I am running an infrequent file integration, processing a CSV file into a database. The file contains about 1500 records, which are processed in a loop (for each database record). The memory consumption for this set of records goes up to 5GB which are subsequently not released.

A) is there a way to control memory consumption of the integration service

B) can the memory be released without having to bounce the service?

We are running BT 2022 R6 - the R7 Release notes make no mention of improved memory handling, so I would not want to upgrade "just because".