY;
Table dom.EXTN_STORE_REGION_CACHE;
目标库：----------VDBCB3(ce_hub schema):
MAP dom.YFS_ITEM, TARGET ce_hub.YFS_ITEM;
MAP dom.YFS_ITEM_ALIAS, TARGET ce_hub.YFS_ITEM_ALIAS;
MAP dom.YFS_INVENTORY_ITEM, TARGET ce_hub.YFS_INVENTORY_ITEM;
MAP dom.YFS_INVENTORY_ALERTS, TARGET ce_hub.YFS_INVENTORY_ALERTS;
MAP dom.YFS_INVENTORY_SUPPLY, TARGET ce_hub.YFS_INVENTORY_SUPPLY;
MAP dom.EXTN_STORE_REGION_CACHE, TARGET ce_hub.EXTN_STORE_REGION_CACHE;
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com) 1>
Program Status Group Lag at Chkpt Time Since Chkpt
MANAGER RUNNING
EXTRACT ABENDED E_ICUB3 00:00:00 2122:38:24
REPLICAT ABENDED R_ICUB3 00:00:07 2155:36:01 可见废弃很久了
===================================================================================
现在开始重新配置了
把extract和replicate进程delete掉，清下相关的rpt file,dsc file,trail file
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com) 5> dblogin USERID ggadmin, PASSWORD xxx
Successfully logged into database.
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com as ggadmin@VDBSL3) 7> delete extract E_ICUB3
Deleted EXTRACT E_ICUB3.
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com as ggadmin@VDBCB3) 3> delete replicat R_ICUB3
Deleted REPLICAT R_ICUB3.
Delete掉之后，清除一些文件，主要是dirrpt下的rpt和dsc file和dirdat下的trail file，当然还可以去清下gglog dump file
做完这些基本上就可以了，然后replicate那边去对应的数据库里看看GGADMIN.GG_CHECKPOINT这张表是否还有这个replicate的记录，理论上是应该没有的，如果有的话，那么把记录delete掉
=================================================================================================
接下来就是做下exp和impdp的操作，在做这之前，因为这个是delphix database，定期会被刷新的，先来检查一些必要的参数
SELECT supplemental_log_data_min, force_logging FROM v$database;
ALTER DATABASE FORCE LOGGING;
alter database add supplemental log data;
alter system switch logfile;
alter system set enable_goldengate_replication=TRUE scope=both； （this parameter used in 11.2.0.4）
顺便建下directory
create directory DATA_PUMP_DIR1 as '/u01/home/oracle/datapump/vdbsl3';
然后的话准备添加extract进程并起起来，如下：
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com) 2> obey /u01/app/ggadmin/product/12.1.2.1.0/dirprm/e_icub3.oby
也可以不执行obey这个文件，手动执行这几条命令
more /u01/app/ggadmin/product/12.1.2.1.0/dirprm/e_icub3.oby
-- Create the change data capture extract, and local extract trail of 1024mb
dblogin userid ggadmin, password xxxxx
add extract e_icub3, tranlog, begin now
add rmttrail /u01/app/ggadmin/product/12.1.2.1.0/dirdat/extract/VDBCB3/e1, extract e_icub3 , megabytes 1024
然后的话就准备起起来，但是要注意的是数据库当前不要有long running transaction在跑，可以用如下命令确认当前的数据库事务情况：
select min(start_time) from gv$transaction;
select count(*) from gv$transaction;
select t.start_time, t.xidusn||'.'||t.xidslot||'.'||t.xidsqn xid, s.status, s.sid,s.serial#,s.username,s.status,s.schemaname, decode(s.sql_id,null,s.prev_sql_id) sqlid, decode(s.sql_child_number,null,s.prev_child_number) child from v$transaction t, v$session s where s.saddr = t.ses_addr order by t.start_time ;
