2
异步模式需要在数据库中做一些准备工作:
改数据为归档并启用logminer:
SQL> shutdown immediate
数据库已经关闭。
已经卸载数据库。
ORACLE 例程已经关闭。
SQL> startup mount
ORACLE 例程已经启动。
Total System Global Area 599785472 bytes
Fixed Size 1288820 bytes
Variable Size 171967884 bytes
Database Buffers 423624704 bytes
Redo Buffers 2904064 bytes
数据库装载完毕。
SQL> alter database force logging;
数据库已更改。
SQL> alter database archivelog;
数据库已更改。
SQL> alter database add supplemental log data;
数据库已更改。
设置权限:
发布用户必须有管理Oracle Stream的权限.
用DMBS_STREAMS_AUTH.GRANT_ADMIN_PRIVILEGE赋予权限.
授权管理权限给用户户,允许用户户改实例参数和创建对象.
SQL> grant select_catalog_role to scott;
Grant succeeded.
SQL> grant execute_catalog_role to scott;
Grant succeeded.
SQL> grant create table to scott;
Grant succeeded.
SQL> grant create session to scott;
Grant succeeded.
SQL> grant dba to scott;
Grant succeeded.
SQL> grant execute on dbms_cdc_publish to scott;
Grant succeeded.
SQL> grant select on change_sets to scott;
Grant succeeded.
SQL> grant create tablespace to scott;
Grant succeeded.
SQL> grant create sequence to scott;
Grant succeeded.
SQL> grant unlimited tablespace to scott;
Grant succeeded.
SQL> execute dbms_streams_auth.grant_admin_privilege(grantee=>'scott');
PL/SQL procedure successfully completed.
工作对象(Work schema)必须有select any table权限
SQL> grant select any table to scott;
Grant succeeded.
设置数据库参数,因为要用logminer和stream,设置相应内存参数;
SQL> alter system set java_pool_size=50m;
系统已更改。
SQL> alter system set streams_pool_size=100m;
系统已更改。
需要把table添加到日志组 (后来测试,这步骤在KM已经有,而且做了以后反而导致错误)
SQL> conn scott/tiger
Connected.
SQL> alter table dept add supplemental log data(all) columns;
Table altered.
这个时候要确认数据库打开了Streams Capture功能
SQL> select * from v$option where parameter =’Streams Capture’;
PARAMETER VALUE
-------------------- --------
Streams Capture TRUE
标准版不支持Advanced Replication 和 Stream. 1. 不支持Streams
2. 不支持Advanced Replication的multimaster方式
解决需：1.安装企业版 2.安装Advanced Replication 组件 3.执行CATREP.SQL 。 fact: Oracle Server - Standard Edition symptom: ORA-00439: feature not enabled: %s fix: 1. Install Enterprise Edition (see Note 117048.1) 2. Install Advanced Replication Option using installer 3. Run CATREP.SQL script. UNIX: $ORACLE_HOME/rdbms/admin/catrep.sql NT: ORACLE_HOME\rdbms\admin\catrep.sql
导入JKM的知识模块:
这里,我们把三种模式的JKM都导入,然后引入时候,我们做一个对比,看看JKM的差别
编辑模型,点击日志记录 TAB页,选择相应的日志记录模式和JKM:
我们选择一致性设置,可以看到我们刚才引入的三个JKM,都属于一致性读
每个JKM具体是怎么实现,差别是什么呢?我们通过编辑JKM来看他们具体做什么?
选择项目栏:
我们可以看到这个JKM的顺序180的命令,删除触发器,通过下面知道,它还是利用触发器实现.
我们那么再看JKM Oracle consistent (update date)
看到顺序500创建了一个删除后的触发器,双机这个命令,可以看到相应的创建语法,这些JKM虽然是是一致的模式,但不是异步CDC的概念.
