Journal settings from the dspfd screen of the file i want to monitor:
Journal images . . . . . . . . . . . . . . : IMAGES *BOTH
Journal entries to be omitted . . . . . . . : OMTJRNE *OPNCLO
Last journal start date/time . . . . . . . : 07/25/10 01:16:59
Background Info:
Lets call the physical file in question FOO. Lets call the specific field in question City. For some time now I have been receiving complaints from end users that the data in the City field is incorrect for many records in the physical file FOO. No problem, I happen to have a master file that I can update this field with to obtain the correct City field value for every record in file FOO. I proceed to update the file and record with the correct data.
1 month passes. I start getting the same user complaints that the City field has incorrect data. I check the file and some of the values have changed again and become incorrect. The bad data is actually a phone number.
I device a plan to watch the data change from correct to incorrect. I monitor the journal for any insert, update, delete, etc... that would indicate a change to the record. I write that info to a table for easy retrieval. I test my process by modifying various records in various ways like DFU, SQL, RPGLE, etc... Sure enough, my tests show up in my tracking table. I proceed to update table FOO record City with correct data. All the updates I make show up in my tracking table. Another month passes.
Sure enough, the City field again is populated with phone numbers (bad data).
Problem:
There are no corresponding entries in my tracking table for updates made to cause the bad data.
Question:
Is it possible to modify the data in a table without causing journal entries?
My thought is that we have a rogue program that is sitting somewhere, perhaps on the job scheduler, perhaps as a trigger on a table, that is updating this table FOO field City with the wrong data. Unfortunately our system is too large for this to be easily found thru the process of elimination.
Please help :-)
Journal images . . . . . . . . . . . . . . : IMAGES *BOTH
Journal entries to be omitted . . . . . . . : OMTJRNE *OPNCLO
Last journal start date/time . . . . . . . : 07/25/10 01:16:59
Background Info:
Lets call the physical file in question FOO. Lets call the specific field in question City. For some time now I have been receiving complaints from end users that the data in the City field is incorrect for many records in the physical file FOO. No problem, I happen to have a master file that I can update this field with to obtain the correct City field value for every record in file FOO. I proceed to update the file and record with the correct data.
1 month passes. I start getting the same user complaints that the City field has incorrect data. I check the file and some of the values have changed again and become incorrect. The bad data is actually a phone number.
I device a plan to watch the data change from correct to incorrect. I monitor the journal for any insert, update, delete, etc... that would indicate a change to the record. I write that info to a table for easy retrieval. I test my process by modifying various records in various ways like DFU, SQL, RPGLE, etc... Sure enough, my tests show up in my tracking table. I proceed to update table FOO record City with correct data. All the updates I make show up in my tracking table. Another month passes.
Sure enough, the City field again is populated with phone numbers (bad data).
Problem:
There are no corresponding entries in my tracking table for updates made to cause the bad data.
Question:
Is it possible to modify the data in a table without causing journal entries?
My thought is that we have a rogue program that is sitting somewhere, perhaps on the job scheduler, perhaps as a trigger on a table, that is updating this table FOO field City with the wrong data. Unfortunately our system is too large for this to be easily found thru the process of elimination.
Please help :-)
Comment