>In Calvin Hsias blog was an article about fragmented memory, including some code.

Do you have a direct link, please? I checked two blogs so far, both are nice, but I just need some number I can use as my threshold assuming we identified the leak correctly as ctSchedule being the culprit.