Just testing out a scenrio.
I got an updatable mv site and a master site.
i performed some transactions on the mv site .. they reflect back to the master site, after that i disabled the network .. performed some transaction on the mv site .. and wait till the # of failures count to 16, which in returns change the status of my job to be BROKEN.
I used DBMS_JOB.BROKEN with the false parameter . It turns the staus of my job to NORMAL, but the issue in this case is .. i have to manually refresh my refresh group. i set the refresh time to one minute .. it was working fine .. before executing the DBMS_JOB.BROKEN package .. On the other hand if i repeat the same scenrio of disabling the network .. and after # of failures counts to be 16 , it wont change the status of my JOB.. Y?
Just want to know, how to eable my BROKEN JOBS in such a way that i dont have to manually refresh my refresh group.
07-14-2003, 08:17 PM
Have you tried issuing a commit after your DBMS_REFRESH.BROKEN execution.
I know sometimes the snp processes will not recognise a change in a job until a commit has been made...
could be worht a try..
07-15-2003, 01:20 AM
May be try just executing the job as
dbms_job.run(job#). If it executes without any errors, then you have accomplished it.
07-15-2003, 01:22 AM
Yes i tried that out.. issuing commit after DBMS_JOB.BROKEN..
Have to dig out for some other solution. Well .. what is the difference between executing DBMS_JOB.BROKEN and DBMS_JOB.RUN .. its is mentioned in the Oracle Documentation that they both can be used to run the BROKEN JOBS..