I need suggestion on the solution proposed for one of our customers requirement. We are yet to begin the development activities, so trying it out in real environment region is not feasible at this time.
Customer wants a set of data (from 3 different tables) to be sent every time a row is inserted or updated in a particular table. I could use a DB2 trigger to achieve this but wanted to know :
1. How it would perform when there is a real time situation of more than 10,000 data updated in less than a minute.
2. Will the trigger send data one after the other when some other rows are parallely updated
3. Our MIPS consumption is close to 80%, will these triggers eat up more MIPS and cause performance issue for UI applications
4. Some of the application program performs COMMIT only at the end of the program and some programs perform COMMIT after specific count of records. How these trigger behave if programs abends in the above said 2 scenarios.