OGG新版本出来后，大部分情况都会选择升级到新版本，不过也存在部分原因导致只能用11G的ogg，但是又因为存在部分bug等原因，要做11G小版本之间的升级，如从11.1.x.x.x升级到11.2.1.0.1，下面对这种情况做个简单的步骤说明。
1.停止所有进程
源端和目标端的进程全部停止，包括mgr进程
GGSCI>stop *
GGSCI>stop mgr
本文采用用新版本覆盖就版本的方法升级，故防止进程信息丢失，记录当前各进程的信息
源端：
GGSCI (qzdb2) 2> info * EXTRACT EXT1 Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag unknown (updated 02:07:16 ago) Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:30:09 Thread 1, Seqno 162502, RBA 373017020 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:28:59 Thread 2, Seqno 85757, RBA 1029120 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:29:14 Thread 3, Seqno 39513, RBA 1349712 EXTRACT EXT1DP Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:14 ago) Log Read Checkpoint File ./trails/w1002906 2016-12-08 07:29:49.000000 RBA 9404504 EXTRACT EXT2 Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:13 ago) Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:30:09 Thread 1, Seqno 162502, RBA 373017020 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:28:59 Thread 2, Seqno 85757, RBA 1029632 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:29:14 Thread 3, Seqno 39513, RBA 1349712 EXTRACT EXT2DP Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:11 ago) Log Read Checkpoint File ./trails/w2008990 2016-12-08 07:30:09.000000 RBA 6816055 EXTRACT EXT3 Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:09 ago) Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:30:09 Thread 1, Seqno 162502, RBA 373017020 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:28:59 Thread 2, Seqno 85757, RBA 1029632 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:29:14 Thread 3, Seqno 39513, RBA 1349712 EXTRACT EXT3DP Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:08 ago) Log Read Checkpoint File ./trails/w3000321 2016-12-08 07:30:08.000000 RBA 3873403 EXTRACT EXT4 Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:07 ago) Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:30:09 Thread 1, Seqno 162502, RBA 373017020 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:28:59 Thread 2, Seqno 85757, RBA 1030144 Log Read Checkpoint Oracle Redo Logs 2016-12-08 07:29:14 Thread 3, Seqno 39513, RBA 1349712 EXTRACT EXT4DP Last Started 2016-12-07 12:00 Status STOPPED Checkpoint Lag 00:00:00 (updated 02:07:05 ago) Log Read Checkpoint File ./trails/w4002813 2016-12-08 07:30:09.000000 RBA 8638179
目标端：
GGSCI (QZJHK) 8> info * REPLICAT REP1 Last Started 2016-11-14 18:46 Status STOPPED Checkpoint Lag 00:00:00 (updated 00:10:27 ago) Log Read Checkpoint File /oradata/ggs/trails/w1007920 2016-12-08 07:29:49.236830 RBA 9404532 REPLICAT REP2 Last Started 2016-11-14 18:46 Status STOPPED Checkpoint Lag 00:00:00 (updated 00:10:26 ago) Log Read Checkpoint File /oradata/ggs/trails/w2026025 2016-12-08 07:30:09.236829 RBA 6816246 REPLICAT REP3 Last Started 2016-11-14 18:46 Status STOPPED Checkpoint Lag 00:00:00 (updated 00:10:24 ago) Log Read Checkpoint File /oradata/ggs/trails/w3000956 2016-12-08 07:30:08.236788 RBA 3873431 REPLICAT REP4 Last Started 2016-11-16 14:31 Status STOPPED Checkpoint Lag 00:00:00 (updated 00:10:22 ago) Log Read Checkpoint File /oradata/ggs/trails/w4007019 2016-12-08 07:30:09.236898 RBA 8640562
2.备份原有目录
源端：
$ cd /u01
$ tar -cf ggs.tar ggs
目标端：
$ cd /oradata
$ tar -cv ggs.tar ggs
3.使用新版本覆盖就版本
上传新版本到ggs目录并解压
$ cd /u01/ggs
$ unzip ogg112101_ggs_AIX5_AIX6.1_ppc_ora11g_64bit.zip
$ tar -xf ggs_AIX_ppc_ora11g_64bit.tar
在这里解压的时候可能会报错
解决方法：
在ggs目录下随便建个目录，如tmp
$ mkdir tmp
将ggs_AIX_ppc_ora11g_64bit.tar一到tmp目录并解压
$ mv ggs_AIX_ppc_ora11g_64bit.tar tmp
$ cd tmp
$ tar -xf ggs_AIX_ppc_ora11g_64bit.tar
然后将新解压出来的文件mv到ggs目录覆盖原来的旧版本（dirprm文件不要覆盖，不然进程要重建）
$ pwd
/u01/ggs/tpm
$ rm -rf dirprm
$ mv * /u01/ggs
这样新的版本就覆盖就版本了
ggsci命令进入ogg控制台
$ ./ggsci Oracle GoldenGate Command Interpreter for Oracle Version 11.2.1.0.1 OGGCORE_11.2.1.0.1_PLATFORMS_120423.0230 AIX 5L, ppc, 64bit (optimized), Oracle 10.2 on Apr 23 2012 07:27:46 Copyright (C) 1995, 2012, Oracle and/or its affiliates. All rights reserved.
