OGG-01028錯誤處理
OGG抽取程式報錯如下:
ERROR OGG-01028 encountered commit SCN 3090.2667129739 (13274116074379) that is not greater than the highest
SCN already processed 3090.2667130101 (13274116074741) Redo Thread 2 (2) xid 18.24.28309117 (0x0012.0018.01aff67d), starting seq.rb
a 3509.60173328, scn 3090.2667129738 (13274116074378), commit seq.rba 3509.60174196 commit timestamp 2013-03-18 22:02:35.000000.
GGSCI (nhdb16) 16> info ESCDB002
EXTRACT ESCDB002 Last Started 2013-03-19 11:03 Status ABENDED
Checkpoint Lag 00:00:02 (updated 11:24:37 ago)
Log Read Checkpoint Oracle Redo Logs
2013-03-19 21:28:38 Thread 1, Seqno 4735, RBA 21603584
Log Read Checkpoint Oracle Redo Logs
2013-03-19 21:28:54 Thread 2, Seqno 3585, RBA 50791440
-rw-rw-rw- 1 oracle oinstall 9999525 Mar 20 16:36 esc00019
-rw-rw-rw- 1 oracle oinstall 7943251 Mar 20 18:55 esc00020
Solution Details:
1. Do an ETROLLOVER on Extract, and take note of the new sequence number of the trail file.
ALTER EXTRACT ESCDB002, ETROLLOVER
2. Start extract
START EXTRACT ESCDB002
GGSCI (nhdb16) 76> info ESCDB002
EXTRACT ESCDB002 Last Started 2013-03-21 15:05 Status RUNNING
Checkpoint Lag 20:10:38 (updated 00:00:08 ago)
Log Read Checkpoint Oracle Redo Logs
2013-03-20 18:54:46 Thread 1, Seqno 4791, RBA 18211856
Log Read Checkpoint Oracle Redo Logs
2013-03-20 18:55:00 Thread 2, Seqno 3661, RBA 87376180
3. Send PUMP, LOGEND, to see if it's at the end of the previous trail.
SEND EXTRACT PNFHC002, LOGEND
4. Once it is at the end of the trail file, You must stop the pump, and do an ETROLLOVER for it too. Take note of the new trail file sequence number that is created from this step.
STOP EXTRACT PNFHC002
ALTER EXTRACT PNFHC002, ETROLLOVER
GGSCI (nhdb16) 80> info PNFHC002
EXTRACT PNFHC002 Initialized 2013-03-20 14:35 Status STOPPED
Checkpoint Lag 00:00:00 (updated 00:00:11 ago)
Log Read Checkpoint File /gglog/dirdat/enfzh002/esc00020
2013-03-20 18:54:47.000000 RBA 7943251
2013-03-20 14:17:30 INFO OGG-01520 Rollover performed. For each affected output trail of Version 10 or higher format, after starting the source extract, issue ALTER EXTSEQNO for that trail's reader (either pump EXTRACT or REPLICAT) to move the reader's scan to the new trail file; it will not happen automatically.
EXTRACT altered.
[nhdb25:oracle]ls -la
total 0
drwxr-xr-x 2 oracle oinstall 256 Mar 20 21:29 .
drwxr-xr-x 12 oracle oinstall 4096 Mar 14 16:53 ..
-rw-rw-rw- 1 oracle oinstall 47937377 Mar 20 18:55 rsc0004
5. Alter the pump to SEQNO to the new trail file created from step #1.
ALTER EXTRACT PNFHC002, EXTSEQNO 21 EXTRBA 0
6. Restart pump
START EXTRACT PNFHC002
7. Send Replicat, LOGEND to make sure it has processed all the remaining data, and stop Replicat.
SEND REPLICAT RSCDB002, LOGEND
STOP REPLICAT RSCDB002
8. If replicat is not at end of trail, generate a report and forcestop replicat
SEND REPLICAT RSCDB002, REPORT
STOP REPLICAT RSCDB002
9. Add the following parameters to replicat parameter file to allow replicat to process each trail record as a single transaction, or set them to 1 if you have any of these parameters:
GROUPTRANSOPS 1
MAXTRANSOPS 1
10. Restart replicat
START REPLICAT RSCDB002
11. Once replicat has completely processed the trail, stop the replicat
STOP REPLICAT RSCDB002
12. Edit the replicat parameter file:
- Add parameter HANDLECOLLISIONS to Replicat parameter file
- Remove or comment out GROUPTRANSOPS and MAXTRANSOPS or revert them back to their original values.
Note: There are pre-conditions for using HANDLECOLLISIONS . Ther emust be either
a. no pkupdates
or
b. extract has "FETCHOPTIONS FETCHPKUPDATECOLS"Also all the tables should have Primary key or unique index on the table to avoid data integrity issues when using handlecollisions.
13. ALTER REPLICAT, SEQNO to the new trail file created in step #4.
ALTER REPLICAT RSCDB002, EXTSEQNO 5 EXTRBA 0
14. Start Replicat
START REPLICAT RSCDB002
15. Once Replicat has processed the out of order SCN operations, disable HANDLECOLLISIONS. You could also look for the CSN and wait for Replicat to checkpoint past it.
SEND REPLICAT RSCDB002, NOHANDLECOLLISIONS.
16.Edit the replicat parameter and comment out the HANDLECOLLISIONS parameter. You do not need to stop/restart replicat. This will ensure that on any subsequent replicat restarts the parameter is disabled.
根據oracle給出的文件,測試引數如下:
THREADOPTIONS MAXCOMMITPROPAGATIONDELAY 60000 IOLATENCY 90000
來自 “ ITPUB部落格 ” ,連結:http://blog.itpub.net/29468144/viewspace-1079330/,如需轉載,請註明出處,否則將追究法律責任。
相關文章
- 錯誤處理
- PHP 錯誤處理PHP
- php錯誤處理PHP
- Go 錯誤處理Go
- Swift錯誤處理Swift
- Zabbix錯誤處理
- mysqldump錯誤處理MySql
- 錯誤處理:如何通過 error、deferred、panic 等處理錯誤?Error
- PHP錯誤處理和異常處理PHP
- go的錯誤處理Go
- Python錯誤處理Python
- 異常錯誤資訊處理
- PHP 核心特性 - 錯誤處理PHP
- 常用模組 PHP 錯誤處理PHP
- laravel9 錯誤處理Laravel
- 淺談前端錯誤處理前端
- Oracle異常錯誤處理Oracle
- ORACLE 異常錯誤處理Oracle
- 15-錯誤處理(Error)Error
- 學習Rust 錯誤處理Rust
- axios 的錯誤處理iOS
- Go語言之錯誤處理Go
- GOLANG錯誤處理最佳方案Golang
- Objective-C:錯誤處理Object
- javascript之處理Ajax錯誤JavaScript
- 搭建dataguard時,錯誤處理
- Oracle錯誤處理思路(一)Oracle
- COM的錯誤處理 (轉)
- 使用PHP錯誤處理 (轉)PHP
- openGauss 處理錯誤表
- Python錯誤處理和異常處理(二)Python
- 【故障處理】CRS-1153錯誤處理
- 【故障處理】ORA-19809錯誤處理
- rust學習十、異常處理(錯誤處理)Rust
- oracle ora-00054錯誤處理Oracle
- 程式錯誤型別及其處理型別
- 請教 Element 的錯誤處理
- Restful API 中的錯誤處理RESTAPI