我们看JKM Oracle consistent (LOGMINER)
我们可以看到它是通过强制数据库记录日志实现的,这个是异步CDC的方法.这时,我们在模型编辑的时候,选择这个JKM.
选择JKM Oracle 10g consistent (LOGMINER).CDC,可以看到有3个选项,一个是异步模式,就是我们提到的HotLog模式,还有一个是自动配置,是下一节要介绍的AUTOLOG模式.
最后一个参数是定义日志表的存储表空间,手工输入就可以,这里输入USERS.
看下面的说明: 如果选择异步模式,JKM最小化在源系统上的负载,用Oracle CDC HotLog模式,.这个选项要求归档模式,强制日志.
启动日志报错,查看operator
SQL> l
1 BEGIN
2 DBMS_CDC_PUBLISH.CREATE_CHANGE_SET(
3 change_set_name => 'CDC',
4 description => 'Sunopsis change set for model : CDC',
5
6 change_source_name => 'HOTLOG_SOURCE',
7 begin_date => sysdate
8
9 );
10* END;
SQL> /
*
1031 : 42000 : java.sql.SQLException: ORA-01031: insufficient privileges
ORA-06512: at "SYS.DBMS_CDC_PUBLISH", line 705
ORA-06512: at line 1
因为没有select any change_sets权限,查了10多分钟.
再次启动日志又报错:
查看描述:
30567 : 99999 : java.sql.BatchUpdateException: ORA-30567: name already used by an existing log group
30567 : 99999 : java.sql.SQLException: ORA-30567: name already used by an existing log group
java.sql.BatchUpdateException: ORA-30567: name already used by an existing log group
因为我们在前面做了下面动作
alter table dept add supplemental log data(all) columns;
删除这个日志数据
SQL> alter table dept drop supplemental log data(all) columns;
Table altered.
检查日志组
SQL> select owner,table_name,log_group_name from all_log_groups;
OWNER TABLE_NAME LOG_GROUP_NAME
---------- ---------- ------------------------------
SYS ENC$ ENC$_LOG_GRP
SCOTT EMP LOG_GROUP_EMP
SCOTT DEPT LOG_GROUP_DEPT
SCOTT EMP SYS_C0011135
SCOTT EMP SYS_C0011136
SCOTT EMP SYS_C0011137
SCOTT DEPT SYS_C0011175
SCOTT DEPT SYS_C0011176
SCOTT DEPT SYS_C0011177
SCOTT DEPT DEPT
SCOTT DEPT SYS_C0011442
11 rows selected.
删除同名日志组
SQL> alter table dept drop supplemental log group dept;
Table altered.
重新启动 operator
又报了一个新错误
BEGIN
DBMS_CDC_PUBLISH.CREATE_CHANGE_TABLE(
owner => 'SCOTT',
change_table_name => 'J$DEPT',
change_set_name => 'CDC',
source_schema => 'SCOTT',
source_table => 'DEPT',
column_type_list => 'DEPTNO NUMBER(2) ',
capture_values => 'new',
rs_id => 'n',
row_id => 'n',
user_id => 'n',
timestamp => 'y',
object_id => 'n',
source_colmap => 'n',
target_colmap => 'n',
ddl_markers => 'n',
options_string => 'users'
);
END;
ERROR at line 1:
ORA-00922: missing or invalid option
ORA-06512: at "SYS.DBMS_CDC_PUBLISH", line 783
ORA-06512: at line 2
做10046跟踪:
SQL> conn scott/tiger
Connected.
SQL> alter session set events '10046 trace name context forever, level 8';
Session altered.
