asked on

Reg : DB Jobs

Hi

Good Evening

I have faced an issue with db job while we migrate WebMethods6.1 to Webmethods 6.5. When we try to run the db migration scripts. we had db lock on wmlog table. The job also running in the same schema which does the cleanup activity. The job suppose to run at everyday 5 AM. But,we executed the scripts in 7:15 Pm

I would like know the magic behind how the job has been initiated automatically. when shutdown and start the system.

Also, our system doesn't enabled with sql_trace,audit_trace

Please find hereunder the trace file for your kind perusal and advice.

Oracle9i Enterprise Edition Release 9.2.0.7.0 - 64bit Production
With the Partitioning, OLAP and Oracle Data Mining options
JServer Release 9.2.0.7.0 - Production
ORACLE_HOME = /ora/920
System name: AIX
Node name: sss
Release: 3
Version: 5
Machine: 00C6E13E4C00
Instance name: webm
Redo thread mounted by this instance: 1
Oracle process number: 18
Unix process pid: 2936954, image: oracle@sss (J001)

*** SESSION ID:(23.1) 2008-06-14 10:42:55.143
OPIRIP: Uncaught error 1089. Error stack:
ORA-01089: immediate shutdown in progress - no operations are permitted
ORA-00448: normal completion of background process

Thanks in advance

ASKER CERTIFIED SOLUTION

Sean Stuber

membership

Create a free account to see this answer

Signing up is free and takes 30 seconds. No credit card required.

See answer

rengudi

ASKER

Hi,

Actually that job wasnt failed on that day and hardly it tooks 2-3 hrs to complete the job.we scheduled the job at 5 Am everyday. we have shutdown the database at 10:42 , and started up 4:30 PM in the evening.

One more thing we have killed the job forcibly from OS level using kill -9 option for making other script run sucessfully. Actually the job cleaning up the data from the few tables which are old 30 days.

My question here is "By killing the job is there any chance to corrupt the data"?
And is there any chance to get the logs to investigation on this. we are not enable the trace on.

Thanks in Advance.

Sean Stuber

different kinds of corruption could occur. If you have commits in your job and part A commits but part B did not get a chance to before it was killed then your data is now out of synch.
trace would not help you here, except possibly to tell you what step in your job you were on.

depending on what kind of logs you are talking about definitely. If your application writes logs itself (it should) maybe, maybe not. Depends on if those logs were written before the kill
if you're talking about oracle's archive and/or redo logs. it depends on if your database is in archive log mode.