oracle goldengate非并行进程转换为并行进程

oracle goldengate非并行进程转换为并行进程

在上一期的文章中写道了直接创建并行进程的方式对大事务进行分解,这对于新建立同步进程的时候提前规划是很有帮助的,但是如果对已经进行了同步的进程重新建立需要耗时比较长,Oracle提供了非并行进程转换到并行进程的方式。

参考连接如下:

https://docs.oracle.com/en/middleware/goldengate/core/21.3/coredoc/quickstart-switching-non-integrated-replicat-parallel-non-integrated-replicat.html

已经按照Oracle提供的方法进行测试验证,可以直接切换到并行复制进程。

具体步骤:

Switching from Nonintegrated Replicat to Parallel Nonintegrated Replicat

The process for switching to parallel integrated or parallel nonintegrated Replicat is the same for all Replicat modes. This topic describes the process to switch from nonintegrated Replicat to parallel nonintegrated Replicat.

Before Starting the Switching Process

1.Create a parallel nonintegrated Replicat process, repe that reads from the exisiting trail file:

复制代码
ADD REPLICAT repe, PARALLEL, EXTTRAIL ea, checkpointtable ggadmin.ggs_checkpoint1

2.Do not start the parallel nonintegrated Replicat (repe).

3.Stop the current nonintegrated Replicat, repea.

复制代码
STOP REPLICAT repea

4.On the target side, access the Replicat report file (.rpt) to know the values of the following components:

Last applied CSN by the current nonintegrated Replicat process.

Trail sequence and RBA of the exisitng Replicat process.

To access the details of the Replicat, run the command:

复制代码
INFO REPLICAT repea DETAIL

The output for this command shows an output similar to the following:

复制代码
Replicat   REPEA     Last Started 2022-06-16 04:21   Status STOPPED
Description          eastt
Checkpoint Lag       00:00:00 (updated 01:59:59 ago)
Log Read Checkpoint  File east/ea000000009
                     2022-06-14 04:38:34.084220  RBA 9382
Settings Profile     Default
Encryption Profile   LocalWallet

Current Log BSN value: (no data)

Last Committed Transaction CSN value: 50698907

  Extract Source                          Begin             End             

  east/ea000000009                        2022-06-16 04:21  2022-06-14 04:38
  east/ea000000000                        * Initialized *   2022-06-16 04:21
  east/ea000000009                        2022-06-14 04:38  2022-06-14 04:38
  east/ea000000009                        2022-06-14 04:38  2022-06-14 04:38
  east/ea000000009                        2022-06-16 03:55  2022-06-14 04:38
  east/ea000000000                        * Initialized *   2022-06-16 03:55
  east/ea000000000                        * Initialized *   First Record    
  east/ea000000000                        * Initialized *   First Record    
  east/ea000000000                        * Initialized *   First Record    
  east/ea000000000                        * Initialized *   First Record    

Current directory    /scratch/preeshuk/ggtest/install_ogg21.3_210725/bin

Report file          /scratch/oggoradep/var/lib/report/REPEA.rpt
Parameter file       /scratch/oggoradep/etc/conf/ogg/REPEA.prm
Checkpoint file      /scratch/oggoradep/var/lib/checkpt/REPEA.cpr
Checkpoint table     DBEAST.GGADMIN.GGS_CHECKPOINT
Process file         /scratch/oggoradep/var/run/REPEA.pcr
Error log            /scratch/oggoradep/var/log/ggserr.log

Start the Switching Process

To start using the nonintegrated parallel Replicat, you need to alter it to port the content from the other Replicat. Use the following steps to perform this task:

1.Run the ALTER REPLICAT command as follows:

复制代码
ALTER REPLICAT replicat_name, EXTSEQNO extseqno,  EXTRBA extrba

For example, for the Replicat repe, here's the command:

复制代码
ALTER REPLICAT repe, EXTSEQNO 9, EXTRBAm 9382

2.Start the newly created parallel nonintegrated Replicat process using the following command:

复制代码
START REPLICAT repe AFTERCSN csn_value

For example:

复制代码
START REPLICAT repe AFTERCSN 50698907

This starts the Replicat at the specified CSN value in the trail file.

上期文章连接:

https://blog.csdn.net/bing_yuan/article/details/147766818?spm=1011.2124.3001.6209

相关推荐
西柚小萌新31 分钟前
【大模型:知识图谱】--3.py2neo连接图数据库neo4j
数据库·知识图谱·neo4j
wangfenglei12345638 分钟前
mybatis打印完整的SQL,p6spy
数据库·sql·mybatis
__风__41 分钟前
PostgreSQL ERROR: out of shared memory处理
数据库·postgresql
占星安啦1 小时前
一个html实现数据库自定义查询
java·前端·javascript·数据库·动态查询
天空之城夢主1 小时前
MySQL 全量、增量备份与恢复
数据库·mysql·oracle
Elastic 中国社区官方博客1 小时前
连接关键点:使用 ES|QL 联接实现更丰富的可观测性洞察
大数据·数据库·sql·elasticsearch·搜索引擎·全文检索
元闰子1 小时前
AI Agent需要什么样的数据库?
数据库·人工智能·后端
什么都想学的阿超1 小时前
【PostgreSQL 02】PostgreSQL数据类型革命:JSON、数组与地理信息让你的应用飞起来
数据库·postgresql·json
小陈又菜2 小时前
SQL Indexes(索引)
数据库·sql·索引
TDengine (老段)3 小时前
TDengine 的 AI 应用实战——运维异常检测
大数据·数据库·人工智能·物联网·时序数据库·tdengine·涛思数据