SQL> BEGIN
2 DBMS_CDC_PUBLISH.CREATE_CHANGE_TABLE(
3 owner => 'SCOTT',
change_table_name => 'J$DEPT',
4 5 change_set_name => 'CDC',
source_schema => 'SCOTT',
source_table => 'DEPT',
6 7 8 column_type_list => 'DEPTNO NUMBER(2) ',
capture_values => 'new',
rs_id => 'n',
row_id => 'n',
9 10 11 12 user_id => 'n',
timestamp => 'y',
13 14 object_id => 'n',
source_colmap => 'n',
target_colmap => 'n',
15 16 17 ddl_markers => 'n',
18 options_string => 'users'
19 );
20 END;
21 /
BEGIN
*
ERROR at line 1:
ORA-00922: missing or invalid option
ORA-06512: at "SYS.DBMS_CDC_PUBLISH", line 783
ORA-06512: at line 2
SQL> alter session set events '10046 trace name context off';
Session altered.
查看trace发现是创建J$DEPT表时出现的错误
PARSE ERROR #7:len=253 dep=1 uid=81 ct=1 lid=81 tim=1241053855723263 err=922
CREATE TABLE SCOTT.J$DEPT ( OPERATION$ CHAR(2), CSCN$ NUMBER, COMMIT_TIMESTAMP$ DATE, XIDUSN$ NUMBER, XIDSLT$ NUMBER, XIDSEQ$ NUMBER
, TIMESTAMP$ DATE, DEPTNO NUMBER(2) ) USERS PARTITION BY RANGE (cscn$) (PARTITION p1 VALUES LESS THAN (281474976710656))
EXEC #3:c=0,e=56,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1241053855731599
Akkkkkkkkkkkkkkkk
FETCH #3:c=0,e=514,p=0,cr=4,cu=0,mis=0,r=1,dep=1,og=4,tim=1241053855732165
=====================
PARSING IN CURSOR #6 len=83 dep=1 uid=0 ct=3 lid=0 tim=1241053855735395 hv=1936881461 ad='2fa86004' sqlid='9gbtbxttr4xtp'
select jarname from javajarobjects$ where bjname=:1 and owner#=:2 and namespace=:3
END OF STMT
PARSE #6:c=2000,e=1106,p=0,cr=0,cu=0,mis=1,r=0,dep=1,og=4,tim=1241053855735387
EXEC #6:c=999,e=1650,p=0,cr=0,cu=0,mis=1,r=0,dep=1,og=4,tim=1241053855737156
FETCH #6:c=1000,e=280,p=0,cr=4,cu=0,mis=0,r=0,dep=1,og=4,tim=1241053855737494
STAT #6 id=1 cnt=0 pid=0 pos=1 bj=12468 p='TABLE ACCESS FULL JAVAJAROBJECTS$ (cr=4 pr=0 pw=0 time=0 us cost=2 size=44 card=1)'
oracle.jdbc.driver.OracleSQLException: ORA-00922: missing or invalid option
at oracle.jdbc.driver.DatabaseError.throwSqlException(DatabaseError.java)
at oracle.jdbc.driver.T2SConnection.check_error(T2SConnection.java)
at oracle.jdbc.driver.T2SStatement.checkError(T2SStatement.java)
at oracle.jdbc.driver.T2SStatement.executeForRows(T2SStatement.java)
at oracle.jdbc.driver.OracleStatement.doExecuteWithTimeout(OracleStatement.java)
at oracle.jdbc.driver.OracleStatement.executeInternal(OracleStatement.java)
at oracle.jdbc.driver.OracleStatement.execute(OracleStatement.java)
at oracle.jdbc.driver.OracleStatementWrapper.execute(OracleStatementWrapper.java)
at oracle.CDC.ChangeTable.createTable(ChangeTable.java:1657)
at oracle.CDC.ChangeTable.create(ChangeTable.java:1527)
at oracle.CDC.PublishApi.createChangeTable(PublishApi.java:427)
我们到数据库执行发现,原来是创建表时,语法有错,没有加tablespace 参数.