可以看到源端版本已经升级到11.2.1.0.1
目标端也用同样的方法升级
4.源端启动进程
升级好后，在源端启动进程
GGSCI > start mgr
GGSCI > start *
发现抽取进程和传输进程起不来，查看日志报错：
ERROR OGG-01416 File ./trails/w1002906, with format RELEASE 10.4/11.1, does not match current format specification of RELEASE 11.2.Modify the parameter file to specify format RELEASE 10.4/11.1 or issueETROLLOVER prior to restar
因为ogg版本换过了则trail文件不能匹配，报错中已经提示解决办法，将参数文件中制定版本的参数改成现在的版本，或者使用ETROLLOVER命令
查看参数文件的配置，本来就没制定版本，所以不用修改，采用ETROLLOVER命令，如下：
对抽取进程和传输进程做ETROLLOVER操作
GGSCI > alter ext1 ETROLLOVER GGSCI > alter ext1dp ETROLLOVER GGSCI > alter ext2 ETROLLOVER GGSCI > alter ext2dp ETROLLOVER GGSCI > alter ext3 ETROLLOVER GGSCI > alter ext3dp ETROLLOVER GGSCI > alter ext4 ETROLLOVER GGSCI > alter ext4dp ETROLLOVER
再次启动进程：
GGSCI > start *
进程起来后，抽取进程正常挖掘
传输进程虽然是running的，但是一直停留在原来的记录点，目标端存放trail文件的目录下文件大小为0，不增长
-rw-rw-rw- 1 oracle dba 9404532 Dec 08 07:29 w1007920 -rw-rw-rw- 1 oracle dba 3873431 Dec 08 07:29 w3000956 -rw-rw-rw- 1 oracle dba 6816246 Dec 08 07:29 w2026025 -rw-rw-rw- 1 oracle dba 8640562 Dec 08 07:29 w4007019 -rw-rw-rw- 1 oracle dba 0 Dec 08 11:38 w1007921 -rw-rw-rw- 1 oracle dba 0 Dec 08 11:38 w3000957 -rw-rw-rw- 1 oracle dba 0 Dec 08 11:38 w2026026 -rw-rw-rw- 1 oracle dba 0 Dec 08 11:39 w4007020
查看ggserr.log日志如下提示：
WARNING OGG-01519 Oracle GoldenGate Capture for Oracle, ext1dp.prm: Waiting at EOF on input trail file ./trai ls/w1002906, which is not marked as complete; but succeeding trail file ./trails/w1002907 exists. If ALTER ETROLLOVER has been perfo rmed on source extract, ALTER EXTSEQNO must be performed on each corresponding downstream reader.
