Originally posted by: SystemAdmin

Hello All...

Currently we have a map that needs to read data from two oracle tables and then transform the data. There are over 300,000 records and the map seems to be performing the transactions in a linear fashion. This is taking way too long as it is taking about 2-3 seconds per transaction. What I need is for the map to read and write transactions in parallel mode but without duplicating data, to cut down on transaction time.

Any ideas?

Thanks,
Kris
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: SystemAdmin

Try using the burst mode in the input card settings (FetchAs)
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

Will this put the map into parallel mode? I want the map to hit the database mutiple times at once without duplicating the data.

Thanks,
Kris
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

Have you got map tracing switched off? This can be a big overhead.

To parallel process, you could read the database in bursts of data, transform the data and output it to a file of unique name. Then have a second map for the database update triggering on a wild card for these files so that you can run multiple instances of the map.
I'm not sure that this would be faster as there will be contention on the database.
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender