goldengate维护要点

oracle Goldengate 维护要点

1.查看各个进程状态和日志

1.1 查看所有进程状态，如

GGSCI (IP0106) 36> info all

Program     Status      Group       Lag           Time Since Chkpt

MANAGER     RUNNING
EXTRACT     RUNNING     EX1         00:00:00      00:00:07
EXTRACT     RUNNING     PMP1        00:00:00      00:00:01
REPLICAT    RUNNING     REP1        00:00:00      00:00:02
REPLICAT    RUNNING     REP2        00:00:00      00:00:04

status是running即为正常，如果是abended或者stopped就需要查看错误原因,lag正常为0

1.2 查看单个进程状态 如

GGSCI (IP0106) 37> info ex1,detail

EXTRACT    EX1       Last Started 2011-06-04 11:35   Status RUNNING
Checkpoint Lag       00:00:00 (updated 00:00:00 ago)
Log Read Checkpoint  Oracle Redo Logs
                     2011-06-12 13:29:37  Seqno 1763, RBA 179811328

  Target Extract Trails:

  Remote Trail Name                                Seqno        RBA     Max MB

  ./dirdat/ex                                        134  834768473       1000

  Extract Source                          Begin             End

  /u01/app/oracle/oradata/orcl/redo04.log  2011-06-04 10:29  2011-06-12 13:29
  /u01/app/oracle/oradata/orcl/redo05.log  2011-05-10 11:17  2011-06-04 10:29
  /u01/app/oracle/oradata/orcl/redo05.log  2011-05-06 16:49  2011-05-10 11:17
  /u01/app/oracle/oradata/orcl/redo04.log  2011-05-02 13:16  2011-05-06 16:49
  /u01/app/oracle/oradata/orcl/redo04.log  2011-04-30 14:05  2011-05-02 13:16
  /u01/app/oracle/oradata/orcl/redo02.log  2011-04-30 12:48  2011-04-30 14:05
  /u01/app/oracle/oradata/orcl/redo05.log  2011-04-30 11:51  2011-04-30 12:48
  /u01/app/oracle/oradata/orcl/redo05.log  2011-04-30 11:29  2011-04-30 11:51
  /u01/app/oracle/oradata/orcl/redo05.log  2011-04-30 06:11  2011-04-30 11:29
  /u01/app/oracle/oradata/orcl/redo05.log  2011-04-29 16:29  2011-04-30 06:11
  Not Available                           * Initialized *   2011-04-29 16:29
  /u01/app/oracle/arch/1_144_745802503.dbf  2011-04-29 15:13  2011-04-29 15:13
  /u01/app/oracle/arch/1_144_745802503.dbf  2011-04-29 15:13  2011-04-29 15:13
  Not Available                           * Initialized *   2011-04-29 15:13


Current directory    /u01/app/ggs

Report file          /u01/app/ggs/dirrpt/EX1.rpt
Parameter file       /u01/app/ggs/dirprm/ex1.prm
Checkpoint file      /u01/app/ggs/dirchk/EX1.cpe
Process file         /u01/app/ggs/dirpcs/EX1.pce
Stdout file          /u01/app/ggs/dirout/EX1.out
Error log            /u01/app/ggs/ggserr.log

主要查看log read checkpoint的时间是否还停留在过去某一时间点，说明extract进程没有在读日志




2.查看系统和进程日志

2.1 查看系统日志 如

GGSCI>view ggsevt

2.2 查看进程日志，在排错时常常用到 如

GGSCI>view report ex1,detail

查看ogg错误和数据库对应错误编号




3.启动各个进程

3.1 启动extract，如

GGSCI>start ex1

3.2 启动datapump 如

GGSCI>start pmp1

3.3 启动replicat 如

GGSCI>start rep1




4.复制停止后的处理

4.1 extract进程因为错误而停止

查看进程错误日志，一般错误都为日志读取错误，可以start ex1手动启动，如果状态还是abended,日志读取还是有问题，建议等源端没有同步任务的时候
delete extract，然后再重新添加add extract的方式重建进程，如

GGSCI>delete ex1

GGSCI>add ex1,vam,begin now              --mysql extract的语法

GGSCI>add exttract ./sq2/ex1,extract ex1 --队列位置要和重建前的队列一致，通过

GGSCI>info exttrail *                    --查看

4.2 datapump因为错误而停止

一般为远端mgr的端口连接不上导致错误，查看远端mgr的端口和防火墙的配置

4.3 replicat因为错误而停止

建议查看replicat进程的错误日志，查看具体原因

如果需要跳过问题事务而恢复复制

GGSCI>start rep1,skiptransaction

一般等replicat的lag时间为0且状态为running说明恢复正常




5.队列的维护

一般都在mgr进程中配置了队列文件的自动删除，如

GGSCI>edit params mgr

purgeoldextracts ./dirdat/ex*,minkeepdays 3,minkeepfiles 10

可以手动修改保持时间，然后refresh mgr,如

GGSCI>refresh mgr