因为源端的抽取进程执行ETROLLOVER操作后，进程的extseqno和extrba号被重置至下一个extseqno号的第0号extrba，
而datapump传输进程并不知道extract进程发生了这个变化，仍然守望在extract ETROLLOVER前的extseqno和extrba号上，
永远也等不到这个extseqno上有新的RBA变化，造成就无法将extract新抽取到的数据，传递到目标端去。
所以，在extract进程ETROLLOVER后，需要使用"Alterextract group_name EXTSEQNO X, EXTRBA 0"的命令，重置checkpoint位置，datadump进程才能正常的将数据继续传递到远端。
查看trail文件目录，发现进程的extseqno号已经跳到下一个
故重新指定extseqno和extrba，如下：
GGSCI > alter ext1dp extseqno 2907 extrba 0 GGSCI > alter ext2dp extseqno 8991 extrba 0 GGSCI > alter ext3dp extseqno 322 extrba 0 GGSCI > alter ext4dp extseqno 2814 extrba 0
重新启动传输进程，trail文件正常传输到目标端。
5.目标端启动进程
GGSCI > start *
应用进程无法启动，日志报错如下：
ERROR OGG-00665 Oracle GoldenGate Delivery for Oracle, rep4.prm: OCI Error describe for query (status = 942 -ORA-00942: table or view does not exist), SQL<SELECT a.current_dir, a.seqno, a.rba, a.audit_ts, a.log_csn, a.log_xid, a.log_cmplt_c sn, a.log_cmplt_xids, b.log_cmplt_xids FROM GOLDENGATE.CHKPOINT a LEFT JOIN GOLDENGATE.CHKPOINT_lox b ON a.group_name = b.group_name AND a.group_key = b.group_key AND a.log_cmplt_csn = b.log_cmplt_csn WHERE a.group_name = 'REP4' AND a.group_key = 2420532828>. 2016-12-08 11:30:30 ERROR OGG-01668 Oracle GoldenGate Delivery for Oracle, rep4.prm: PROCESS ABENDING.
原因：ogg_11.1.1.1.2每条replicat进程只有CHECKPOINT一张表，而ogg_11.2.x.x.1每条replicat进程有CHECKPOINT,CHECKPOINT_LOX两张表，
所以，如果在升级后，直接启动replicat进程，是无法启动
提示表不存在，这张表是指：GOLDENGATE.CHKPOINT表（CHECKPOINT_LOX表），重建replicat的最终目的是在重建进程时，自动将两张checkpoint表都自动创建起来。
删除replicat进程与checkpoint表
GGSCI > dblogin userid GOLDENGATE password GOLDENGATE GGSCI > delete CHECKPOINTTABLE GOLDENGATE.CHKPOINT GGSCI > ADD CHECKPOINTTABLE GOLDENGATE.CHKPOIN
重建应用进程
GGSCI > add replicat rep1 exttrail /oradata/ggs/trails/w1 CHECKPOINTTABLE GOLDENGATE.CHKPOINT GGSCI > alter rep1 extseqno 7920 extrba 9404532 --可不做，直接指定到下一个 GGSCI > add replicat rep2 exttrail /oradata/ggs/trails/w2 CHECKPOINTTABLE GOLDENGATE.CHKPOINT GGSCI > alter rep2 extseqno 26025 extrba 6816246 --可不做，直接指定到下一个 GGSCI > add replicat rep3 exttrail /oradata/ggs/trails/w3 CHECKPOINTTABLE GOLDENGATE.CHKPOINT GGSCI > alter rep3 extseqno 956 extrba 3873431 --可不做，直接指定到下一个 GGSCI > add replicat rep4 exttrail /oradata/ggs/trails/w4 CHECKPOINTTABLE GOLDENGATE.CHKPOINT GGSCI > alter rep4 extseqno 7019 extrba 8640562 --可不做，直接指定到下一个
若是将应用进程制定到原来的点的话，会和传输进程一样找不到这个点，无法应用，报错如下：
WARNING OGG-01519 Oracle GoldenGate Delivery for Oracle, rep1.prm: Waiting at EOF on input trail file /oradat a/ggs/trails/w1007920, which is not marked as complete; but succeeding trail file /oradata/ggs/trails/w1007921 exists. If ALTER ETRO LLOVER has been performed on source extract, ALTER EXTSEQNO must be performed on each corresponding downstream reader.
重新指定应用进程记录点
查看trail文件目录下文件，找到源端etrollover后传输过来的第一个文件
GGSCI > alter rep1 extseqno 7921 extrba 0 GGSCI > alter rep2 extseqno 26026 extrba 0 GGSCI > alter rep3 extseqno 957 extrba 0 GGSCI > alter rep4 extseqno 7020 extrba 0
重新启动进程
GGSCI > start *
发现进程正常应用
至此，ogg版本升级结束。
来自 “ ITPUB博客 ” ，链接：https://blog.itpub.net/29956245/viewspace-3011263/ ，如需转载，请注明出处，否则将追究法律责任。
请登录后发表评论
登录
全部评论