SQL> CREATE TABLE SCOTT.J$DEPT (
2 OPERATION$ CHAR(2), CSCN$ NUMBER, COMMIT_TIMESTAMP$ DATE, XIDUSN$ NUMBER,
3 XIDSLT$ NUMBER, XIDSEQ$ NUMBER,
4 TIMESTAMP$ DATE, DEPTNO NUMBER(2) )
5 users
6 PARTITION BY RANGE (cscn$) (PARTITION p1 VALUES LESS THAN (281474976710656));
users
*
ERROR at line 5:
ORA-00922: missing or invalid option
SQL> l5
5* users
SQL> c/users/tablespace users
5* tablespace users
SQL> l
1 CREATE TABLE SCOTT.J$DEPT (
2 OPERATION$ CHAR(2), CSCN$ NUMBER, COMMIT_TIMESTAMP$ DATE, XIDUSN$ NUMBER,
3 XIDSLT$ NUMBER, XIDSEQ$ NUMBER,
4 TIMESTAMP$ DATE, DEPTNO NUMBER(2) )
5 tablespace users
6* PARTITION BY RANGE (cscn$) (PARTITION p1 VALUES LESS THAN (281474976710656))
SQL> /
Table created.
到此我们已经找到原因是ODI JKM的参数定义有问题
这个时候,我们进行修改,再次执行
SQL> BEGIN
2 DBMS_CDC_PUBLISH.CREATE_CHANGE_TABLE(
3 owner => 'SCOTT',
4 change_table_name => 'J$DEPT',
change_set_name => 'CDC',
source_schema => 'SCOTT',
5 6 7 source_table => 'DEPT',
8 column_type_list => 'DEPTNO NUMBER(2) ',
9 capture_values => 'new',
rs_id => 'n',
row_id => 'n',
10 11 12 user_id => 'n',
13 timestamp => 'y',
14 object_id => 'n',
source_colmap => 'n',
15 16 target_colmap => 'n',
17 ddl_markers => 'n',
18 options_string => 'tablespace users'
);
19 20 END;
21 /
PL/SQL procedure successfully completed.
可以看到执行成功.
修改JKM,加入tablespace
重新启动日志,需要做下面动作
SQL> alter table dept drop supplemental log group dept;
Table altered.
SQL> BEGIN
2 DBMS_CDC_PUBLISH.DROP_CHANGE_TABLE(
3 owner => 'SCOTT',
4 change_table_name => 'J$DEPT',
5 FORCE_FLAG => 'Y');
6 end;
7 /
PL/SQL procedure successfully completed.
BEGIN
DBMS_CDC_PUBLISH.DROP_CHANGE_TABLE(
owner => 'SCOTT',
change_table_name => 'J$DEPT',
FORCE_FLAG => 'Y');
end;
/
删除创建CDC对象表的数据
SQL> delete from SNP_CDC_OBJECTS;
3 rows deleted.
SQL> commit;
Commit complete.
Operator创建订阅又出现错误
继续跟踪错误,后来测试发现订阅名称不能和变化集名称一致.
查了1天时间,耗时啊.
SQL> alter session set events '10046 trace name context forever, level 12';
Session altered.
SQL> oradebug setmypid
Statement processed.
SQL> BEGIN
2 DBMS_CDC_SUBSCRIBE.CREATE_SUBSCRIPTION(
3 change_set_name => 'CDC',
4 description => 'Common subscription to be used to manage journalized data',
subscription_name => 'CDC'
);
5 6 7 END;
8 /
BEGIN
*
ERROR at line 1:
ORA-29532: Java call terminated by uncaught Java exception:
java.lang.StringIndexOutOfBoundsException: String index out of range: 4
ORA-06512: at "SYS.DBMS_CDC_SUBSCRIBE", line 12
ORA-06512: at line 2
SQL> oradebug tracefile_name
/u01/app/oracle/diag/rdbms/devdb/devdb/trace/devdb_ora_5895.trc
修改JKM:
把订阅名称多加个C.
再次启动日志,发现执行成功,数据源出现绿色图标.
查看 operator