Manipulating large amounts of data requires enough RAM allocated to the DTM process as well as enough disk space to deal with the staging files and temp data.
Depending on your design, the general recommendation is to always try to shift the heavy lifting to the DBMS side, filtering and sorting should definitely be handled by the DBMS whenever possible.
you can start the tuning of your agent here:
Thank you for your reply, do you mean its always recommended to do transformations or such changes inside the source system itself and then run the integration process through informatica?
1 of 1 people found this helpful
Please note, the design of your mapping might be such that the option of shifting sorting and filtering to the DBMS side is not possible. As in, not the best solution or not compliant to your requirements. In this case Informatica IPS services can help with optimizing the task.
When the source data can be sorted and filtered at its origin, the IICS job will run more efficiently and therefore faster.
For example consider that columns can be dropped before processing them through IICS and rows that should not be processed by the job can be filtered out at the DBMS level.
thanks again, are there any documentations or links that I can refer to regarding this subject other than the link you shared previously. I am looking for documentations that mentions what you said above.
Can you please let us know which source/target connections that you are using for the integration in IICS so that we can assist you with optimization configurations/flags with regards to those specific connectors.