要是确认好了，没有长事务在跑的话，就把extract起起来吧，然后记一下scn点，待会expdp根据这个scn点导出，start replicat也是根据这个时间点
源库执行scn查询：
select to_char(current_scn) from gv$database;
OR
select
to_char(dbms_flashback.get_system_change_number) from dual;
主要是要保证在scn点抓取的时候没有长shiwu6在跑，那种典型的时间短的OLTP事务可以忽略，因为extract开始了就一直抓取数据库的DML变化，然后目标库的数据初始化时以这个scn开始的，所以只要在抓scn的时候没有长事务，那么两边是可以保持一致的。如果有长事务，这个可能就会在目标库里miss掉！
在原库vdbsl3上做expdp
expdp \'/ as sysdba\'
parfile=gg_reint_1213.par
oracle@delphix-vdb-n-3:VDBSL3:/u01/home/oracle/datapump/vdbsl3
$ more gg_reint_1213.par
directory=DATA_PUMP_DIR1
dumpfile=gg_reint_1213_%U.dmp
logfile=ogg_1213.log
cluster=N
parallel=8
Compression=ALL
tables=dom.YFS_ITEM,dom.YFS_ITEM_ALIAS,dom.YFS_INVENTORY_ITEM,dom.YFS_INVENTORY_ALERTS,dom.YFS_INVENTORY_SUPPLY,dom.EXTN_STORE_REGION_CACHE
flashback_scn=14193095280953
（可以加上grants=n）
导出之后，便可以做导入了
impdp \'/ as
sysdba\' parfile=ogg_1213_imp.par
remap_schema=dom:ce_hub
remap_tablespace='(DOM_DATA_HIO:ICUB_DATA,DOM_DATA:ICUB_DATA,DOM_INDEX_HIO:ICUB_INDEX)'
oracle@delphix-vdb-n-3:VDBCB3:/u01/home/oracle/datapump/vdbsl3
$ more ogg_1213_imp.par
directory=DATA_PUMP_DIR1
dumpfile=gg_reint_1213_%U.dmp
logfile=gg_imp_1213.log
cluster=N
parallel=8
tables=dom.YFS_ITEM,dom.YFS_ITEM_ALIAS,dom.YFS_INVENTORY_ITEM,dom.YFS_INVENTORY_ALERTS,dom.YFS_INVENTORY_SUPPLY,dom.EXTN_STORE_REGION_CACHE
table_exists_action=replace
和以前的expdp/impdp一样，需要注意的是表对应的表空间是不是存在，需不需要remap或者新建表空间
在这里我是用了remap_tablespace重定向三个表空间
如下是查看表空间的方法：
源库上：
SQL> select owner,table_name,TABLESPACE_NAME from
dba_tables
where table_name in
('YFS_ITEM','YFS_ITEM_ALIAS','YFS_INVENTORY_ITEM','YFS_INVENTORY_ALERTS','YFS_INVENTORY_SUPPLY','EXTN_STORE_REGION_CACHE');
OWNER TABLE_NAME TABLESPACE_NAME
------------------------------
------------------------------------------------
DOM
YFS_INVENTORY_SUPPLY DOM_DATA_HIO
DOM
YFS_INVENTORY_ITEM DOM_DATA_HIO
DOM YFS_ITEM DOM_DATA_HIO
DOM YFS_ITEM_ALIAS DOM_DATA
DOM
EXTN_STORE_REGION_CACHE DOM_DATA
DOM YFS_INVENTORY_ALERTS DOM_DATA
6 rows selected.
select TABLESPACE_NAME,table_name,index_name,owner from
dba_indexes
where table_name in
('YFS_ITEM','YFS_ITEM_ALIAS','YFS_INVENTORY_ITEM','YFS_INVENTORY_ALERTS','YFS_INVENTORY_SUPPLY','EXTN_STORE_REGION_CACHE');
TABLESPACE_NAME TABLE_NAME INDEX_NAME OWNER
------------------------------
------------------------------ ------------------------------
------------------------------
DOM_DATA
EXTN_STORE_REGION_CACHE
ST_R_CACHE_PK DOM
DOM_DATA YFS_INVENTORY_ALERTS SYS_IL0000228939C00018$$ DOM
DOM_INDEX
EXTN_STORE_REGION_CACHE
EXTN_STORE_REGION_CACHE_I1 DOM
DOM_INDEX YFS_INVENTORY_SUPPLY EXTN_YFS_INVENTORY_SUPPLY_I3 DOM
DOM_INDEX YFS_INVENTORY_SUPPLY YFS_INVENTORY_SUPPLY_I3 DOM
DOM_INDEX YFS_ITEM_ALIAS EXTN_YFS_ITEM_ALIAS_I3 DOM
DOM_INDEX YFS_ITEM_ALIAS YFS_ITEM_ALIAS_PK DOM
DOM_INDEX YFS_ITEM_ALIAS YFS_ITEM_ALIAS_I2 DOM
DOM_INDEX YFS_ITEM_ALIAS YFS_ITEM_ALIAS_I1 DOM
DOM_INDEX YFS_INVENTORY_ALERTS YFS_INVENTORY_ALERTS_PK DOM
DOM_INDEX YFS_INVENTORY_ALERTS YFS_INVENTORY_ALERTS_I4 DOM
DOM_INDEX
YFS_INVENTORY_ALERTS YFS_INVENTORY_ALERTS_I3 DOM
DOM_INDEX YFS_INVENTORY_ALERTS YFS_INVENTORY_ALERTS_I2 DOM
DOM_INDEX YFS_INVENTORY_ALERTS YFS_INVENTORY_ALERTS_I1 DOM
DOM_INDEX_HIO YFS_INVENTORY_SUPPLY YFS_INVENTORY_SUPPLY_PK DOM
DOM_INDEX_HIO YFS_INVENTORY_SUPPLY YFS_INVENTORY_SUPPLY_I2 DOM
DOM_INDEX_HIO YFS_INVENTORY_SUPPLY YFS_INVENTORY_SUPPLY_I1 DOM
DOM_INDEX_HIO YFS_ITEM EXTN_YFS_ITEM_I12 DOM
DOM_INDEX_HIO YFS_ITEM EXTN_YFS_ITEM_I11 DOM
DOM_INDEX_HIO YFS_ITEM EXTN_YFS_ITEM_I10 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_PK DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I6 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I5 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I4 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I3 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I2 DOM
DOM_INDEX_HIO YFS_ITEM YFS_ITEM_I1 DOM
DOM_INDEX_HIO YFS_ITEM EXTN_YFS_ITEM_I9 DOM
DOM_INDEX_HIO YFS_INVENTORY_ITEM YFS_INVENTORY_ITEM_PK DOM
DOM_INDEX_HIO YFS_INVENTORY_ITEM YFS_INVENTORY_ITEM_I1 DOM
30 rows selected.
目标库执行：
select distinct TABLESPACE_NAME from dba_tablespaces;
TABLESPACE_NAME
------------------------------
GGADMIN_DATA
DOM_INDEX
SYSAUX
UNDOTBS1
TEMP
USERS
SYSTEM
NIKE_AUDIT_TBS
ICUB_DATA
ICUB_INDEX
UNDOTBS2
11 rows selected.
可对比发现有几个表空间在目标库是不存在的，需要remap_tablespace
当然也可以用如下的方法来看表空间的问题：
impdp
\'/ as sysdba\' directory=DATA_PUMP_DIR1
dumpfile=gg_reint_1213_%U.dmp sqlfile=myddl.sql
grep
TABLESPACE myddl.sql | sort | uniq | awk -F" " '{print $2}' | uniq
"DOM_DATA"
"DOM_DATA_HIO"
"DOM_DATA"
"DOM_INDEX_HIO"
"DOM_INDEX"
无论怎么看，总而让impdp成功即可
在impdp之后，则新建下replicate进程，并start
replicat aftercsn起起来，如下：
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com) 2> obey
/u01/app/ggadmin/product/12.1.2.1.0/dirprm/r_icub4.oby
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com) 3> dblogin userid
ggadmin, password eur3k2
Successfully logged into database.
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com as ggadmin@VDBCB3)
4> add replicat r_icub3, exttrail /u01/app/ggadmin/product/12.1.2.1.0/dirdat/extract/VDBCB3/e1
REPLICAT added.
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com as ggadmin@VDBCB3)
5> info all
Program Status Group
Lag at Chkpt Time Since Chkpt
MANAGER RUNNING
EXTRACT
RUNNING E_ICUB3 00:00:00 00:00:09
REPLICAT
STOPPED R_ICUB3 00:00:00 00:00:07
GGSCI (delphix-vdb-n-3.va2.b2c.nike.com as ggadmin@VDBCB3)
6> start R_ICUB3 aftercsn 14193095280953
Sending START request to MANAGER ...
REPLICAT R_ICUB3 starting
最后的话要是数据库事务量比较多的话，是可以及时看到replicate的RBA在增长的，或者stats可以看到有DML的操作数目
当然这里的库很闲，自己可以拿相关的表做做DML去测试一下整体功能，这里就省略了。
======================== ENDED ===============================================